diff --git a/Lib/importlib/__init__.py b/Lib/importlib/__init__.py index 707c081cb2c..a7d57561ead 100644 --- a/Lib/importlib/__init__.py +++ b/Lib/importlib/__init__.py @@ -54,8 +54,6 @@ # Fully bootstrapped at this point, import whatever you like, circular # dependencies and startup overhead minimisation permitting :) -import warnings - # Public API ######################################################### @@ -105,7 +103,7 @@ def reload(module): try: name = module.__name__ except AttributeError: - raise TypeError("reload() argument must be a module") + raise TypeError("reload() argument must be a module") from None if sys.modules.get(name) is not module: raise ImportError(f"module {name} not in sys.modules", name=name) diff --git a/Lib/importlib/_bootstrap.py b/Lib/importlib/_bootstrap.py index 093a0b82456..68d993cacae 100644 --- a/Lib/importlib/_bootstrap.py +++ b/Lib/importlib/_bootstrap.py @@ -53,7 +53,7 @@ def _new_module(name): # For a list that can have a weakref to it. class _List(list): - pass + __slots__ = ("__weakref__",) # Copied from weakref.py with some simplifications and modifications unique to @@ -527,7 +527,7 @@ def _load_module_shim(self, fullname): """ msg = ("the load_module() method is deprecated and slated for removal in " - "Python 3.12; use exec_module() instead") + "Python 3.15; use exec_module() instead") _warnings.warn(msg, DeprecationWarning) spec = spec_from_loader(fullname, self) if fullname in sys.modules: @@ -825,10 +825,16 @@ def _module_repr_from_spec(spec): """Return the repr to use for the module.""" name = '?' if spec.name is None else spec.name if spec.origin is None: - if spec.loader is None: + loader = spec.loader + if loader is None: return f'' + elif ( + _bootstrap_external is not None + and isinstance(loader, _bootstrap_external.NamespaceLoader) + ): + return f'' else: - return f'' + return f'' else: if spec.has_location: return f'' @@ -1129,7 +1135,7 @@ def find_spec(cls, fullname, path=None, target=None): # part of the importer), instead of here (the finder part). # The loader is the usual place to get the data that will # be loaded into the module. (For example, see _LoaderBasics - # in _bootstra_external.py.) Most importantly, this importer + # in _bootstrap_external.py.) Most importantly, this importer # is simpler if we wait to get the data. # However, getting as much data in the finder as possible # to later load the module is okay, and sometimes important. diff --git a/Lib/importlib/_bootstrap_external.py b/Lib/importlib/_bootstrap_external.py index 73ac4405cb5..41f538acb03 100644 --- a/Lib/importlib/_bootstrap_external.py +++ b/Lib/importlib/_bootstrap_external.py @@ -52,7 +52,7 @@ # Bootstrap-related code ###################################################### _CASE_INSENSITIVE_PLATFORMS_STR_KEY = 'win', -_CASE_INSENSITIVE_PLATFORMS_BYTES_KEY = 'cygwin', 'darwin' +_CASE_INSENSITIVE_PLATFORMS_BYTES_KEY = 'cygwin', 'darwin', 'ios', 'tvos', 'watchos' _CASE_INSENSITIVE_PLATFORMS = (_CASE_INSENSITIVE_PLATFORMS_BYTES_KEY + _CASE_INSENSITIVE_PLATFORMS_STR_KEY) @@ -81,6 +81,11 @@ def _pack_uint32(x): return (int(x) & 0xFFFFFFFF).to_bytes(4, 'little') +def _unpack_uint64(data): + """Convert 8 bytes in little-endian to an integer.""" + assert len(data) == 8 + return int.from_bytes(data, 'little') + def _unpack_uint32(data): """Convert 4 bytes in little-endian to an integer.""" assert len(data) == 4 @@ -203,7 +208,7 @@ def _write_atomic(path, data, mode=0o666): try: # We first write data to a temporary file, and then use os.replace() to # perform an atomic rename. - with _io.FileIO(fd, 'wb') as file: + with _io.open(fd, 'wb') as file: file.write(data) _os.replace(path_tmp, path) except OSError: @@ -413,6 +418,7 @@ def _write_atomic(path, data, mode=0o666): # Python 3.11a7 3492 (make POP_JUMP_IF_NONE/NOT_NONE/TRUE/FALSE relative) # Python 3.11a7 3493 (Make JUMP_IF_TRUE_OR_POP/JUMP_IF_FALSE_OR_POP relative) # Python 3.11a7 3494 (New location info table) +# Python 3.11b4 3495 (Set line number of module's RESUME instr to 0 per PEP 626) # Python 3.12a1 3500 (Remove PRECALL opcode) # Python 3.12a1 3501 (YIELD_VALUE oparg == stack_depth) # Python 3.12a1 3502 (LOAD_FAST_CHECK, no NULL-check in LOAD_FAST) @@ -445,8 +451,30 @@ def _write_atomic(path, data, mode=0o666): # Python 3.12b1 3529 (Inline list/dict/set comprehensions) # Python 3.12b1 3530 (Shrink the LOAD_SUPER_ATTR caches) # Python 3.12b1 3531 (Add PEP 695 changes) - -# Python 3.13 will start with 3550 +# Python 3.13a1 3550 (Plugin optimizer support) +# Python 3.13a1 3551 (Compact superinstructions) +# Python 3.13a1 3552 (Remove LOAD_FAST__LOAD_CONST and LOAD_CONST__LOAD_FAST) +# Python 3.13a1 3553 (Add SET_FUNCTION_ATTRIBUTE) +# Python 3.13a1 3554 (more efficient bytecodes for f-strings) +# Python 3.13a1 3555 (generate specialized opcodes metadata from bytecodes.c) +# Python 3.13a1 3556 (Convert LOAD_CLOSURE to a pseudo-op) +# Python 3.13a1 3557 (Make the conversion to boolean in jumps explicit) +# Python 3.13a1 3558 (Reorder the stack items for CALL) +# Python 3.13a1 3559 (Generate opcode IDs from bytecodes.c) +# Python 3.13a1 3560 (Add RESUME_CHECK instruction) +# Python 3.13a1 3561 (Add cache entry to branch instructions) +# Python 3.13a1 3562 (Assign opcode IDs for internal ops in separate range) +# Python 3.13a1 3563 (Add CALL_KW and remove KW_NAMES) +# Python 3.13a1 3564 (Removed oparg from YIELD_VALUE, changed oparg values of RESUME) +# Python 3.13a1 3565 (Oparg of YIELD_VALUE indicates whether it is in a yield-from) +# Python 3.13a1 3566 (Emit JUMP_NO_INTERRUPT instead of JUMP for non-loop no-lineno cases) +# Python 3.13a1 3567 (Reimplement line number propagation by the compiler) +# Python 3.13a1 3568 (Change semantics of END_FOR) +# Python 3.13a5 3569 (Specialize CONTAINS_OP) +# Python 3.13a6 3570 (Add __firstlineno__ class attribute) +# Python 3.13b1 3571 (Fix miscompilation of private names in generic classes) + +# Python 3.14 will start with 3600 # Please don't copy-paste the same pre-release tag for new entries above!!! # You should always use the *upcoming* tag. For example, if 3.12a6 came out @@ -461,7 +489,7 @@ def _write_atomic(path, data, mode=0o666): # Whenever MAGIC_NUMBER is changed, the ranges in the magic_values array # in PC/launcher.c must also be updated. -MAGIC_NUMBER = (3531).to_bytes(2, 'little') + b'\r\n' +MAGIC_NUMBER = (3571).to_bytes(2, 'little') + b'\r\n' _RAW_MAGIC_NUMBER = int.from_bytes(MAGIC_NUMBER, 'little') # For import.c @@ -535,7 +563,8 @@ def cache_from_source(path, debug_override=None, *, optimization=None): # Strip initial drive from a Windows path. We know we have an absolute # path here, so the second part of the check rules out a POSIX path that # happens to contain a colon at the second character. - if head[1] == ':' and head[0] not in path_separators: + # Slicing avoids issues with an empty (or short) `head`. + if head[1:2] == ':' and head[0:1] not in path_separators: head = head[2:] # Strip initial path separator from `head` to complete the conversion @@ -1437,7 +1466,7 @@ class PathFinder: @staticmethod def invalidate_caches(): """Call the invalidate_caches() method on all path entry finders - stored in sys.path_importer_caches (where implemented).""" + stored in sys.path_importer_cache (where implemented).""" for name, finder in list(sys.path_importer_cache.items()): # Drop entry if finder name is a relative path. The current # working directory may have changed. @@ -1449,6 +1478,9 @@ def invalidate_caches(): # https://bugs.python.org/issue45703 _NamespacePath._epoch += 1 + from importlib.metadata import MetadataPathFinder + MetadataPathFinder.invalidate_caches() + @staticmethod def _path_hooks(path): """Search sys.path_hooks for a finder for 'path'.""" @@ -1690,6 +1722,52 @@ def __repr__(self): return f'FileFinder({self.path!r})' +class AppleFrameworkLoader(ExtensionFileLoader): + """A loader for modules that have been packaged as frameworks for + compatibility with Apple's iOS App Store policies. + """ + def create_module(self, spec): + # If the ModuleSpec has been created by the FileFinder, it will have + # been created with an origin pointing to the .fwork file. We need to + # redirect this to the location in the Frameworks folder, using the + # content of the .fwork file. + if spec.origin.endswith(".fwork"): + with _io.FileIO(spec.origin, 'r') as file: + framework_binary = file.read().decode().strip() + bundle_path = _path_split(sys.executable)[0] + spec.origin = _path_join(bundle_path, framework_binary) + + # If the loader is created based on the spec for a loaded module, the + # path will be pointing at the Framework location. If this occurs, + # get the original .fwork location to use as the module's __file__. + if self.path.endswith(".fwork"): + path = self.path + else: + with _io.FileIO(self.path + ".origin", 'r') as file: + origin = file.read().decode().strip() + bundle_path = _path_split(sys.executable)[0] + path = _path_join(bundle_path, origin) + + module = _bootstrap._call_with_frames_removed(_imp.create_dynamic, spec) + + _bootstrap._verbose_message( + "Apple framework extension module {!r} loaded from {!r} (path {!r})", + spec.name, + spec.origin, + path, + ) + + # Ensure that the __file__ points at the .fwork location + try: + module.__file__ = path + except AttributeError: + # Not important enough to report. + # (The error is also ignored in _bootstrap._init_module_attrs or + # import_run_extension in import.c) + pass + + return module + # Import setup ############################################################### def _fix_up_module(ns, name, pathname, cpathname=None): @@ -1722,10 +1800,17 @@ def _get_supported_file_loaders(): Each item is a tuple (loader, suffixes). """ - extensions = ExtensionFileLoader, _imp.extension_suffixes() + extension_loaders = [] + if hasattr(_imp, 'create_dynamic'): + if sys.platform in {"ios", "tvos", "watchos"}: + extension_loaders = [(AppleFrameworkLoader, [ + suffix.replace(".so", ".fwork") + for suffix in _imp.extension_suffixes() + ])] + extension_loaders.append((ExtensionFileLoader, _imp.extension_suffixes())) source = SourceFileLoader, SOURCE_SUFFIXES bytecode = SourcelessFileLoader, BYTECODE_SUFFIXES - return [extensions, source, bytecode] + return extension_loaders + [source, bytecode] def _set_bootstrap_module(_bootstrap_module): diff --git a/Lib/importlib/abc.py b/Lib/importlib/abc.py index b56fa94eb9c..37fef357fe2 100644 --- a/Lib/importlib/abc.py +++ b/Lib/importlib/abc.py @@ -180,7 +180,11 @@ def get_code(self, fullname): else: return self.source_to_code(source, path) -_register(ExecutionLoader, machinery.ExtensionFileLoader) +_register( + ExecutionLoader, + machinery.ExtensionFileLoader, + machinery.AppleFrameworkLoader, +) class FileLoader(_bootstrap_external.FileLoader, ResourceLoader, ExecutionLoader): diff --git a/Lib/importlib/machinery.py b/Lib/importlib/machinery.py index d9a19a13f7b..fbd30b159fb 100644 --- a/Lib/importlib/machinery.py +++ b/Lib/importlib/machinery.py @@ -12,6 +12,7 @@ from ._bootstrap_external import SourceFileLoader from ._bootstrap_external import SourcelessFileLoader from ._bootstrap_external import ExtensionFileLoader +from ._bootstrap_external import AppleFrameworkLoader from ._bootstrap_external import NamespaceLoader diff --git a/Lib/importlib/metadata/__init__.py b/Lib/importlib/metadata/__init__.py index 82e0ce1b281..8ce62dd864f 100644 --- a/Lib/importlib/metadata/__init__.py +++ b/Lib/importlib/metadata/__init__.py @@ -1,9 +1,13 @@ +from __future__ import annotations + import os import re import abc -import csv import sys +import json import email +import types +import inspect import pathlib import zipfile import operator @@ -12,11 +16,9 @@ import functools import itertools import posixpath -import contextlib import collections -import inspect -from . import _adapters, _meta +from . import _meta from ._collections import FreezableDefaultDict, Pair from ._functools import method_cache, pass_none from ._itertools import always_iterable, unique_everseen @@ -26,8 +28,7 @@ from importlib import import_module from importlib.abc import MetaPathFinder from itertools import starmap -from typing import List, Mapping, Optional, cast - +from typing import Any, Iterable, List, Mapping, Match, Optional, Set, cast __all__ = [ 'Distribution', @@ -48,11 +49,11 @@ class PackageNotFoundError(ModuleNotFoundError): """The package was not found.""" - def __str__(self): + def __str__(self) -> str: return f"No package metadata was found for {self.name}" @property - def name(self): + def name(self) -> str: # type: ignore[override] (name,) = self.args return name @@ -118,38 +119,11 @@ def read(text, filter_=None): yield Pair(name, value) @staticmethod - def valid(line): + def valid(line: str): return line and not line.startswith('#') -class DeprecatedTuple: - """ - Provide subscript item access for backward compatibility. - - >>> recwarn = getfixture('recwarn') - >>> ep = EntryPoint(name='name', value='value', group='group') - >>> ep[:] - ('name', 'value', 'group') - >>> ep[0] - 'name' - >>> len(recwarn) - 1 - """ - - # Do not remove prior to 2023-05-01 or Python 3.13 - _warn = functools.partial( - warnings.warn, - "EntryPoint tuple interface is deprecated. Access members by name.", - DeprecationWarning, - stacklevel=2, - ) - - def __getitem__(self, item): - self._warn() - return self._key()[item] - - -class EntryPoint(DeprecatedTuple): +class EntryPoint: """An entry point as defined by Python packaging conventions. See `the packaging docs on entry points @@ -191,34 +165,37 @@ class EntryPoint(DeprecatedTuple): value: str group: str - dist: Optional['Distribution'] = None + dist: Optional[Distribution] = None - def __init__(self, name, value, group): + def __init__(self, name: str, value: str, group: str) -> None: vars(self).update(name=name, value=value, group=group) - def load(self): + def load(self) -> Any: """Load the entry point from its definition. If only a module is indicated by the value, return that module. Otherwise, return the named object. """ - match = self.pattern.match(self.value) + match = cast(Match, self.pattern.match(self.value)) module = import_module(match.group('module')) attrs = filter(None, (match.group('attr') or '').split('.')) return functools.reduce(getattr, attrs, module) @property - def module(self): + def module(self) -> str: match = self.pattern.match(self.value) + assert match is not None return match.group('module') @property - def attr(self): + def attr(self) -> str: match = self.pattern.match(self.value) + assert match is not None return match.group('attr') @property - def extras(self): + def extras(self) -> List[str]: match = self.pattern.match(self.value) + assert match is not None return re.findall(r'\w+', match.group('extras') or '') def _for(self, dist): @@ -266,7 +243,7 @@ def __repr__(self): f'group={self.group!r})' ) - def __hash__(self): + def __hash__(self) -> int: return hash(self._key()) @@ -277,7 +254,7 @@ class EntryPoints(tuple): __slots__ = () - def __getitem__(self, name): # -> EntryPoint: + def __getitem__(self, name: str) -> EntryPoint: # type: ignore[override] """ Get the EntryPoint in self matching name. """ @@ -286,7 +263,14 @@ def __getitem__(self, name): # -> EntryPoint: except StopIteration: raise KeyError(name) - def select(self, **params): + def __repr__(self): + """ + Repr with classname and tuple constructor to + signal that we deviate from regular tuple behavior. + """ + return '%s(%r)' % (self.__class__.__name__, tuple(self)) + + def select(self, **params) -> EntryPoints: """ Select entry points from self that match the given parameters (typically group and/or name). @@ -294,14 +278,14 @@ def select(self, **params): return EntryPoints(ep for ep in self if ep.matches(**params)) @property - def names(self): + def names(self) -> Set[str]: """ Return the set of all names of all entry points. """ return {ep.name for ep in self} @property - def groups(self): + def groups(self) -> Set[str]: """ Return the set of all groups of all entry points. """ @@ -322,28 +306,31 @@ def _from_text(text): class PackagePath(pathlib.PurePosixPath): """A reference to a path in a package""" - def read_text(self, encoding='utf-8'): - with self.locate().open(encoding=encoding) as stream: - return stream.read() + hash: Optional[FileHash] + size: int + dist: Distribution - def read_binary(self): - with self.locate().open('rb') as stream: - return stream.read() + def read_text(self, encoding: str = 'utf-8') -> str: # type: ignore[override] + return self.locate().read_text(encoding=encoding) - def locate(self): + def read_binary(self) -> bytes: + return self.locate().read_bytes() + + def locate(self) -> SimplePath: """Return a path-like object for this path""" return self.dist.locate_file(self) class FileHash: - def __init__(self, spec): + def __init__(self, spec: str) -> None: self.mode, _, self.value = spec.partition('=') - def __repr__(self): + def __repr__(self) -> str: return f'' class DeprecatedNonAbstract: + # Required until Python 3.14 def __new__(cls, *args, **kwargs): all_names = { name for subclass in inspect.getmro(cls) for name in vars(subclass) @@ -363,25 +350,48 @@ def __new__(cls, *args, **kwargs): class Distribution(DeprecatedNonAbstract): - """A Python distribution package.""" + """ + An abstract Python distribution package. + + Custom providers may derive from this class and define + the abstract methods to provide a concrete implementation + for their environment. Some providers may opt to override + the default implementation of some properties to bypass + the file-reading mechanism. + """ @abc.abstractmethod def read_text(self, filename) -> Optional[str]: """Attempt to load metadata file given by the name. + Python distribution metadata is organized by blobs of text + typically represented as "files" in the metadata directory + (e.g. package-1.0.dist-info). These files include things + like: + + - METADATA: The distribution metadata including fields + like Name and Version and Description. + - entry_points.txt: A series of entry points as defined in + `the entry points spec `_. + - RECORD: A record of files according to + `this recording spec `_. + + A package may provide any set of files, including those + not listed here or none at all. + :param filename: The name of the file in the distribution info. :return: The text if found, otherwise None. """ @abc.abstractmethod - def locate_file(self, path): + def locate_file(self, path: str | os.PathLike[str]) -> SimplePath: """ - Given a path to a file in this distribution, return a path + Given a path to a file in this distribution, return a SimplePath to it. """ @classmethod - def from_name(cls, name: str): + def from_name(cls, name: str) -> Distribution: """Return the Distribution for the given package name. :param name: The name of the distribution package to search for. @@ -394,21 +404,23 @@ def from_name(cls, name: str): if not name: raise ValueError("A distribution name is required.") try: - return next(cls.discover(name=name)) + return next(iter(cls.discover(name=name))) except StopIteration: raise PackageNotFoundError(name) @classmethod - def discover(cls, **kwargs): + def discover( + cls, *, context: Optional[DistributionFinder.Context] = None, **kwargs + ) -> Iterable[Distribution]: """Return an iterable of Distribution objects for all packages. Pass a ``context`` or pass keyword arguments for constructing a context. :context: A ``DistributionFinder.Context`` object. - :return: Iterable of Distribution objects for all packages. + :return: Iterable of Distribution objects for packages matching + the context. """ - context = kwargs.pop('context', None) if context and kwargs: raise ValueError("cannot accept context and kwargs") context = context or DistributionFinder.Context(**kwargs) @@ -417,8 +429,8 @@ def discover(cls, **kwargs): ) @staticmethod - def at(path): - """Return a Distribution for the indicated metadata path + def at(path: str | os.PathLike[str]) -> Distribution: + """Return a Distribution for the indicated metadata path. :param path: a string or path-like object :return: a concrete Distribution instance for the path @@ -427,7 +439,7 @@ def at(path): @staticmethod def _discover_resolvers(): - """Search the meta_path for resolvers.""" + """Search the meta_path for resolvers (MetadataPathFinders).""" declared = ( getattr(finder, 'find_distributions', None) for finder in sys.meta_path ) @@ -438,8 +450,15 @@ def metadata(self) -> _meta.PackageMetadata: """Return the parsed metadata for this Distribution. The returned object will have keys that name the various bits of - metadata. See PEP 566 for details. + metadata per the + `Core metadata specifications `_. + + Custom providers may provide the METADATA file or override this + property. """ + # deferred for performance (python/cpython#109829) + from . import _adapters + opt_text = ( self.read_text('METADATA') or self.read_text('PKG-INFO') @@ -452,7 +471,7 @@ def metadata(self) -> _meta.PackageMetadata: return _adapters.Message(email.message_from_string(text)) @property - def name(self): + def name(self) -> str: """Return the 'Name' metadata for the distribution package.""" return self.metadata['Name'] @@ -462,16 +481,22 @@ def _normalized_name(self): return Prepared.normalize(self.name) @property - def version(self): + def version(self) -> str: """Return the 'Version' metadata for the distribution package.""" return self.metadata['Version'] @property - def entry_points(self): + def entry_points(self) -> EntryPoints: + """ + Return EntryPoints for this distribution. + + Custom providers may provide the ``entry_points.txt`` file + or override this property. + """ return EntryPoints._from_text_for(self.read_text('entry_points.txt'), self) @property - def files(self): + def files(self) -> Optional[List[PackagePath]]: """Files in this distribution. :return: List of PackagePath for this distribution or None @@ -480,6 +505,10 @@ def files(self): (i.e. RECORD for dist-info, or installed-files.txt or SOURCES.txt for egg-info) is missing. Result may be empty if the metadata exists but is empty. + + Custom providers are recommended to provide a "RECORD" file (in + ``read_text``) or override this property to allow for callers to be + able to resolve filenames provided by the package. """ def make_file(name, hash=None, size_str=None): @@ -491,6 +520,10 @@ def make_file(name, hash=None, size_str=None): @pass_none def make_files(lines): + # Delay csv import, since Distribution.files is not as widely used + # as other parts of importlib.metadata + import csv + return starmap(make_file, csv.reader(lines)) @pass_none @@ -507,7 +540,7 @@ def skip_missing_files(package_paths): def _read_files_distinfo(self): """ - Read the lines of RECORD + Read the lines of RECORD. """ text = self.read_text('RECORD') return text and text.splitlines() @@ -534,7 +567,7 @@ def _read_files_egginfo_installed(self): paths = ( (subdir / name) .resolve() - .relative_to(self.locate_file('').resolve()) + .relative_to(self.locate_file('').resolve(), walk_up=True) .as_posix() for name in text.splitlines() ) @@ -556,7 +589,7 @@ def _read_files_egginfo_sources(self): return text and map('"{}"'.format, text.splitlines()) @property - def requires(self): + def requires(self) -> Optional[List[str]]: """Generated requirements specified for this Distribution""" reqs = self._read_dist_info_reqs() or self._read_egg_info_reqs() return reqs and list(reqs) @@ -607,10 +640,23 @@ def url_req_space(req): space = url_req_space(section.value) yield section.value + space + quoted_marker(section.name) + @property + def origin(self): + return self._load_json('direct_url.json') + + def _load_json(self, filename): + return pass_none(json.loads)( + self.read_text(filename), + object_hook=lambda data: types.SimpleNamespace(**data), + ) + class DistributionFinder(MetaPathFinder): """ A MetaPathFinder capable of discovering installed distributions. + + Custom providers should implement this interface in order to + supply metadata. """ class Context: @@ -623,6 +669,17 @@ class Context: Each DistributionFinder may expect any parameters and should attempt to honor the canonical parameters defined below when appropriate. + + This mechanism gives a custom provider a means to + solicit additional details from the caller beyond + "name" and "path" when searching distributions. + For example, imagine a provider that exposes suites + of packages in either a "public" or "private" ``realm``. + A caller may wish to query only for distributions in + a particular realm and could call + ``distributions(realm="private")`` to signal to the + custom provider to only include distributions from that + realm. """ name = None @@ -635,7 +692,7 @@ def __init__(self, **kwargs): vars(self).update(kwargs) @property - def path(self): + def path(self) -> List[str]: """ The sequence of directory path that a distribution finder should search. @@ -646,7 +703,7 @@ def path(self): return vars(self).get('path', sys.path) @abc.abstractmethod - def find_distributions(self, context=Context()): + def find_distributions(self, context=Context()) -> Iterable[Distribution]: """ Find distributions. @@ -658,11 +715,18 @@ def find_distributions(self, context=Context()): class FastPath: """ - Micro-optimized class for searching a path for - children. + Micro-optimized class for searching a root for children. + + Root is a path on the file system that may contain metadata + directories either as natural directories or within a zip file. >>> FastPath('').children() ['...'] + + FastPath objects are cached and recycled for any given root. + + >>> FastPath('foobar') is FastPath('foobar') + True """ @functools.lru_cache() # type: ignore @@ -704,7 +768,19 @@ def lookup(self, mtime): class Lookup: + """ + A micro-optimized class for searching a (fast) path for metadata. + """ + def __init__(self, path: FastPath): + """ + Calculate all of the children representing metadata. + + From the children in the path, calculate early all of the + children that appear to represent metadata (infos) or legacy + metadata (eggs). + """ + base = os.path.basename(path.root).lower() base_is_egg = base.endswith(".egg") self.infos = FreezableDefaultDict(list) @@ -725,7 +801,10 @@ def __init__(self, path: FastPath): self.infos.freeze() self.eggs.freeze() - def search(self, prepared): + def search(self, prepared: Prepared): + """ + Yield all infos and eggs matching the Prepared query. + """ infos = ( self.infos[prepared.normalized] if prepared @@ -741,13 +820,28 @@ def search(self, prepared): class Prepared: """ - A prepared search for metadata on a possibly-named package. + A prepared search query for metadata on a possibly-named package. + + Pre-calculates the normalization to prevent repeated operations. + + >>> none = Prepared(None) + >>> none.normalized + >>> none.legacy_normalized + >>> bool(none) + False + >>> sample = Prepared('Sample__Pkg-name.foo') + >>> sample.normalized + 'sample_pkg_name_foo' + >>> sample.legacy_normalized + 'sample__pkg_name.foo' + >>> bool(sample) + True """ normalized = None legacy_normalized = None - def __init__(self, name): + def __init__(self, name: Optional[str]): self.name = name if name is None: return @@ -775,7 +869,9 @@ def __bool__(self): class MetadataPathFinder(DistributionFinder): @classmethod - def find_distributions(cls, context=DistributionFinder.Context()): + def find_distributions( + cls, context=DistributionFinder.Context() + ) -> Iterable[PathDistribution]: """ Find distributions. @@ -795,19 +891,20 @@ def _search_paths(cls, name, paths): path.search(prepared) for path in map(FastPath, paths) ) - def invalidate_caches(cls): + @classmethod + def invalidate_caches(cls) -> None: FastPath.__new__.cache_clear() class PathDistribution(Distribution): - def __init__(self, path: SimplePath): + def __init__(self, path: SimplePath) -> None: """Construct a distribution. :param path: SimplePath indicating the metadata directory. """ self._path = path - def read_text(self, filename): + def read_text(self, filename: str | os.PathLike[str]) -> Optional[str]: with suppress( FileNotFoundError, IsADirectoryError, @@ -817,9 +914,11 @@ def read_text(self, filename): ): return self._path.joinpath(filename).read_text(encoding='utf-8') + return None + read_text.__doc__ = Distribution.read_text.__doc__ - def locate_file(self, path): + def locate_file(self, path: str | os.PathLike[str]) -> SimplePath: return self._path.parent / path @property @@ -852,7 +951,7 @@ def _name_from_stem(stem): return name -def distribution(distribution_name): +def distribution(distribution_name: str) -> Distribution: """Get the ``Distribution`` instance for the named package. :param distribution_name: The name of the distribution package as a string. @@ -861,7 +960,7 @@ def distribution(distribution_name): return Distribution.from_name(distribution_name) -def distributions(**kwargs): +def distributions(**kwargs) -> Iterable[Distribution]: """Get all ``Distribution`` instances in the current environment. :return: An iterable of ``Distribution`` instances. @@ -869,7 +968,7 @@ def distributions(**kwargs): return Distribution.discover(**kwargs) -def metadata(distribution_name) -> _meta.PackageMetadata: +def metadata(distribution_name: str) -> _meta.PackageMetadata: """Get the metadata for the named package. :param distribution_name: The name of the distribution package to query. @@ -878,7 +977,7 @@ def metadata(distribution_name) -> _meta.PackageMetadata: return Distribution.from_name(distribution_name).metadata -def version(distribution_name): +def version(distribution_name: str) -> str: """Get the version string for the named package. :param distribution_name: The name of the distribution package to query. @@ -912,7 +1011,7 @@ def entry_points(**params) -> EntryPoints: return EntryPoints(eps).select(**params) -def files(distribution_name): +def files(distribution_name: str) -> Optional[List[PackagePath]]: """Return a list of files for the named package. :param distribution_name: The name of the distribution package to query. @@ -921,11 +1020,11 @@ def files(distribution_name): return distribution(distribution_name).files -def requires(distribution_name): +def requires(distribution_name: str) -> Optional[List[str]]: """ Return a list of requirements for the named package. - :return: An iterator of requirements, suitable for + :return: An iterable of requirements, suitable for packaging.requirement.Requirement. """ return distribution(distribution_name).requires @@ -952,13 +1051,42 @@ def _top_level_declared(dist): return (dist.read_text('top_level.txt') or '').split() +def _topmost(name: PackagePath) -> Optional[str]: + """ + Return the top-most parent as long as there is a parent. + """ + top, *rest = name.parts + return top if rest else None + + +def _get_toplevel_name(name: PackagePath) -> str: + """ + Infer a possibly importable module name from a name presumed on + sys.path. + + >>> _get_toplevel_name(PackagePath('foo.py')) + 'foo' + >>> _get_toplevel_name(PackagePath('foo')) + 'foo' + >>> _get_toplevel_name(PackagePath('foo.pyc')) + 'foo' + >>> _get_toplevel_name(PackagePath('foo/__init__.py')) + 'foo' + >>> _get_toplevel_name(PackagePath('foo.pth')) + 'foo.pth' + >>> _get_toplevel_name(PackagePath('foo.dist-info')) + 'foo.dist-info' + """ + return _topmost(name) or ( + # python/typeshed#10328 + inspect.getmodulename(name) # type: ignore + or str(name) + ) + + def _top_level_inferred(dist): - opt_names = { - f.parts[0] if len(f.parts) > 1 else inspect.getmodulename(f) - for f in always_iterable(dist.files) - } + opt_names = set(map(_get_toplevel_name, always_iterable(dist.files))) - @pass_none def importable_name(name): return '.' not in name diff --git a/Lib/importlib/metadata/_adapters.py b/Lib/importlib/metadata/_adapters.py index 6aed69a3085..59116880895 100644 --- a/Lib/importlib/metadata/_adapters.py +++ b/Lib/importlib/metadata/_adapters.py @@ -53,7 +53,7 @@ def __iter__(self): def __getitem__(self, item): """ Warn users that a ``KeyError`` can be expected when a - mising key is supplied. Ref python/importlib_metadata#371. + missing key is supplied. Ref python/importlib_metadata#371. """ res = super().__getitem__(item) if res is None: diff --git a/Lib/importlib/metadata/_meta.py b/Lib/importlib/metadata/_meta.py index c9a7ef906a8..1927d0f624d 100644 --- a/Lib/importlib/metadata/_meta.py +++ b/Lib/importlib/metadata/_meta.py @@ -1,3 +1,6 @@ +from __future__ import annotations + +import os from typing import Protocol from typing import Any, Dict, Iterator, List, Optional, TypeVar, Union, overload @@ -6,30 +9,27 @@ class PackageMetadata(Protocol): - def __len__(self) -> int: - ... # pragma: no cover + def __len__(self) -> int: ... # pragma: no cover - def __contains__(self, item: str) -> bool: - ... # pragma: no cover + def __contains__(self, item: str) -> bool: ... # pragma: no cover - def __getitem__(self, key: str) -> str: - ... # pragma: no cover + def __getitem__(self, key: str) -> str: ... # pragma: no cover - def __iter__(self) -> Iterator[str]: - ... # pragma: no cover + def __iter__(self) -> Iterator[str]: ... # pragma: no cover @overload - def get(self, name: str, failobj: None = None) -> Optional[str]: - ... # pragma: no cover + def get( + self, name: str, failobj: None = None + ) -> Optional[str]: ... # pragma: no cover @overload - def get(self, name: str, failobj: _T) -> Union[str, _T]: - ... # pragma: no cover + def get(self, name: str, failobj: _T) -> Union[str, _T]: ... # pragma: no cover # overload per python/importlib_metadata#435 @overload - def get_all(self, name: str, failobj: None = None) -> Optional[List[Any]]: - ... # pragma: no cover + def get_all( + self, name: str, failobj: None = None + ) -> Optional[List[Any]]: ... # pragma: no cover @overload def get_all(self, name: str, failobj: _T) -> Union[List[Any], _T]: @@ -44,20 +44,24 @@ def json(self) -> Dict[str, Union[str, List[str]]]: """ -class SimplePath(Protocol[_T]): +class SimplePath(Protocol): """ - A minimal subset of pathlib.Path required by PathDistribution. + A minimal subset of pathlib.Path required by Distribution. """ - def joinpath(self) -> _T: - ... # pragma: no cover + def joinpath( + self, other: Union[str, os.PathLike[str]] + ) -> SimplePath: ... # pragma: no cover - def __truediv__(self, other: Union[str, _T]) -> _T: - ... # pragma: no cover + def __truediv__( + self, other: Union[str, os.PathLike[str]] + ) -> SimplePath: ... # pragma: no cover @property - def parent(self) -> _T: - ... # pragma: no cover + def parent(self) -> SimplePath: ... # pragma: no cover + + def read_text(self, encoding=None) -> str: ... # pragma: no cover + + def read_bytes(self) -> bytes: ... # pragma: no cover - def read_text(self) -> str: - ... # pragma: no cover + def exists(self) -> bool: ... # pragma: no cover diff --git a/Lib/importlib/metadata/diagnose.py b/Lib/importlib/metadata/diagnose.py new file mode 100644 index 00000000000..e405471ac4d --- /dev/null +++ b/Lib/importlib/metadata/diagnose.py @@ -0,0 +1,21 @@ +import sys + +from . import Distribution + + +def inspect(path): + print("Inspecting", path) + dists = list(Distribution.discover(path=[path])) + if not dists: + return + print("Found", len(dists), "packages:", end=' ') + print(', '.join(dist.name for dist in dists)) + + +def run(): + for path in sys.path: + inspect(path) + + +if __name__ == '__main__': + run() diff --git a/Lib/importlib/resources/__init__.py b/Lib/importlib/resources/__init__.py index 34e3a9950cc..723c9f9eb33 100644 --- a/Lib/importlib/resources/__init__.py +++ b/Lib/importlib/resources/__init__.py @@ -1,20 +1,27 @@ -"""Read resources contained within a package.""" +""" +Read resources contained within a package. + +This codebase is shared between importlib.resources in the stdlib +and importlib_resources in PyPI. See +https://github.com/python/importlib_metadata/wiki/Development-Methodology +for more detail. +""" from ._common import ( as_file, files, Package, + Anchor, ) -from ._legacy import ( +from ._functional import ( contents, + is_resource, open_binary, - read_binary, open_text, - read_text, - is_resource, path, - Resource, + read_binary, + read_text, ) from .abc import ResourceReader @@ -22,11 +29,11 @@ __all__ = [ 'Package', - 'Resource', + 'Anchor', 'ResourceReader', 'as_file', - 'contents', 'files', + 'contents', 'is_resource', 'open_binary', 'open_text', diff --git a/Lib/importlib/resources/_common.py b/Lib/importlib/resources/_common.py index b402e05116e..d9128266a2e 100644 --- a/Lib/importlib/resources/_common.py +++ b/Lib/importlib/resources/_common.py @@ -12,8 +12,6 @@ from typing import Union, Optional, cast from .abc import ResourceReader, Traversable -from ._adapters import wrap_spec - Package = Union[types.ModuleType, str] Anchor = Package @@ -27,6 +25,8 @@ def package_to_anchor(func): >>> files('a', 'b') Traceback (most recent call last): TypeError: files() takes from 0 to 1 positional arguments but 2 were given + + Remove this compatibility in Python 3.14. """ undefined = object() @@ -66,10 +66,10 @@ def get_resource_reader(package: types.ModuleType) -> Optional[ResourceReader]: # zipimport.zipimporter does not support weak references, resulting in a # TypeError. That seems terrible. spec = package.__spec__ - reader = getattr(spec.loader, 'get_resource_reader', None) # type: ignore + reader = getattr(spec.loader, 'get_resource_reader', None) # type: ignore[union-attr] if reader is None: return None - return reader(spec.name) # type: ignore + return reader(spec.name) # type: ignore[union-attr] @functools.singledispatch @@ -93,12 +93,13 @@ def _infer_caller(): """ def is_this_file(frame_info): - return frame_info.filename == __file__ + return frame_info.filename == stack[0].filename def is_wrapper(frame_info): return frame_info.function == 'wrapper' - not_this_file = itertools.filterfalse(is_this_file, inspect.stack()) + stack = inspect.stack() + not_this_file = itertools.filterfalse(is_this_file, stack) # also exclude 'wrapper' due to singledispatch in the call stack callers = itertools.filterfalse(is_wrapper, not_this_file) return next(callers).frame @@ -109,6 +110,9 @@ def from_package(package: types.ModuleType): Return a Traversable object for the given package. """ + # deferred for performance (python/cpython#109829) + from ._adapters import wrap_spec + spec = wrap_spec(package) reader = spec.loader.get_resource_reader(spec.name) return reader.files() diff --git a/Lib/importlib/resources/_functional.py b/Lib/importlib/resources/_functional.py new file mode 100644 index 00000000000..f59416f2dd6 --- /dev/null +++ b/Lib/importlib/resources/_functional.py @@ -0,0 +1,81 @@ +"""Simplified function-based API for importlib.resources""" + +import warnings + +from ._common import files, as_file + + +_MISSING = object() + + +def open_binary(anchor, *path_names): + """Open for binary reading the *resource* within *package*.""" + return _get_resource(anchor, path_names).open('rb') + + +def open_text(anchor, *path_names, encoding=_MISSING, errors='strict'): + """Open for text reading the *resource* within *package*.""" + encoding = _get_encoding_arg(path_names, encoding) + resource = _get_resource(anchor, path_names) + return resource.open('r', encoding=encoding, errors=errors) + + +def read_binary(anchor, *path_names): + """Read and return contents of *resource* within *package* as bytes.""" + return _get_resource(anchor, path_names).read_bytes() + + +def read_text(anchor, *path_names, encoding=_MISSING, errors='strict'): + """Read and return contents of *resource* within *package* as str.""" + encoding = _get_encoding_arg(path_names, encoding) + resource = _get_resource(anchor, path_names) + return resource.read_text(encoding=encoding, errors=errors) + + +def path(anchor, *path_names): + """Return the path to the *resource* as an actual file system path.""" + return as_file(_get_resource(anchor, path_names)) + + +def is_resource(anchor, *path_names): + """Return ``True`` if there is a resource named *name* in the package, + + Otherwise returns ``False``. + """ + return _get_resource(anchor, path_names).is_file() + + +def contents(anchor, *path_names): + """Return an iterable over the named resources within the package. + + The iterable returns :class:`str` resources (e.g. files). + The iterable does not recurse into subdirectories. + """ + warnings.warn( + "importlib.resources.contents is deprecated. " + "Use files(anchor).iterdir() instead.", + DeprecationWarning, + stacklevel=1, + ) + return (resource.name for resource in _get_resource(anchor, path_names).iterdir()) + + +def _get_encoding_arg(path_names, encoding): + # For compatibility with versions where *encoding* was a positional + # argument, it needs to be given explicitly when there are multiple + # *path_names*. + # This limitation can be removed in Python 3.15. + if encoding is _MISSING: + if len(path_names) > 1: + raise TypeError( + "'encoding' argument required with multiple path names", + ) + else: + return 'utf-8' + return encoding + + +def _get_resource(anchor, path_names): + if anchor is None: + raise TypeError("anchor must be module or string, got None") + return files(anchor).joinpath(*path_names) diff --git a/Lib/importlib/resources/readers.py b/Lib/importlib/resources/readers.py index c3cdf769cbe..70fc7e2b9c0 100644 --- a/Lib/importlib/resources/readers.py +++ b/Lib/importlib/resources/readers.py @@ -1,8 +1,14 @@ +from __future__ import annotations + import collections +import contextlib import itertools import pathlib import operator +import re +import warnings import zipfile +from collections.abc import Iterator from . import abc @@ -31,8 +37,10 @@ def files(self): class ZipReader(abc.TraversableResources): def __init__(self, loader, module): - _, _, name = module.rpartition('.') - self.prefix = loader.prefix.replace('\\', '/') + name + '/' + self.prefix = loader.prefix.replace('\\', '/') + if loader.is_package(module): + _, _, name = module.rpartition('.') + self.prefix += name + '/' self.archive = loader.archive def open_resource(self, resource): @@ -62,7 +70,7 @@ class MultiplexedPath(abc.Traversable): """ def __init__(self, *paths): - self._paths = list(map(pathlib.Path, remove_duplicates(paths))) + self._paths = list(map(_ensure_traversable, remove_duplicates(paths))) if not self._paths: message = 'MultiplexedPath must contain at least one path' raise FileNotFoundError(message) @@ -130,7 +138,40 @@ class NamespaceReader(abc.TraversableResources): def __init__(self, namespace_path): if 'NamespacePath' not in str(namespace_path): raise ValueError('Invalid path') - self.path = MultiplexedPath(*list(namespace_path)) + self.path = MultiplexedPath(*filter(bool, map(self._resolve, namespace_path))) + + @classmethod + def _resolve(cls, path_str) -> abc.Traversable | None: + r""" + Given an item from a namespace path, resolve it to a Traversable. + + path_str might be a directory on the filesystem or a path to a + zipfile plus the path within the zipfile, e.g. ``/foo/bar`` or + ``/foo/baz.zip/inner_dir`` or ``foo\baz.zip\inner_dir\sub``. + + path_str might also be a sentinel used by editable packages to + trigger other behaviors (see python/importlib_resources#311). + In that case, return None. + """ + dirs = (cand for cand in cls._candidate_paths(path_str) if cand.is_dir()) + return next(dirs, None) + + @classmethod + def _candidate_paths(cls, path_str: str) -> Iterator[abc.Traversable]: + yield pathlib.Path(path_str) + yield from cls._resolve_zip_path(path_str) + + @staticmethod + def _resolve_zip_path(path_str: str): + for match in reversed(list(re.finditer(r'[\\/]', path_str))): + with contextlib.suppress( + FileNotFoundError, + IsADirectoryError, + NotADirectoryError, + PermissionError, + ): + inner = path_str[match.end() :].replace('\\', '/') + '/' + yield zipfile.Path(path_str[: match.start()], inner.lstrip('/')) def resource_path(self, resource): """ @@ -142,3 +183,21 @@ def resource_path(self, resource): def files(self): return self.path + + +def _ensure_traversable(path): + """ + Convert deprecated string arguments to traversables (pathlib.Path). + + Remove with Python 3.15. + """ + if not isinstance(path, str): + return path + + warnings.warn( + "String arguments are deprecated. Pass a Traversable instead.", + DeprecationWarning, + stacklevel=3, + ) + + return pathlib.Path(path) diff --git a/Lib/importlib/resources/simple.py b/Lib/importlib/resources/simple.py index 7770c922c84..2e75299b13a 100644 --- a/Lib/importlib/resources/simple.py +++ b/Lib/importlib/resources/simple.py @@ -77,7 +77,7 @@ class ResourceHandle(Traversable): def __init__(self, parent: ResourceContainer, name: str): self.parent = parent - self.name = name # type: ignore + self.name = name # type: ignore[misc] def is_file(self): return True @@ -88,7 +88,7 @@ def is_dir(self): def open(self, mode='r', *args, **kwargs): stream = self.parent.reader.open_binary(self.name) if 'b' not in mode: - stream = io.TextIOWrapper(*args, **kwargs) + stream = io.TextIOWrapper(stream, *args, **kwargs) return stream def joinpath(self, name): diff --git a/Lib/importlib/util.py b/Lib/importlib/util.py index f4d6e823315..284206b62f9 100644 --- a/Lib/importlib/util.py +++ b/Lib/importlib/util.py @@ -135,7 +135,7 @@ class _incompatible_extension_module_restrictions: may not be imported in a subinterpreter. That implies modules that do not implement multi-phase init or that explicitly of out. - Likewise for modules import in a subinterpeter with its own GIL + Likewise for modules import in a subinterpreter with its own GIL when the extension does not support a per-interpreter GIL. This implies the module does not have a Py_mod_multiple_interpreters slot set to Py_MOD_PER_INTERPRETER_GIL_SUPPORTED. @@ -145,7 +145,7 @@ class _incompatible_extension_module_restrictions: You can get the same effect as this function by implementing the basic interface of multi-phase init (PEP 489) and lying about - support for mulitple interpreters (or per-interpreter GIL). + support for multiple interpreters (or per-interpreter GIL). """ def __init__(self, *, disable_check): @@ -171,36 +171,57 @@ class _LazyModule(types.ModuleType): def __getattribute__(self, attr): """Trigger the load of the module and return the attribute.""" - # All module metadata must be garnered from __spec__ in order to avoid - # using mutated values. - # Stop triggering this method. - self.__class__ = types.ModuleType - # Get the original name to make sure no object substitution occurred - # in sys.modules. - original_name = self.__spec__.name - # Figure out exactly what attributes were mutated between the creation - # of the module and now. - attrs_then = self.__spec__.loader_state['__dict__'] - attrs_now = self.__dict__ - attrs_updated = {} - for key, value in attrs_now.items(): - # Code that set the attribute may have kept a reference to the - # assigned object, making identity more important than equality. - if key not in attrs_then: - attrs_updated[key] = value - elif id(attrs_now[key]) != id(attrs_then[key]): - attrs_updated[key] = value - self.__spec__.loader.exec_module(self) - # If exec_module() was used directly there is no guarantee the module - # object was put into sys.modules. - if original_name in sys.modules: - if id(self) != id(sys.modules[original_name]): - raise ValueError(f"module object for {original_name!r} " - "substituted in sys.modules during a lazy " - "load") - # Update after loading since that's what would happen in an eager - # loading situation. - self.__dict__.update(attrs_updated) + __spec__ = object.__getattribute__(self, '__spec__') + loader_state = __spec__.loader_state + with loader_state['lock']: + # Only the first thread to get the lock should trigger the load + # and reset the module's class. The rest can now getattr(). + if object.__getattribute__(self, '__class__') is _LazyModule: + __class__ = loader_state['__class__'] + + # Reentrant calls from the same thread must be allowed to proceed without + # triggering the load again. + # exec_module() and self-referential imports are the primary ways this can + # happen, but in any case we must return something to avoid deadlock. + if loader_state['is_loading']: + return __class__.__getattribute__(self, attr) + loader_state['is_loading'] = True + + __dict__ = __class__.__getattribute__(self, '__dict__') + + # All module metadata must be gathered from __spec__ in order to avoid + # using mutated values. + # Get the original name to make sure no object substitution occurred + # in sys.modules. + original_name = __spec__.name + # Figure out exactly what attributes were mutated between the creation + # of the module and now. + attrs_then = loader_state['__dict__'] + attrs_now = __dict__ + attrs_updated = {} + for key, value in attrs_now.items(): + # Code that set an attribute may have kept a reference to the + # assigned object, making identity more important than equality. + if key not in attrs_then: + attrs_updated[key] = value + elif id(attrs_now[key]) != id(attrs_then[key]): + attrs_updated[key] = value + __spec__.loader.exec_module(self) + # If exec_module() was used directly there is no guarantee the module + # object was put into sys.modules. + if original_name in sys.modules: + if id(self) != id(sys.modules[original_name]): + raise ValueError(f"module object for {original_name!r} " + "substituted in sys.modules during a lazy " + "load") + # Update after loading since that's what would happen in an eager + # loading situation. + __dict__.update(attrs_updated) + # Finally, stop triggering this method, if the module did not + # already update its own __class__. + if isinstance(self, _LazyModule): + object.__setattr__(self, '__class__', __class__) + return getattr(self, attr) def __delattr__(self, attr): @@ -235,6 +256,9 @@ def create_module(self, spec): def exec_module(self, module): """Make the module load lazily.""" + # Threading is only needed for lazy loading, and importlib.util can + # be pulled in at interpreter startup, so defer until needed. + import threading module.__spec__.loader = self.loader module.__loader__ = self.loader # Don't need to worry about deep-copying as trying to set an attribute @@ -244,5 +268,7 @@ def exec_module(self, module): loader_state = {} loader_state['__dict__'] = module.__dict__.copy() loader_state['__class__'] = module.__class__ + loader_state['lock'] = threading.RLock() + loader_state['is_loading'] = False module.__spec__.loader_state = loader_state module.__class__ = _LazyModule diff --git a/Lib/nturl2path.py b/Lib/nturl2path.py index 61852aff589..757fd01bec8 100644 --- a/Lib/nturl2path.py +++ b/Lib/nturl2path.py @@ -15,32 +15,29 @@ def url2pathname(url): # become # C:\foo\bar\spam.foo import string, urllib.parse + if url[:3] == '///': + # URL has an empty authority section, so the path begins on the third + # character. + url = url[2:] + elif url[:12] == '//localhost/': + # Skip past 'localhost' authority. + url = url[11:] + if url[:3] == '///': + # Skip past extra slash before UNC drive in URL path. + url = url[1:] # Windows itself uses ":" even in URLs. url = url.replace(':', '|') if not '|' in url: # No drive specifier, just convert slashes - if url[:4] == '////': - # path is something like ////host/path/on/remote/host - # convert this to \\host\path\on\remote\host - # (notice halving of slashes at the start of the path) - url = url[2:] - components = url.split('/') # make sure not to convert quoted slashes :-) - return urllib.parse.unquote('\\'.join(components)) + return urllib.parse.unquote(url.replace('/', '\\')) comp = url.split('|') if len(comp) != 2 or comp[0][-1] not in string.ascii_letters: error = 'Bad URL: ' + url raise OSError(error) drive = comp[0][-1].upper() - components = comp[1].split('/') - path = drive + ':' - for comp in components: - if comp: - path = path + '\\' + urllib.parse.unquote(comp) - # Issue #11474 - handing url such as |c/| - if path.endswith(':') and url.endswith('/'): - path += '\\' - return path + tail = urllib.parse.unquote(comp[1].replace('/', '\\')) + return drive + ':' + tail def pathname2url(p): """OS-specific conversion from a file system path to a relative URL @@ -52,30 +49,21 @@ def pathname2url(p): import urllib.parse # First, clean up some special forms. We are going to sacrifice # the additional information anyway - if p[:4] == '\\\\?\\': + p = p.replace('\\', '/') + if p[:4] == '//?/': p = p[4:] - if p[:4].upper() == 'UNC\\': - p = '\\' + p[4:] + if p[:4].upper() == 'UNC/': + p = '//' + p[4:] elif p[1:2] != ':': raise OSError('Bad path: ' + p) if not ':' in p: - # No drive specifier, just convert slashes and quote the name - if p[:2] == '\\\\': - # path is something like \\host\path\on\remote\host - # convert this to ////host/path/on/remote/host - # (notice doubling of slashes at the start of the path) - p = '\\\\' + p - components = p.split('\\') - return urllib.parse.quote('/'.join(components)) + # No DOS drive specified, just quote the pathname + return urllib.parse.quote(p) comp = p.split(':', maxsplit=2) if len(comp) != 2 or len(comp[0]) > 1: error = 'Bad path: ' + p raise OSError(error) drive = urllib.parse.quote(comp[0].upper()) - components = comp[1].split('\\') - path = '///' + drive + ':' - for comp in components: - if comp: - path = path + '/' + urllib.parse.quote(comp) - return path + tail = urllib.parse.quote(comp[1]) + return '///' + drive + ':' + tail diff --git a/Lib/pathlib.py b/Lib/pathlib.py deleted file mode 100644 index bd5a096f9e3..00000000000 --- a/Lib/pathlib.py +++ /dev/null @@ -1,1435 +0,0 @@ -"""Object-oriented filesystem paths. - -This module provides classes to represent abstract paths and concrete -paths with operations that have semantics appropriate for different -operating systems. -""" - -import fnmatch -import functools -import io -import ntpath -import os -import posixpath -import re -import sys -import warnings -from _collections_abc import Sequence -from errno import ENOENT, ENOTDIR, EBADF, ELOOP -from stat import S_ISDIR, S_ISLNK, S_ISREG, S_ISSOCK, S_ISBLK, S_ISCHR, S_ISFIFO -from urllib.parse import quote_from_bytes as urlquote_from_bytes - - -__all__ = [ - "PurePath", "PurePosixPath", "PureWindowsPath", - "Path", "PosixPath", "WindowsPath", - ] - -# -# Internals -# - -# Reference for Windows paths can be found at -# https://learn.microsoft.com/en-gb/windows/win32/fileio/naming-a-file . -_WIN_RESERVED_NAMES = frozenset( - {'CON', 'PRN', 'AUX', 'NUL', 'CONIN$', 'CONOUT$'} | - {f'COM{c}' for c in '123456789\xb9\xb2\xb3'} | - {f'LPT{c}' for c in '123456789\xb9\xb2\xb3'} -) - -_WINERROR_NOT_READY = 21 # drive exists but is not accessible -_WINERROR_INVALID_NAME = 123 # fix for bpo-35306 -_WINERROR_CANT_RESOLVE_FILENAME = 1921 # broken symlink pointing to itself - -# EBADF - guard against macOS `stat` throwing EBADF -_IGNORED_ERRNOS = (ENOENT, ENOTDIR, EBADF, ELOOP) - -_IGNORED_WINERRORS = ( - _WINERROR_NOT_READY, - _WINERROR_INVALID_NAME, - _WINERROR_CANT_RESOLVE_FILENAME) - -def _ignore_error(exception): - return (getattr(exception, 'errno', None) in _IGNORED_ERRNOS or - getattr(exception, 'winerror', None) in _IGNORED_WINERRORS) - - -@functools.cache -def _is_case_sensitive(flavour): - return flavour.normcase('Aa') == 'Aa' - -# -# Globbing helpers -# - - -# fnmatch.translate() returns a regular expression that includes a prefix and -# a suffix, which enable matching newlines and ensure the end of the string is -# matched, respectively. These features are undesirable for our implementation -# of PurePatch.match(), which represents path separators as newlines and joins -# pattern segments together. As a workaround, we define a slice object that -# can remove the prefix and suffix from any translate() result. See the -# _compile_pattern_lines() function for more details. -_FNMATCH_PREFIX, _FNMATCH_SUFFIX = fnmatch.translate('_').split('_') -_FNMATCH_SLICE = slice(len(_FNMATCH_PREFIX), -len(_FNMATCH_SUFFIX)) -_SWAP_SEP_AND_NEWLINE = { - '/': str.maketrans({'/': '\n', '\n': '/'}), - '\\': str.maketrans({'\\': '\n', '\n': '\\'}), -} - - -@functools.lru_cache() -def _make_selector(pattern_parts, flavour, case_sensitive): - pat = pattern_parts[0] - if not pat: - return _TerminatingSelector() - if pat == '**': - child_parts_idx = 1 - while child_parts_idx < len(pattern_parts) and pattern_parts[child_parts_idx] == '**': - child_parts_idx += 1 - child_parts = pattern_parts[child_parts_idx:] - if '**' in child_parts: - cls = _DoubleRecursiveWildcardSelector - else: - cls = _RecursiveWildcardSelector - else: - child_parts = pattern_parts[1:] - if pat == '..': - cls = _ParentSelector - elif '**' in pat: - raise ValueError("Invalid pattern: '**' can only be an entire path component") - else: - cls = _WildcardSelector - return cls(pat, child_parts, flavour, case_sensitive) - - -@functools.lru_cache(maxsize=256) -def _compile_pattern(pat, case_sensitive): - flags = re.NOFLAG if case_sensitive else re.IGNORECASE - return re.compile(fnmatch.translate(pat), flags).match - - -@functools.lru_cache() -def _compile_pattern_lines(pattern_lines, case_sensitive): - """Compile the given pattern lines to an `re.Pattern` object. - - The *pattern_lines* argument is a glob-style pattern (e.g. '*/*.py') with - its path separators and newlines swapped (e.g. '*\n*.py`). By using - newlines to separate path components, and not setting `re.DOTALL`, we - ensure that the `*` wildcard cannot match path separators. - - The returned `re.Pattern` object may have its `match()` method called to - match a complete pattern, or `search()` to match from the right. The - argument supplied to these methods must also have its path separators and - newlines swapped. - """ - - # Match the start of the path, or just after a path separator - parts = ['^'] - for part in pattern_lines.splitlines(keepends=True): - if part == '*\n': - part = r'.+\n' - elif part == '*': - part = r'.+' - else: - # Any other component: pass to fnmatch.translate(). We slice off - # the common prefix and suffix added by translate() to ensure that - # re.DOTALL is not set, and the end of the string not matched, - # respectively. With DOTALL not set, '*' wildcards will not match - # path separators, because the '.' characters in the pattern will - # not match newlines. - part = fnmatch.translate(part)[_FNMATCH_SLICE] - parts.append(part) - # Match the end of the path, always. - parts.append(r'\Z') - flags = re.MULTILINE - if not case_sensitive: - flags |= re.IGNORECASE - return re.compile(''.join(parts), flags=flags) - - -class _Selector: - """A selector matches a specific glob pattern part against the children - of a given path.""" - - def __init__(self, child_parts, flavour, case_sensitive): - self.child_parts = child_parts - if child_parts: - self.successor = _make_selector(child_parts, flavour, case_sensitive) - self.dironly = True - else: - self.successor = _TerminatingSelector() - self.dironly = False - - def select_from(self, parent_path): - """Iterate over all child paths of `parent_path` matched by this - selector. This can contain parent_path itself.""" - path_cls = type(parent_path) - scandir = path_cls._scandir - if not parent_path.is_dir(): - return iter([]) - return self._select_from(parent_path, scandir) - - -class _TerminatingSelector: - - def _select_from(self, parent_path, scandir): - yield parent_path - - -class _ParentSelector(_Selector): - - def __init__(self, name, child_parts, flavour, case_sensitive): - _Selector.__init__(self, child_parts, flavour, case_sensitive) - - def _select_from(self, parent_path, scandir): - path = parent_path._make_child_relpath('..') - for p in self.successor._select_from(path, scandir): - yield p - - -class _WildcardSelector(_Selector): - - def __init__(self, pat, child_parts, flavour, case_sensitive): - _Selector.__init__(self, child_parts, flavour, case_sensitive) - if case_sensitive is None: - # TODO: evaluate case-sensitivity of each directory in _select_from() - case_sensitive = _is_case_sensitive(flavour) - self.match = _compile_pattern(pat, case_sensitive) - - def _select_from(self, parent_path, scandir): - try: - # We must close the scandir() object before proceeding to - # avoid exhausting file descriptors when globbing deep trees. - with scandir(parent_path) as scandir_it: - entries = list(scandir_it) - except OSError: - pass - else: - for entry in entries: - if self.dironly: - try: - if not entry.is_dir(): - continue - except OSError: - continue - name = entry.name - if self.match(name): - path = parent_path._make_child_relpath(name) - for p in self.successor._select_from(path, scandir): - yield p - - -class _RecursiveWildcardSelector(_Selector): - - def __init__(self, pat, child_parts, flavour, case_sensitive): - _Selector.__init__(self, child_parts, flavour, case_sensitive) - - def _iterate_directories(self, parent_path): - yield parent_path - for dirpath, dirnames, _ in parent_path.walk(): - for dirname in dirnames: - yield dirpath._make_child_relpath(dirname) - - def _select_from(self, parent_path, scandir): - successor_select = self.successor._select_from - for starting_point in self._iterate_directories(parent_path): - for p in successor_select(starting_point, scandir): - yield p - - -class _DoubleRecursiveWildcardSelector(_RecursiveWildcardSelector): - """ - Like _RecursiveWildcardSelector, but also de-duplicates results from - successive selectors. This is necessary if the pattern contains - multiple non-adjacent '**' segments. - """ - - def _select_from(self, parent_path, scandir): - yielded = set() - try: - for p in super()._select_from(parent_path, scandir): - if p not in yielded: - yield p - yielded.add(p) - finally: - yielded.clear() - - -# -# Public API -# - -class _PathParents(Sequence): - """This object provides sequence-like access to the logical ancestors - of a path. Don't try to construct it yourself.""" - __slots__ = ('_path', '_drv', '_root', '_tail') - - def __init__(self, path): - self._path = path - self._drv = path.drive - self._root = path.root - self._tail = path._tail - - def __len__(self): - return len(self._tail) - - def __getitem__(self, idx): - if isinstance(idx, slice): - return tuple(self[i] for i in range(*idx.indices(len(self)))) - - if idx >= len(self) or idx < -len(self): - raise IndexError(idx) - if idx < 0: - idx += len(self) - return self._path._from_parsed_parts(self._drv, self._root, - self._tail[:-idx - 1]) - - def __repr__(self): - return "<{}.parents>".format(type(self._path).__name__) - - -class PurePath(object): - """Base class for manipulating paths without I/O. - - PurePath represents a filesystem path and offers operations which - don't imply any actual filesystem I/O. Depending on your system, - instantiating a PurePath will return either a PurePosixPath or a - PureWindowsPath object. You can also instantiate either of these classes - directly, regardless of your system. - """ - - __slots__ = ( - # The `_raw_paths` slot stores unnormalized string paths. This is set - # in the `__init__()` method. - '_raw_paths', - - # The `_drv`, `_root` and `_tail_cached` slots store parsed and - # normalized parts of the path. They are set when any of the `drive`, - # `root` or `_tail` properties are accessed for the first time. The - # three-part division corresponds to the result of - # `os.path.splitroot()`, except that the tail is further split on path - # separators (i.e. it is a list of strings), and that the root and - # tail are normalized. - '_drv', '_root', '_tail_cached', - - # The `_str` slot stores the string representation of the path, - # computed from the drive, root and tail when `__str__()` is called - # for the first time. It's used to implement `_str_normcase` - '_str', - - # The `_str_normcase_cached` slot stores the string path with - # normalized case. It is set when the `_str_normcase` property is - # accessed for the first time. It's used to implement `__eq__()` - # `__hash__()`, and `_parts_normcase` - '_str_normcase_cached', - - # The `_parts_normcase_cached` slot stores the case-normalized - # string path after splitting on path separators. It's set when the - # `_parts_normcase` property is accessed for the first time. It's used - # to implement comparison methods like `__lt__()`. - '_parts_normcase_cached', - - # The `_lines_cached` slot stores the string path with path separators - # and newlines swapped. This is used to implement `match()`. - '_lines_cached', - - # The `_hash` slot stores the hash of the case-normalized string - # path. It's set when `__hash__()` is called for the first time. - '_hash', - ) - _flavour = os.path - - def __new__(cls, *args, **kwargs): - """Construct a PurePath from one or several strings and or existing - PurePath objects. The strings and path objects are combined so as - to yield a canonicalized path, which is incorporated into the - new PurePath object. - """ - if cls is PurePath: - cls = PureWindowsPath if os.name == 'nt' else PurePosixPath - return object.__new__(cls) - - def __reduce__(self): - # Using the parts tuple helps share interned path parts - # when pickling related paths. - return (self.__class__, self.parts) - - def __init__(self, *args): - paths = [] - for arg in args: - if isinstance(arg, PurePath): - if arg._flavour is ntpath and self._flavour is posixpath: - # GH-103631: Convert separators for backwards compatibility. - paths.extend(path.replace('\\', '/') for path in arg._raw_paths) - else: - paths.extend(arg._raw_paths) - else: - try: - path = os.fspath(arg) - except TypeError: - path = arg - if not isinstance(path, str): - raise TypeError( - "argument should be a str or an os.PathLike " - "object where __fspath__ returns a str, " - f"not {type(path).__name__!r}") - paths.append(path) - self._raw_paths = paths - - def with_segments(self, *pathsegments): - """Construct a new path object from any number of path-like objects. - Subclasses may override this method to customize how new path objects - are created from methods like `iterdir()`. - """ - return type(self)(*pathsegments) - - @classmethod - def _parse_path(cls, path): - if not path: - return '', '', [] - sep = cls._flavour.sep - altsep = cls._flavour.altsep - if altsep: - path = path.replace(altsep, sep) - drv, root, rel = cls._flavour.splitroot(path) - if not root and drv.startswith(sep) and not drv.endswith(sep): - drv_parts = drv.split(sep) - if len(drv_parts) == 4 and drv_parts[2] not in '?.': - # e.g. //server/share - root = sep - elif len(drv_parts) == 6: - # e.g. //?/unc/server/share - root = sep - parsed = [sys.intern(str(x)) for x in rel.split(sep) if x and x != '.'] - return drv, root, parsed - - def _load_parts(self): - paths = self._raw_paths - if len(paths) == 0: - path = '' - elif len(paths) == 1: - path = paths[0] - else: - path = self._flavour.join(*paths) - drv, root, tail = self._parse_path(path) - self._drv = drv - self._root = root - self._tail_cached = tail - - def _from_parsed_parts(self, drv, root, tail): - path_str = self._format_parsed_parts(drv, root, tail) - path = self.with_segments(path_str) - path._str = path_str or '.' - path._drv = drv - path._root = root - path._tail_cached = tail - return path - - @classmethod - def _format_parsed_parts(cls, drv, root, tail): - if drv or root: - return drv + root + cls._flavour.sep.join(tail) - elif tail and cls._flavour.splitdrive(tail[0])[0]: - tail = ['.'] + tail - return cls._flavour.sep.join(tail) - - def __str__(self): - """Return the string representation of the path, suitable for - passing to system calls.""" - try: - return self._str - except AttributeError: - self._str = self._format_parsed_parts(self.drive, self.root, - self._tail) or '.' - return self._str - - def __fspath__(self): - return str(self) - - def as_posix(self): - """Return the string representation of the path with forward (/) - slashes.""" - f = self._flavour - return str(self).replace(f.sep, '/') - - def __bytes__(self): - """Return the bytes representation of the path. This is only - recommended to use under Unix.""" - return os.fsencode(self) - - def __repr__(self): - return "{}({!r})".format(self.__class__.__name__, self.as_posix()) - - def as_uri(self): - """Return the path as a 'file' URI.""" - if not self.is_absolute(): - raise ValueError("relative path can't be expressed as a file URI") - - drive = self.drive - if len(drive) == 2 and drive[1] == ':': - # It's a path on a local drive => 'file:///c:/a/b' - prefix = 'file:///' + drive - path = self.as_posix()[2:] - elif drive: - # It's a path on a network drive => 'file://host/share/a/b' - prefix = 'file:' - path = self.as_posix() - else: - # It's a posix path => 'file:///etc/hosts' - prefix = 'file://' - path = str(self) - return prefix + urlquote_from_bytes(os.fsencode(path)) - - @property - def _str_normcase(self): - # String with normalized case, for hashing and equality checks - try: - return self._str_normcase_cached - except AttributeError: - if _is_case_sensitive(self._flavour): - self._str_normcase_cached = str(self) - else: - self._str_normcase_cached = str(self).lower() - return self._str_normcase_cached - - @property - def _parts_normcase(self): - # Cached parts with normalized case, for comparisons. - try: - return self._parts_normcase_cached - except AttributeError: - self._parts_normcase_cached = self._str_normcase.split(self._flavour.sep) - return self._parts_normcase_cached - - @property - def _lines(self): - # Path with separators and newlines swapped, for pattern matching. - try: - return self._lines_cached - except AttributeError: - path_str = str(self) - if path_str == '.': - self._lines_cached = '' - else: - trans = _SWAP_SEP_AND_NEWLINE[self._flavour.sep] - self._lines_cached = path_str.translate(trans) - return self._lines_cached - - def __eq__(self, other): - if not isinstance(other, PurePath): - return NotImplemented - return self._str_normcase == other._str_normcase and self._flavour is other._flavour - - def __hash__(self): - try: - return self._hash - except AttributeError: - self._hash = hash(self._str_normcase) - return self._hash - - def __lt__(self, other): - if not isinstance(other, PurePath) or self._flavour is not other._flavour: - return NotImplemented - return self._parts_normcase < other._parts_normcase - - def __le__(self, other): - if not isinstance(other, PurePath) or self._flavour is not other._flavour: - return NotImplemented - return self._parts_normcase <= other._parts_normcase - - def __gt__(self, other): - if not isinstance(other, PurePath) or self._flavour is not other._flavour: - return NotImplemented - return self._parts_normcase > other._parts_normcase - - def __ge__(self, other): - if not isinstance(other, PurePath) or self._flavour is not other._flavour: - return NotImplemented - return self._parts_normcase >= other._parts_normcase - - @property - def drive(self): - """The drive prefix (letter or UNC path), if any.""" - try: - return self._drv - except AttributeError: - self._load_parts() - return self._drv - - @property - def root(self): - """The root of the path, if any.""" - try: - return self._root - except AttributeError: - self._load_parts() - return self._root - - @property - def _tail(self): - try: - return self._tail_cached - except AttributeError: - self._load_parts() - return self._tail_cached - - @property - def anchor(self): - """The concatenation of the drive and root, or ''.""" - anchor = self.drive + self.root - return anchor - - @property - def name(self): - """The final path component, if any.""" - tail = self._tail - if not tail: - return '' - return tail[-1] - - @property - def suffix(self): - """ - The final component's last suffix, if any. - - This includes the leading period. For example: '.txt' - """ - name = self.name - i = name.rfind('.') - if 0 < i < len(name) - 1: - return name[i:] - else: - return '' - - @property - def suffixes(self): - """ - A list of the final component's suffixes, if any. - - These include the leading periods. For example: ['.tar', '.gz'] - """ - name = self.name - if name.endswith('.'): - return [] - name = name.lstrip('.') - return ['.' + suffix for suffix in name.split('.')[1:]] - - @property - def stem(self): - """The final path component, minus its last suffix.""" - name = self.name - i = name.rfind('.') - if 0 < i < len(name) - 1: - return name[:i] - else: - return name - - def with_name(self, name): - """Return a new path with the file name changed.""" - if not self.name: - raise ValueError("%r has an empty name" % (self,)) - f = self._flavour - if not name or f.sep in name or (f.altsep and f.altsep in name) or name == '.': - raise ValueError("Invalid name %r" % (name)) - return self._from_parsed_parts(self.drive, self.root, - self._tail[:-1] + [name]) - - def with_stem(self, stem): - """Return a new path with the stem changed.""" - return self.with_name(stem + self.suffix) - - def with_suffix(self, suffix): - """Return a new path with the file suffix changed. If the path - has no suffix, add given suffix. If the given suffix is an empty - string, remove the suffix from the path. - """ - f = self._flavour - if f.sep in suffix or f.altsep and f.altsep in suffix: - raise ValueError("Invalid suffix %r" % (suffix,)) - if suffix and not suffix.startswith('.') or suffix == '.': - raise ValueError("Invalid suffix %r" % (suffix)) - name = self.name - if not name: - raise ValueError("%r has an empty name" % (self,)) - old_suffix = self.suffix - if not old_suffix: - name = name + suffix - else: - name = name[:-len(old_suffix)] + suffix - return self._from_parsed_parts(self.drive, self.root, - self._tail[:-1] + [name]) - - def relative_to(self, other, /, *_deprecated, walk_up=False): - """Return the relative path to another path identified by the passed - arguments. If the operation is not possible (because this is not - related to the other path), raise ValueError. - - The *walk_up* parameter controls whether `..` may be used to resolve - the path. - """ - if _deprecated: - msg = ("support for supplying more than one positional argument " - "to pathlib.PurePath.relative_to() is deprecated and " - "scheduled for removal in Python {remove}") - warnings._deprecated("pathlib.PurePath.relative_to(*args)", msg, - remove=(3, 14)) - other = self.with_segments(other, *_deprecated) - for step, path in enumerate([other] + list(other.parents)): - if self.is_relative_to(path): - break - elif not walk_up: - raise ValueError(f"{str(self)!r} is not in the subpath of {str(other)!r}") - elif path.name == '..': - raise ValueError(f"'..' segment in {str(other)!r} cannot be walked") - else: - raise ValueError(f"{str(self)!r} and {str(other)!r} have different anchors") - parts = ['..'] * step + self._tail[len(path._tail):] - return self.with_segments(*parts) - - def is_relative_to(self, other, /, *_deprecated): - """Return True if the path is relative to another path or False. - """ - if _deprecated: - msg = ("support for supplying more than one argument to " - "pathlib.PurePath.is_relative_to() is deprecated and " - "scheduled for removal in Python {remove}") - warnings._deprecated("pathlib.PurePath.is_relative_to(*args)", - msg, remove=(3, 14)) - other = self.with_segments(other, *_deprecated) - return other == self or other in self.parents - - @property - def parts(self): - """An object providing sequence-like access to the - components in the filesystem path.""" - if self.drive or self.root: - return (self.drive + self.root,) + tuple(self._tail) - else: - return tuple(self._tail) - - def joinpath(self, *pathsegments): - """Combine this path with one or several arguments, and return a - new path representing either a subpath (if all arguments are relative - paths) or a totally different path (if one of the arguments is - anchored). - """ - return self.with_segments(self, *pathsegments) - - def __truediv__(self, key): - try: - return self.joinpath(key) - except TypeError: - return NotImplemented - - def __rtruediv__(self, key): - try: - return self.with_segments(key, self) - except TypeError: - return NotImplemented - - @property - def parent(self): - """The logical parent of the path.""" - drv = self.drive - root = self.root - tail = self._tail - if not tail: - return self - return self._from_parsed_parts(drv, root, tail[:-1]) - - @property - def parents(self): - """A sequence of this path's logical parents.""" - # The value of this property should not be cached on the path object, - # as doing so would introduce a reference cycle. - return _PathParents(self) - - def is_absolute(self): - """True if the path is absolute (has both a root and, if applicable, - a drive).""" - if self._flavour is ntpath: - # ntpath.isabs() is defective - see GH-44626. - return bool(self.drive and self.root) - elif self._flavour is posixpath: - # Optimization: work with raw paths on POSIX. - for path in self._raw_paths: - if path.startswith('/'): - return True - return False - else: - return self._flavour.isabs(str(self)) - - def is_reserved(self): - """Return True if the path contains one of the special names reserved - by the system, if any.""" - if self._flavour is posixpath or not self._tail: - return False - - # NOTE: the rules for reserved names seem somewhat complicated - # (e.g. r"..\NUL" is reserved but not r"foo\NUL" if "foo" does not - # exist). We err on the side of caution and return True for paths - # which are not considered reserved by Windows. - if self.drive.startswith('\\\\'): - # UNC paths are never reserved. - return False - name = self._tail[-1].partition('.')[0].partition(':')[0].rstrip(' ') - return name.upper() in _WIN_RESERVED_NAMES - - def match(self, path_pattern, *, case_sensitive=None): - """ - Return True if this path matches the given pattern. - """ - if not isinstance(path_pattern, PurePath): - path_pattern = self.with_segments(path_pattern) - if case_sensitive is None: - case_sensitive = _is_case_sensitive(self._flavour) - pattern = _compile_pattern_lines(path_pattern._lines, case_sensitive) - if path_pattern.drive or path_pattern.root: - return pattern.match(self._lines) is not None - elif path_pattern._tail: - return pattern.search(self._lines) is not None - else: - raise ValueError("empty pattern") - - -# Can't subclass os.PathLike from PurePath and keep the constructor -# optimizations in PurePath.__slots__. -os.PathLike.register(PurePath) - - -class PurePosixPath(PurePath): - """PurePath subclass for non-Windows systems. - - On a POSIX system, instantiating a PurePath should return this object. - However, you can also instantiate it directly on any system. - """ - _flavour = posixpath - __slots__ = () - - -class PureWindowsPath(PurePath): - """PurePath subclass for Windows systems. - - On a Windows system, instantiating a PurePath should return this object. - However, you can also instantiate it directly on any system. - """ - _flavour = ntpath - __slots__ = () - - -# Filesystem-accessing classes - - -class Path(PurePath): - """PurePath subclass that can make system calls. - - Path represents a filesystem path but unlike PurePath, also offers - methods to do system calls on path objects. Depending on your system, - instantiating a Path will return either a PosixPath or a WindowsPath - object. You can also instantiate a PosixPath or WindowsPath directly, - but cannot instantiate a WindowsPath on a POSIX system or vice versa. - """ - __slots__ = () - - def stat(self, *, follow_symlinks=True): - """ - Return the result of the stat() system call on this path, like - os.stat() does. - """ - return os.stat(self, follow_symlinks=follow_symlinks) - - def lstat(self): - """ - Like stat(), except if the path points to a symlink, the symlink's - status information is returned, rather than its target's. - """ - return self.stat(follow_symlinks=False) - - - # Convenience functions for querying the stat results - - def exists(self, *, follow_symlinks=True): - """ - Whether this path exists. - - This method normally follows symlinks; to check whether a symlink exists, - add the argument follow_symlinks=False. - """ - try: - self.stat(follow_symlinks=follow_symlinks) - except OSError as e: - if not _ignore_error(e): - raise - return False - except ValueError: - # Non-encodable path - return False - return True - - def is_dir(self): - """ - Whether this path is a directory. - """ - try: - return S_ISDIR(self.stat().st_mode) - except OSError as e: - if not _ignore_error(e): - raise - # Path doesn't exist or is a broken symlink - # (see http://web.archive.org/web/20200623061726/https://bitbucket.org/pitrou/pathlib/issues/12/ ) - return False - except ValueError: - # Non-encodable path - return False - - def is_file(self): - """ - Whether this path is a regular file (also True for symlinks pointing - to regular files). - """ - try: - return S_ISREG(self.stat().st_mode) - except OSError as e: - if not _ignore_error(e): - raise - # Path doesn't exist or is a broken symlink - # (see http://web.archive.org/web/20200623061726/https://bitbucket.org/pitrou/pathlib/issues/12/ ) - return False - except ValueError: - # Non-encodable path - return False - - def is_mount(self): - """ - Check if this path is a mount point - """ - return self._flavour.ismount(self) - - def is_symlink(self): - """ - Whether this path is a symbolic link. - """ - try: - return S_ISLNK(self.lstat().st_mode) - except OSError as e: - if not _ignore_error(e): - raise - # Path doesn't exist - return False - except ValueError: - # Non-encodable path - return False - - def is_junction(self): - """ - Whether this path is a junction. - """ - return self._flavour.isjunction(self) - - def is_block_device(self): - """ - Whether this path is a block device. - """ - try: - return S_ISBLK(self.stat().st_mode) - except OSError as e: - if not _ignore_error(e): - raise - # Path doesn't exist or is a broken symlink - # (see http://web.archive.org/web/20200623061726/https://bitbucket.org/pitrou/pathlib/issues/12/ ) - return False - except ValueError: - # Non-encodable path - return False - - def is_char_device(self): - """ - Whether this path is a character device. - """ - try: - return S_ISCHR(self.stat().st_mode) - except OSError as e: - if not _ignore_error(e): - raise - # Path doesn't exist or is a broken symlink - # (see http://web.archive.org/web/20200623061726/https://bitbucket.org/pitrou/pathlib/issues/12/ ) - return False - except ValueError: - # Non-encodable path - return False - - def is_fifo(self): - """ - Whether this path is a FIFO. - """ - try: - return S_ISFIFO(self.stat().st_mode) - except OSError as e: - if not _ignore_error(e): - raise - # Path doesn't exist or is a broken symlink - # (see http://web.archive.org/web/20200623061726/https://bitbucket.org/pitrou/pathlib/issues/12/ ) - return False - except ValueError: - # Non-encodable path - return False - - def is_socket(self): - """ - Whether this path is a socket. - """ - try: - return S_ISSOCK(self.stat().st_mode) - except OSError as e: - if not _ignore_error(e): - raise - # Path doesn't exist or is a broken symlink - # (see http://web.archive.org/web/20200623061726/https://bitbucket.org/pitrou/pathlib/issues/12/ ) - return False - except ValueError: - # Non-encodable path - return False - - def samefile(self, other_path): - """Return whether other_path is the same or not as this file - (as returned by os.path.samefile()). - """ - st = self.stat() - try: - other_st = other_path.stat() - except AttributeError: - other_st = self.with_segments(other_path).stat() - return self._flavour.samestat(st, other_st) - - def open(self, mode='r', buffering=-1, encoding=None, - errors=None, newline=None): - """ - Open the file pointed by this path and return a file object, as - the built-in open() function does. - """ - if "b" not in mode: - encoding = io.text_encoding(encoding) - return io.open(self, mode, buffering, encoding, errors, newline) - - def read_bytes(self): - """ - Open the file in bytes mode, read it, and close the file. - """ - with self.open(mode='rb') as f: - return f.read() - - def read_text(self, encoding=None, errors=None): - """ - Open the file in text mode, read it, and close the file. - """ - encoding = io.text_encoding(encoding) - with self.open(mode='r', encoding=encoding, errors=errors) as f: - return f.read() - - def write_bytes(self, data): - """ - Open the file in bytes mode, write to it, and close the file. - """ - # type-check for the buffer interface before truncating the file - view = memoryview(data) - with self.open(mode='wb') as f: - return f.write(view) - - def write_text(self, data, encoding=None, errors=None, newline=None): - """ - Open the file in text mode, write to it, and close the file. - """ - if not isinstance(data, str): - raise TypeError('data must be str, not %s' % - data.__class__.__name__) - encoding = io.text_encoding(encoding) - with self.open(mode='w', encoding=encoding, errors=errors, newline=newline) as f: - return f.write(data) - - def iterdir(self): - """Yield path objects of the directory contents. - - The children are yielded in arbitrary order, and the - special entries '.' and '..' are not included. - """ - for name in os.listdir(self): - yield self._make_child_relpath(name) - - def _scandir(self): - # bpo-24132: a future version of pathlib will support subclassing of - # pathlib.Path to customize how the filesystem is accessed. This - # includes scandir(), which is used to implement glob(). - return os.scandir(self) - - def _make_child_relpath(self, name): - path_str = str(self) - tail = self._tail - if tail: - path_str = f'{path_str}{self._flavour.sep}{name}' - elif path_str != '.': - path_str = f'{path_str}{name}' - else: - path_str = name - path = self.with_segments(path_str) - path._str = path_str - path._drv = self.drive - path._root = self.root - path._tail_cached = tail + [name] - return path - - def glob(self, pattern, *, case_sensitive=None): - """Iterate over this subtree and yield all existing files (of any - kind, including directories) matching the given relative pattern. - """ - sys.audit("pathlib.Path.glob", self, pattern) - if not pattern: - raise ValueError("Unacceptable pattern: {!r}".format(pattern)) - drv, root, pattern_parts = self._parse_path(pattern) - if drv or root: - raise NotImplementedError("Non-relative patterns are unsupported") - if pattern[-1] in (self._flavour.sep, self._flavour.altsep): - pattern_parts.append('') - selector = _make_selector(tuple(pattern_parts), self._flavour, case_sensitive) - for p in selector.select_from(self): - yield p - - def rglob(self, pattern, *, case_sensitive=None): - """Recursively yield all existing files (of any kind, including - directories) matching the given relative pattern, anywhere in - this subtree. - """ - sys.audit("pathlib.Path.rglob", self, pattern) - drv, root, pattern_parts = self._parse_path(pattern) - if drv or root: - raise NotImplementedError("Non-relative patterns are unsupported") - if pattern and pattern[-1] in (self._flavour.sep, self._flavour.altsep): - pattern_parts.append('') - selector = _make_selector(("**",) + tuple(pattern_parts), self._flavour, case_sensitive) - for p in selector.select_from(self): - yield p - - def walk(self, top_down=True, on_error=None, follow_symlinks=False): - """Walk the directory tree from this directory, similar to os.walk().""" - sys.audit("pathlib.Path.walk", self, on_error, follow_symlinks) - paths = [self] - - while paths: - path = paths.pop() - if isinstance(path, tuple): - yield path - continue - - # We may not have read permission for self, in which case we can't - # get a list of the files the directory contains. os.walk() - # always suppressed the exception in that instance, rather than - # blow up for a minor reason when (say) a thousand readable - # directories are still left to visit. That logic is copied here. - try: - scandir_it = path._scandir() - except OSError as error: - if on_error is not None: - on_error(error) - continue - - with scandir_it: - dirnames = [] - filenames = [] - for entry in scandir_it: - try: - is_dir = entry.is_dir(follow_symlinks=follow_symlinks) - except OSError: - # Carried over from os.path.isdir(). - is_dir = False - - if is_dir: - dirnames.append(entry.name) - else: - filenames.append(entry.name) - - if top_down: - yield path, dirnames, filenames - else: - paths.append((path, dirnames, filenames)) - - paths += [path._make_child_relpath(d) for d in reversed(dirnames)] - - def __init__(self, *args, **kwargs): - if kwargs: - msg = ("support for supplying keyword arguments to pathlib.PurePath " - "is deprecated and scheduled for removal in Python {remove}") - warnings._deprecated("pathlib.PurePath(**kwargs)", msg, remove=(3, 14)) - super().__init__(*args) - - def __new__(cls, *args, **kwargs): - if cls is Path: - cls = WindowsPath if os.name == 'nt' else PosixPath - return object.__new__(cls) - - def __enter__(self): - # In previous versions of pathlib, __exit__() marked this path as - # closed; subsequent attempts to perform I/O would raise an IOError. - # This functionality was never documented, and had the effect of - # making Path objects mutable, contrary to PEP 428. - # In Python 3.9 __exit__() was made a no-op. - # In Python 3.11 __enter__() began emitting DeprecationWarning. - # In Python 3.13 __enter__() and __exit__() should be removed. - warnings.warn("pathlib.Path.__enter__() is deprecated and scheduled " - "for removal in Python 3.13; Path objects as a context " - "manager is a no-op", - DeprecationWarning, stacklevel=2) - return self - - def __exit__(self, t, v, tb): - pass - - # Public API - - @classmethod - def cwd(cls): - """Return a new path pointing to the current working directory.""" - # We call 'absolute()' rather than using 'os.getcwd()' directly to - # enable users to replace the implementation of 'absolute()' in a - # subclass and benefit from the new behaviour here. This works because - # os.path.abspath('.') == os.getcwd(). - return cls().absolute() - - @classmethod - def home(cls): - """Return a new path pointing to the user's home directory (as - returned by os.path.expanduser('~')). - """ - return cls("~").expanduser() - - def absolute(self): - """Return an absolute version of this path by prepending the current - working directory. No normalization or symlink resolution is performed. - - Use resolve() to get the canonical path to a file. - """ - if self.is_absolute(): - return self - elif self.drive: - # There is a CWD on each drive-letter drive. - cwd = self._flavour.abspath(self.drive) - else: - cwd = os.getcwd() - # Fast path for "empty" paths, e.g. Path("."), Path("") or Path(). - # We pass only one argument to with_segments() to avoid the cost - # of joining, and we exploit the fact that getcwd() returns a - # fully-normalized string by storing it in _str. This is used to - # implement Path.cwd(). - if not self.root and not self._tail: - result = self.with_segments(cwd) - result._str = cwd - return result - return self.with_segments(cwd, self) - - def resolve(self, strict=False): - """ - Make the path absolute, resolving all symlinks on the way and also - normalizing it. - """ - - def check_eloop(e): - winerror = getattr(e, 'winerror', 0) - if e.errno == ELOOP or winerror == _WINERROR_CANT_RESOLVE_FILENAME: - raise RuntimeError("Symlink loop from %r" % e.filename) - - try: - s = self._flavour.realpath(self, strict=strict) - except OSError as e: - check_eloop(e) - raise - p = self.with_segments(s) - - # In non-strict mode, realpath() doesn't raise on symlink loops. - # Ensure we get an exception by calling stat() - if not strict: - try: - p.stat() - except OSError as e: - check_eloop(e) - return p - - def owner(self): - """ - Return the login name of the file owner. - """ - try: - import pwd - return pwd.getpwuid(self.stat().st_uid).pw_name - except ImportError: - raise NotImplementedError("Path.owner() is unsupported on this system") - - def group(self): - """ - Return the group name of the file gid. - """ - - try: - import grp - return grp.getgrgid(self.stat().st_gid).gr_name - except ImportError: - raise NotImplementedError("Path.group() is unsupported on this system") - - def readlink(self): - """ - Return the path to which the symbolic link points. - """ - if not hasattr(os, "readlink"): - raise NotImplementedError("os.readlink() not available on this system") - return self.with_segments(os.readlink(self)) - - def touch(self, mode=0o666, exist_ok=True): - """ - Create this file with the given access mode, if it doesn't exist. - """ - - if exist_ok: - # First try to bump modification time - # Implementation note: GNU touch uses the UTIME_NOW option of - # the utimensat() / futimens() functions. - try: - os.utime(self, None) - except OSError: - # Avoid exception chaining - pass - else: - return - flags = os.O_CREAT | os.O_WRONLY - if not exist_ok: - flags |= os.O_EXCL - fd = os.open(self, flags, mode) - os.close(fd) - - def mkdir(self, mode=0o777, parents=False, exist_ok=False): - """ - Create a new directory at this given path. - """ - try: - os.mkdir(self, mode) - except FileNotFoundError: - if not parents or self.parent == self: - raise - self.parent.mkdir(parents=True, exist_ok=True) - self.mkdir(mode, parents=False, exist_ok=exist_ok) - except OSError: - # Cannot rely on checking for EEXIST, since the operating system - # could give priority to other errors like EACCES or EROFS - if not exist_ok or not self.is_dir(): - raise - - def chmod(self, mode, *, follow_symlinks=True): - """ - Change the permissions of the path, like os.chmod(). - """ - os.chmod(self, mode, follow_symlinks=follow_symlinks) - - def lchmod(self, mode): - """ - Like chmod(), except if the path points to a symlink, the symlink's - permissions are changed, rather than its target's. - """ - self.chmod(mode, follow_symlinks=False) - - def unlink(self, missing_ok=False): - """ - Remove this file or link. - If the path is a directory, use rmdir() instead. - """ - try: - os.unlink(self) - except FileNotFoundError: - if not missing_ok: - raise - - def rmdir(self): - """ - Remove this directory. The directory must be empty. - """ - os.rmdir(self) - - def rename(self, target): - """ - Rename this path to the target path. - - The target path may be absolute or relative. Relative paths are - interpreted relative to the current working directory, *not* the - directory of the Path object. - - Returns the new Path instance pointing to the target path. - """ - os.rename(self, target) - return self.with_segments(target) - - def replace(self, target): - """ - Rename this path to the target path, overwriting if that path exists. - - The target path may be absolute or relative. Relative paths are - interpreted relative to the current working directory, *not* the - directory of the Path object. - - Returns the new Path instance pointing to the target path. - """ - os.replace(self, target) - return self.with_segments(target) - - def symlink_to(self, target, target_is_directory=False): - """ - Make this path a symlink pointing to the target path. - Note the order of arguments (link, target) is the reverse of os.symlink. - """ - if not hasattr(os, "symlink"): - raise NotImplementedError("os.symlink() not available on this system") - os.symlink(target, self, target_is_directory) - - def hardlink_to(self, target): - """ - Make this path a hard link pointing to the same file as *target*. - - Note the order of arguments (self, target) is the reverse of os.link's. - """ - if not hasattr(os, "link"): - raise NotImplementedError("os.link() not available on this system") - os.link(target, self) - - def expanduser(self): - """ Return a new path with expanded ~ and ~user constructs - (as returned by os.path.expanduser) - """ - if (not (self.drive or self.root) and - self._tail and self._tail[0][:1] == '~'): - homedir = self._flavour.expanduser(self._tail[0]) - if homedir[:1] == "~": - raise RuntimeError("Could not determine home directory.") - drv, root, tail = self._parse_path(homedir) - return self._from_parsed_parts(drv, root, tail + self._tail[1:]) - - return self - - -class PosixPath(Path, PurePosixPath): - """Path subclass for non-Windows systems. - - On a POSIX system, instantiating a Path should return this object. - """ - __slots__ = () - - if os.name == 'nt': - def __new__(cls, *args, **kwargs): - raise NotImplementedError( - f"cannot instantiate {cls.__name__!r} on your system") - -class WindowsPath(Path, PureWindowsPath): - """Path subclass for Windows systems. - - On a Windows system, instantiating a Path should return this object. - """ - __slots__ = () - - if os.name != 'nt': - def __new__(cls, *args, **kwargs): - raise NotImplementedError( - f"cannot instantiate {cls.__name__!r} on your system") diff --git a/Lib/pathlib/__init__.py b/Lib/pathlib/__init__.py new file mode 100644 index 00000000000..4b3edf535a6 --- /dev/null +++ b/Lib/pathlib/__init__.py @@ -0,0 +1,12 @@ +"""Object-oriented filesystem paths. + +This module provides classes to represent abstract paths and concrete +paths with operations that have semantics appropriate for different +operating systems. +""" + +from ._abc import * +from ._local import * + +__all__ = (_abc.__all__ + + _local.__all__) diff --git a/Lib/pathlib/_abc.py b/Lib/pathlib/_abc.py new file mode 100644 index 00000000000..4d24146aa53 --- /dev/null +++ b/Lib/pathlib/_abc.py @@ -0,0 +1,930 @@ +""" +Abstract base classes for rich path objects. + +This module is published as a PyPI package called "pathlib-abc". + +This module is also a *PRIVATE* part of the Python standard library, where +it's developed alongside pathlib. If it finds success and maturity as a PyPI +package, it could become a public part of the standard library. + +Two base classes are defined here -- PurePathBase and PathBase -- that +resemble pathlib's PurePath and Path respectively. +""" + +import functools +from glob import _Globber, _no_recurse_symlinks +from errno import ENOENT, ENOTDIR, EBADF, ELOOP, EINVAL +from stat import S_ISDIR, S_ISLNK, S_ISREG, S_ISSOCK, S_ISBLK, S_ISCHR, S_ISFIFO + + +__all__ = ["UnsupportedOperation"] + +# +# Internals +# + +_WINERROR_NOT_READY = 21 # drive exists but is not accessible +_WINERROR_INVALID_NAME = 123 # fix for bpo-35306 +_WINERROR_CANT_RESOLVE_FILENAME = 1921 # broken symlink pointing to itself + +# EBADF - guard against macOS `stat` throwing EBADF +_IGNORED_ERRNOS = (ENOENT, ENOTDIR, EBADF, ELOOP) + +_IGNORED_WINERRORS = ( + _WINERROR_NOT_READY, + _WINERROR_INVALID_NAME, + _WINERROR_CANT_RESOLVE_FILENAME) + +def _ignore_error(exception): + return (getattr(exception, 'errno', None) in _IGNORED_ERRNOS or + getattr(exception, 'winerror', None) in _IGNORED_WINERRORS) + + +@functools.cache +def _is_case_sensitive(parser): + return parser.normcase('Aa') == 'Aa' + + +class UnsupportedOperation(NotImplementedError): + """An exception that is raised when an unsupported operation is called on + a path object. + """ + pass + + +class ParserBase: + """Base class for path parsers, which do low-level path manipulation. + + Path parsers provide a subset of the os.path API, specifically those + functions needed to provide PurePathBase functionality. Each PurePathBase + subclass references its path parser via a 'parser' class attribute. + + Every method in this base class raises an UnsupportedOperation exception. + """ + + @classmethod + def _unsupported_msg(cls, attribute): + return f"{cls.__name__}.{attribute} is unsupported" + + @property + def sep(self): + """The character used to separate path components.""" + raise UnsupportedOperation(self._unsupported_msg('sep')) + + def join(self, path, *paths): + """Join path segments.""" + raise UnsupportedOperation(self._unsupported_msg('join()')) + + def split(self, path): + """Split the path into a pair (head, tail), where *head* is everything + before the final path separator, and *tail* is everything after. + Either part may be empty. + """ + raise UnsupportedOperation(self._unsupported_msg('split()')) + + def splitdrive(self, path): + """Split the path into a 2-item tuple (drive, tail), where *drive* is + a device name or mount point, and *tail* is everything after the + drive. Either part may be empty.""" + raise UnsupportedOperation(self._unsupported_msg('splitdrive()')) + + def normcase(self, path): + """Normalize the case of the path.""" + raise UnsupportedOperation(self._unsupported_msg('normcase()')) + + def isabs(self, path): + """Returns whether the path is absolute, i.e. unaffected by the + current directory or drive.""" + raise UnsupportedOperation(self._unsupported_msg('isabs()')) + + +class PurePathBase: + """Base class for pure path objects. + + This class *does not* provide several magic methods that are defined in + its subclass PurePath. They are: __fspath__, __bytes__, __reduce__, + __hash__, __eq__, __lt__, __le__, __gt__, __ge__. Its initializer and path + joining methods accept only strings, not os.PathLike objects more broadly. + """ + + __slots__ = ( + # The `_raw_path` slot store a joined string path. This is set in the + # `__init__()` method. + '_raw_path', + + # The '_resolving' slot stores a boolean indicating whether the path + # is being processed by `PathBase.resolve()`. This prevents duplicate + # work from occurring when `resolve()` calls `stat()` or `readlink()`. + '_resolving', + ) + parser = ParserBase() + _globber = _Globber + + def __init__(self, path, *paths): + self._raw_path = self.parser.join(path, *paths) if paths else path + if not isinstance(self._raw_path, str): + raise TypeError( + f"path should be a str, not {type(self._raw_path).__name__!r}") + self._resolving = False + + def with_segments(self, *pathsegments): + """Construct a new path object from any number of path-like objects. + Subclasses may override this method to customize how new path objects + are created from methods like `iterdir()`. + """ + return type(self)(*pathsegments) + + def __str__(self): + """Return the string representation of the path, suitable for + passing to system calls.""" + return self._raw_path + + def as_posix(self): + """Return the string representation of the path with forward (/) + slashes.""" + return str(self).replace(self.parser.sep, '/') + + @property + def drive(self): + """The drive prefix (letter or UNC path), if any.""" + return self.parser.splitdrive(self.anchor)[0] + + @property + def root(self): + """The root of the path, if any.""" + return self.parser.splitdrive(self.anchor)[1] + + @property + def anchor(self): + """The concatenation of the drive and root, or ''.""" + return self._stack[0] + + @property + def name(self): + """The final path component, if any.""" + return self.parser.split(self._raw_path)[1] + + @property + def suffix(self): + """ + The final component's last suffix, if any. + + This includes the leading period. For example: '.txt' + """ + name = self.name + i = name.rfind('.') + if 0 < i < len(name) - 1: + return name[i:] + else: + return '' + + @property + def suffixes(self): + """ + A list of the final component's suffixes, if any. + + These include the leading periods. For example: ['.tar', '.gz'] + """ + name = self.name + if name.endswith('.'): + return [] + name = name.lstrip('.') + return ['.' + suffix for suffix in name.split('.')[1:]] + + @property + def stem(self): + """The final path component, minus its last suffix.""" + name = self.name + i = name.rfind('.') + if 0 < i < len(name) - 1: + return name[:i] + else: + return name + + def with_name(self, name): + """Return a new path with the file name changed.""" + split = self.parser.split + if split(name)[0]: + raise ValueError(f"Invalid name {name!r}") + return self.with_segments(split(self._raw_path)[0], name) + + def with_stem(self, stem): + """Return a new path with the stem changed.""" + suffix = self.suffix + if not suffix: + return self.with_name(stem) + elif not stem: + # If the suffix is non-empty, we can't make the stem empty. + raise ValueError(f"{self!r} has a non-empty suffix") + else: + return self.with_name(stem + suffix) + + def with_suffix(self, suffix): + """Return a new path with the file suffix changed. If the path + has no suffix, add given suffix. If the given suffix is an empty + string, remove the suffix from the path. + """ + stem = self.stem + if not stem: + # If the stem is empty, we can't make the suffix non-empty. + raise ValueError(f"{self!r} has an empty name") + elif suffix and not (suffix.startswith('.') and len(suffix) > 1): + raise ValueError(f"Invalid suffix {suffix!r}") + else: + return self.with_name(stem + suffix) + + def relative_to(self, other, *, walk_up=False): + """Return the relative path to another path identified by the passed + arguments. If the operation is not possible (because this is not + related to the other path), raise ValueError. + + The *walk_up* parameter controls whether `..` may be used to resolve + the path. + """ + if not isinstance(other, PurePathBase): + other = self.with_segments(other) + anchor0, parts0 = self._stack + anchor1, parts1 = other._stack + if anchor0 != anchor1: + raise ValueError(f"{self._raw_path!r} and {other._raw_path!r} have different anchors") + while parts0 and parts1 and parts0[-1] == parts1[-1]: + parts0.pop() + parts1.pop() + for part in parts1: + if not part or part == '.': + pass + elif not walk_up: + raise ValueError(f"{self._raw_path!r} is not in the subpath of {other._raw_path!r}") + elif part == '..': + raise ValueError(f"'..' segment in {other._raw_path!r} cannot be walked") + else: + parts0.append('..') + return self.with_segments('', *reversed(parts0)) + + def is_relative_to(self, other): + """Return True if the path is relative to another path or False. + """ + if not isinstance(other, PurePathBase): + other = self.with_segments(other) + anchor0, parts0 = self._stack + anchor1, parts1 = other._stack + if anchor0 != anchor1: + return False + while parts0 and parts1 and parts0[-1] == parts1[-1]: + parts0.pop() + parts1.pop() + for part in parts1: + if part and part != '.': + return False + return True + + @property + def parts(self): + """An object providing sequence-like access to the + components in the filesystem path.""" + anchor, parts = self._stack + if anchor: + parts.append(anchor) + return tuple(reversed(parts)) + + def joinpath(self, *pathsegments): + """Combine this path with one or several arguments, and return a + new path representing either a subpath (if all arguments are relative + paths) or a totally different path (if one of the arguments is + anchored). + """ + return self.with_segments(self._raw_path, *pathsegments) + + def __truediv__(self, key): + try: + return self.with_segments(self._raw_path, key) + except TypeError: + return NotImplemented + + def __rtruediv__(self, key): + try: + return self.with_segments(key, self._raw_path) + except TypeError: + return NotImplemented + + @property + def _stack(self): + """ + Split the path into a 2-tuple (anchor, parts), where *anchor* is the + uppermost parent of the path (equivalent to path.parents[-1]), and + *parts* is a reversed list of parts following the anchor. + """ + split = self.parser.split + path = self._raw_path + parent, name = split(path) + names = [] + while path != parent: + names.append(name) + path = parent + parent, name = split(path) + return path, names + + @property + def parent(self): + """The logical parent of the path.""" + path = self._raw_path + parent = self.parser.split(path)[0] + if path != parent: + parent = self.with_segments(parent) + parent._resolving = self._resolving + return parent + return self + + @property + def parents(self): + """A sequence of this path's logical parents.""" + split = self.parser.split + path = self._raw_path + parent = split(path)[0] + parents = [] + while path != parent: + parents.append(self.with_segments(parent)) + path = parent + parent = split(path)[0] + return tuple(parents) + + def is_absolute(self): + """True if the path is absolute (has both a root and, if applicable, + a drive).""" + return self.parser.isabs(self._raw_path) + + @property + def _pattern_str(self): + """The path expressed as a string, for use in pattern-matching.""" + return str(self) + + def match(self, path_pattern, *, case_sensitive=None): + """ + Return True if this path matches the given pattern. If the pattern is + relative, matching is done from the right; otherwise, the entire path + is matched. The recursive wildcard '**' is *not* supported by this + method. + """ + if not isinstance(path_pattern, PurePathBase): + path_pattern = self.with_segments(path_pattern) + if case_sensitive is None: + case_sensitive = _is_case_sensitive(self.parser) + sep = path_pattern.parser.sep + path_parts = self.parts[::-1] + pattern_parts = path_pattern.parts[::-1] + if not pattern_parts: + raise ValueError("empty pattern") + if len(path_parts) < len(pattern_parts): + return False + if len(path_parts) > len(pattern_parts) and path_pattern.anchor: + return False + globber = self._globber(sep, case_sensitive) + for path_part, pattern_part in zip(path_parts, pattern_parts): + match = globber.compile(pattern_part) + if match(path_part) is None: + return False + return True + + def full_match(self, pattern, *, case_sensitive=None): + """ + Return True if this path matches the given glob-style pattern. The + pattern is matched against the entire path. + """ + if not isinstance(pattern, PurePathBase): + pattern = self.with_segments(pattern) + if case_sensitive is None: + case_sensitive = _is_case_sensitive(self.parser) + globber = self._globber(pattern.parser.sep, case_sensitive, recursive=True) + match = globber.compile(pattern._pattern_str) + return match(self._pattern_str) is not None + + + +class PathBase(PurePathBase): + """Base class for concrete path objects. + + This class provides dummy implementations for many methods that derived + classes can override selectively; the default implementations raise + UnsupportedOperation. The most basic methods, such as stat() and open(), + directly raise UnsupportedOperation; these basic methods are called by + other methods such as is_dir() and read_text(). + + The Path class derives this class to implement local filesystem paths. + Users may derive their own classes to implement virtual filesystem paths, + such as paths in archive files or on remote storage systems. + """ + __slots__ = () + + # Maximum number of symlinks to follow in resolve() + _max_symlinks = 40 + + @classmethod + def _unsupported_msg(cls, attribute): + return f"{cls.__name__}.{attribute} is unsupported" + + def stat(self, *, follow_symlinks=True): + """ + Return the result of the stat() system call on this path, like + os.stat() does. + """ + raise UnsupportedOperation(self._unsupported_msg('stat()')) + + def lstat(self): + """ + Like stat(), except if the path points to a symlink, the symlink's + status information is returned, rather than its target's. + """ + return self.stat(follow_symlinks=False) + + + # Convenience functions for querying the stat results + + def exists(self, *, follow_symlinks=True): + """ + Whether this path exists. + + This method normally follows symlinks; to check whether a symlink exists, + add the argument follow_symlinks=False. + """ + try: + self.stat(follow_symlinks=follow_symlinks) + except OSError as e: + if not _ignore_error(e): + raise + return False + except ValueError: + # Non-encodable path + return False + return True + + def is_dir(self, *, follow_symlinks=True): + """ + Whether this path is a directory. + """ + try: + return S_ISDIR(self.stat(follow_symlinks=follow_symlinks).st_mode) + except OSError as e: + if not _ignore_error(e): + raise + # Path doesn't exist or is a broken symlink + # (see http://web.archive.org/web/20200623061726/https://bitbucket.org/pitrou/pathlib/issues/12/ ) + return False + except ValueError: + # Non-encodable path + return False + + def is_file(self, *, follow_symlinks=True): + """ + Whether this path is a regular file (also True for symlinks pointing + to regular files). + """ + try: + return S_ISREG(self.stat(follow_symlinks=follow_symlinks).st_mode) + except OSError as e: + if not _ignore_error(e): + raise + # Path doesn't exist or is a broken symlink + # (see http://web.archive.org/web/20200623061726/https://bitbucket.org/pitrou/pathlib/issues/12/ ) + return False + except ValueError: + # Non-encodable path + return False + + def is_mount(self): + """ + Check if this path is a mount point + """ + # Need to exist and be a dir + if not self.exists() or not self.is_dir(): + return False + + try: + parent_dev = self.parent.stat().st_dev + except OSError: + return False + + dev = self.stat().st_dev + if dev != parent_dev: + return True + ino = self.stat().st_ino + parent_ino = self.parent.stat().st_ino + return ino == parent_ino + + def is_symlink(self): + """ + Whether this path is a symbolic link. + """ + try: + return S_ISLNK(self.lstat().st_mode) + except OSError as e: + if not _ignore_error(e): + raise + # Path doesn't exist + return False + except ValueError: + # Non-encodable path + return False + + def is_junction(self): + """ + Whether this path is a junction. + """ + # Junctions are a Windows-only feature, not present in POSIX nor the + # majority of virtual filesystems. There is no cross-platform idiom + # to check for junctions (using stat().st_mode). + return False + + def is_block_device(self): + """ + Whether this path is a block device. + """ + try: + return S_ISBLK(self.stat().st_mode) + except OSError as e: + if not _ignore_error(e): + raise + # Path doesn't exist or is a broken symlink + # (see http://web.archive.org/web/20200623061726/https://bitbucket.org/pitrou/pathlib/issues/12/ ) + return False + except ValueError: + # Non-encodable path + return False + + def is_char_device(self): + """ + Whether this path is a character device. + """ + try: + return S_ISCHR(self.stat().st_mode) + except OSError as e: + if not _ignore_error(e): + raise + # Path doesn't exist or is a broken symlink + # (see http://web.archive.org/web/20200623061726/https://bitbucket.org/pitrou/pathlib/issues/12/ ) + return False + except ValueError: + # Non-encodable path + return False + + def is_fifo(self): + """ + Whether this path is a FIFO. + """ + try: + return S_ISFIFO(self.stat().st_mode) + except OSError as e: + if not _ignore_error(e): + raise + # Path doesn't exist or is a broken symlink + # (see http://web.archive.org/web/20200623061726/https://bitbucket.org/pitrou/pathlib/issues/12/ ) + return False + except ValueError: + # Non-encodable path + return False + + def is_socket(self): + """ + Whether this path is a socket. + """ + try: + return S_ISSOCK(self.stat().st_mode) + except OSError as e: + if not _ignore_error(e): + raise + # Path doesn't exist or is a broken symlink + # (see http://web.archive.org/web/20200623061726/https://bitbucket.org/pitrou/pathlib/issues/12/ ) + return False + except ValueError: + # Non-encodable path + return False + + def samefile(self, other_path): + """Return whether other_path is the same or not as this file + (as returned by os.path.samefile()). + """ + st = self.stat() + try: + other_st = other_path.stat() + except AttributeError: + other_st = self.with_segments(other_path).stat() + return (st.st_ino == other_st.st_ino and + st.st_dev == other_st.st_dev) + + def open(self, mode='r', buffering=-1, encoding=None, + errors=None, newline=None): + """ + Open the file pointed to by this path and return a file object, as + the built-in open() function does. + """ + raise UnsupportedOperation(self._unsupported_msg('open()')) + + def read_bytes(self): + """ + Open the file in bytes mode, read it, and close the file. + """ + with self.open(mode='rb') as f: + return f.read() + + def read_text(self, encoding=None, errors=None, newline=None): + """ + Open the file in text mode, read it, and close the file. + """ + with self.open(mode='r', encoding=encoding, errors=errors, newline=newline) as f: + return f.read() + + def write_bytes(self, data): + """ + Open the file in bytes mode, write to it, and close the file. + """ + # type-check for the buffer interface before truncating the file + view = memoryview(data) + with self.open(mode='wb') as f: + return f.write(view) + + def write_text(self, data, encoding=None, errors=None, newline=None): + """ + Open the file in text mode, write to it, and close the file. + """ + if not isinstance(data, str): + raise TypeError('data must be str, not %s' % + data.__class__.__name__) + with self.open(mode='w', encoding=encoding, errors=errors, newline=newline) as f: + return f.write(data) + + def iterdir(self): + """Yield path objects of the directory contents. + + The children are yielded in arbitrary order, and the + special entries '.' and '..' are not included. + """ + raise UnsupportedOperation(self._unsupported_msg('iterdir()')) + + def _glob_selector(self, parts, case_sensitive, recurse_symlinks): + if case_sensitive is None: + case_sensitive = _is_case_sensitive(self.parser) + case_pedantic = False + else: + # The user has expressed a case sensitivity choice, but we don't + # know the case sensitivity of the underlying filesystem, so we + # must use scandir() for everything, including non-wildcard parts. + case_pedantic = True + recursive = True if recurse_symlinks else _no_recurse_symlinks + globber = self._globber(self.parser.sep, case_sensitive, case_pedantic, recursive) + return globber.selector(parts) + + def glob(self, pattern, *, case_sensitive=None, recurse_symlinks=True): + """Iterate over this subtree and yield all existing files (of any + kind, including directories) matching the given relative pattern. + """ + if not isinstance(pattern, PurePathBase): + pattern = self.with_segments(pattern) + anchor, parts = pattern._stack + if anchor: + raise NotImplementedError("Non-relative patterns are unsupported") + select = self._glob_selector(parts, case_sensitive, recurse_symlinks) + return select(self) + + def rglob(self, pattern, *, case_sensitive=None, recurse_symlinks=True): + """Recursively yield all existing files (of any kind, including + directories) matching the given relative pattern, anywhere in + this subtree. + """ + if not isinstance(pattern, PurePathBase): + pattern = self.with_segments(pattern) + pattern = '**' / pattern + return self.glob(pattern, case_sensitive=case_sensitive, recurse_symlinks=recurse_symlinks) + + def walk(self, top_down=True, on_error=None, follow_symlinks=False): + """Walk the directory tree from this directory, similar to os.walk().""" + paths = [self] + while paths: + path = paths.pop() + if isinstance(path, tuple): + yield path + continue + dirnames = [] + filenames = [] + if not top_down: + paths.append((path, dirnames, filenames)) + try: + for child in path.iterdir(): + try: + if child.is_dir(follow_symlinks=follow_symlinks): + if not top_down: + paths.append(child) + dirnames.append(child.name) + else: + filenames.append(child.name) + except OSError: + filenames.append(child.name) + except OSError as error: + if on_error is not None: + on_error(error) + if not top_down: + while not isinstance(paths.pop(), tuple): + pass + continue + if top_down: + yield path, dirnames, filenames + paths += [path.joinpath(d) for d in reversed(dirnames)] + + def absolute(self): + """Return an absolute version of this path + No normalization or symlink resolution is performed. + + Use resolve() to resolve symlinks and remove '..' segments. + """ + raise UnsupportedOperation(self._unsupported_msg('absolute()')) + + @classmethod + def cwd(cls): + """Return a new path pointing to the current working directory.""" + # We call 'absolute()' rather than using 'os.getcwd()' directly to + # enable users to replace the implementation of 'absolute()' in a + # subclass and benefit from the new behaviour here. This works because + # os.path.abspath('.') == os.getcwd(). + return cls('').absolute() + + def expanduser(self): + """ Return a new path with expanded ~ and ~user constructs + (as returned by os.path.expanduser) + """ + raise UnsupportedOperation(self._unsupported_msg('expanduser()')) + + @classmethod + def home(cls): + """Return a new path pointing to expanduser('~'). + """ + return cls("~").expanduser() + + def readlink(self): + """ + Return the path to which the symbolic link points. + """ + raise UnsupportedOperation(self._unsupported_msg('readlink()')) + readlink._supported = False + + def resolve(self, strict=False): + """ + Make the path absolute, resolving all symlinks on the way and also + normalizing it. + """ + if self._resolving: + return self + path_root, parts = self._stack + path = self.with_segments(path_root) + try: + path = path.absolute() + except UnsupportedOperation: + path_tail = [] + else: + path_root, path_tail = path._stack + path_tail.reverse() + + # If the user has *not* overridden the `readlink()` method, then symlinks are unsupported + # and (in non-strict mode) we can improve performance by not calling `stat()`. + querying = strict or getattr(self.readlink, '_supported', True) + link_count = 0 + while parts: + part = parts.pop() + if not part or part == '.': + continue + if part == '..': + if not path_tail: + if path_root: + # Delete '..' segment immediately following root + continue + elif path_tail[-1] != '..': + # Delete '..' segment and its predecessor + path_tail.pop() + continue + path_tail.append(part) + if querying and part != '..': + path = self.with_segments(path_root + self.parser.sep.join(path_tail)) + path._resolving = True + try: + st = path.stat(follow_symlinks=False) + if S_ISLNK(st.st_mode): + # Like Linux and macOS, raise OSError(errno.ELOOP) if too many symlinks are + # encountered during resolution. + link_count += 1 + if link_count >= self._max_symlinks: + raise OSError(ELOOP, "Too many symbolic links in path", self._raw_path) + target_root, target_parts = path.readlink()._stack + # If the symlink target is absolute (like '/etc/hosts'), set the current + # path to its uppermost parent (like '/'). + if target_root: + path_root = target_root + path_tail.clear() + else: + path_tail.pop() + # Add the symlink target's reversed tail parts (like ['hosts', 'etc']) to + # the stack of unresolved path parts. + parts.extend(target_parts) + continue + elif parts and not S_ISDIR(st.st_mode): + raise NotADirectoryError(ENOTDIR, "Not a directory", self._raw_path) + except OSError: + if strict: + raise + else: + querying = False + return self.with_segments(path_root + self.parser.sep.join(path_tail)) + + def symlink_to(self, target, target_is_directory=False): + """ + Make this path a symlink pointing to the target path. + Note the order of arguments (link, target) is the reverse of os.symlink. + """ + raise UnsupportedOperation(self._unsupported_msg('symlink_to()')) + + def hardlink_to(self, target): + """ + Make this path a hard link pointing to the same file as *target*. + + Note the order of arguments (self, target) is the reverse of os.link's. + """ + raise UnsupportedOperation(self._unsupported_msg('hardlink_to()')) + + def touch(self, mode=0o666, exist_ok=True): + """ + Create this file with the given access mode, if it doesn't exist. + """ + raise UnsupportedOperation(self._unsupported_msg('touch()')) + + def mkdir(self, mode=0o777, parents=False, exist_ok=False): + """ + Create a new directory at this given path. + """ + raise UnsupportedOperation(self._unsupported_msg('mkdir()')) + + def rename(self, target): + """ + Rename this path to the target path. + + The target path may be absolute or relative. Relative paths are + interpreted relative to the current working directory, *not* the + directory of the Path object. + + Returns the new Path instance pointing to the target path. + """ + raise UnsupportedOperation(self._unsupported_msg('rename()')) + + def replace(self, target): + """ + Rename this path to the target path, overwriting if that path exists. + + The target path may be absolute or relative. Relative paths are + interpreted relative to the current working directory, *not* the + directory of the Path object. + + Returns the new Path instance pointing to the target path. + """ + raise UnsupportedOperation(self._unsupported_msg('replace()')) + + def chmod(self, mode, *, follow_symlinks=True): + """ + Change the permissions of the path, like os.chmod(). + """ + raise UnsupportedOperation(self._unsupported_msg('chmod()')) + + def lchmod(self, mode): + """ + Like chmod(), except if the path points to a symlink, the symlink's + permissions are changed, rather than its target's. + """ + self.chmod(mode, follow_symlinks=False) + + def unlink(self, missing_ok=False): + """ + Remove this file or link. + If the path is a directory, use rmdir() instead. + """ + raise UnsupportedOperation(self._unsupported_msg('unlink()')) + + def rmdir(self): + """ + Remove this directory. The directory must be empty. + """ + raise UnsupportedOperation(self._unsupported_msg('rmdir()')) + + def owner(self, *, follow_symlinks=True): + """ + Return the login name of the file owner. + """ + raise UnsupportedOperation(self._unsupported_msg('owner()')) + + def group(self, *, follow_symlinks=True): + """ + Return the group name of the file gid. + """ + raise UnsupportedOperation(self._unsupported_msg('group()')) + + @classmethod + def from_uri(cls, uri): + """Return a new path from the given 'file' URI.""" + raise UnsupportedOperation(cls._unsupported_msg('from_uri()')) + + def as_uri(self): + """Return the path as a URI.""" + raise UnsupportedOperation(self._unsupported_msg('as_uri()')) diff --git a/Lib/pathlib/_local.py b/Lib/pathlib/_local.py new file mode 100644 index 00000000000..0188e7c7722 --- /dev/null +++ b/Lib/pathlib/_local.py @@ -0,0 +1,861 @@ +import io +import ntpath +import operator +import os +import posixpath +import sys +import warnings +from glob import _StringGlobber +from itertools import chain +from _collections_abc import Sequence + +try: + import pwd +except ImportError: + pwd = None +try: + import grp +except ImportError: + grp = None + +from ._abc import UnsupportedOperation, PurePathBase, PathBase + + +__all__ = [ + "PurePath", "PurePosixPath", "PureWindowsPath", + "Path", "PosixPath", "WindowsPath", + ] + + +class _PathParents(Sequence): + """This object provides sequence-like access to the logical ancestors + of a path. Don't try to construct it yourself.""" + __slots__ = ('_path', '_drv', '_root', '_tail') + + def __init__(self, path): + self._path = path + self._drv = path.drive + self._root = path.root + self._tail = path._tail + + def __len__(self): + return len(self._tail) + + def __getitem__(self, idx): + if isinstance(idx, slice): + return tuple(self[i] for i in range(*idx.indices(len(self)))) + + if idx >= len(self) or idx < -len(self): + raise IndexError(idx) + if idx < 0: + idx += len(self) + return self._path._from_parsed_parts(self._drv, self._root, + self._tail[:-idx - 1]) + + def __repr__(self): + return "<{}.parents>".format(type(self._path).__name__) + + +class PurePath(PurePathBase): + """Base class for manipulating paths without I/O. + + PurePath represents a filesystem path and offers operations which + don't imply any actual filesystem I/O. Depending on your system, + instantiating a PurePath will return either a PurePosixPath or a + PureWindowsPath object. You can also instantiate either of these classes + directly, regardless of your system. + """ + + __slots__ = ( + # The `_raw_paths` slot stores unnormalized string paths. This is set + # in the `__init__()` method. + '_raw_paths', + + # The `_drv`, `_root` and `_tail_cached` slots store parsed and + # normalized parts of the path. They are set when any of the `drive`, + # `root` or `_tail` properties are accessed for the first time. The + # three-part division corresponds to the result of + # `os.path.splitroot()`, except that the tail is further split on path + # separators (i.e. it is a list of strings), and that the root and + # tail are normalized. + '_drv', '_root', '_tail_cached', + + # The `_str` slot stores the string representation of the path, + # computed from the drive, root and tail when `__str__()` is called + # for the first time. It's used to implement `_str_normcase` + '_str', + + # The `_str_normcase_cached` slot stores the string path with + # normalized case. It is set when the `_str_normcase` property is + # accessed for the first time. It's used to implement `__eq__()` + # `__hash__()`, and `_parts_normcase` + '_str_normcase_cached', + + # The `_parts_normcase_cached` slot stores the case-normalized + # string path after splitting on path separators. It's set when the + # `_parts_normcase` property is accessed for the first time. It's used + # to implement comparison methods like `__lt__()`. + '_parts_normcase_cached', + + # The `_hash` slot stores the hash of the case-normalized string + # path. It's set when `__hash__()` is called for the first time. + '_hash', + ) + parser = os.path + _globber = _StringGlobber + + def __new__(cls, *args, **kwargs): + """Construct a PurePath from one or several strings and or existing + PurePath objects. The strings and path objects are combined so as + to yield a canonicalized path, which is incorporated into the + new PurePath object. + """ + if cls is PurePath: + cls = PureWindowsPath if os.name == 'nt' else PurePosixPath + return object.__new__(cls) + + def __init__(self, *args): + paths = [] + for arg in args: + if isinstance(arg, PurePath): + if arg.parser is not self.parser: + # GH-103631: Convert separators for backwards compatibility. + paths.append(arg.as_posix()) + else: + paths.extend(arg._raw_paths) + else: + try: + path = os.fspath(arg) + except TypeError: + path = arg + if not isinstance(path, str): + raise TypeError( + "argument should be a str or an os.PathLike " + "object where __fspath__ returns a str, " + f"not {type(path).__name__!r}") + paths.append(path) + # Avoid calling super().__init__, as an optimisation + self._raw_paths = paths + + def joinpath(self, *pathsegments): + """Combine this path with one or several arguments, and return a + new path representing either a subpath (if all arguments are relative + paths) or a totally different path (if one of the arguments is + anchored). + """ + return self.with_segments(self, *pathsegments) + + def __truediv__(self, key): + try: + return self.with_segments(self, key) + except TypeError: + return NotImplemented + + def __rtruediv__(self, key): + try: + return self.with_segments(key, self) + except TypeError: + return NotImplemented + + def __reduce__(self): + return self.__class__, tuple(self._raw_paths) + + def __repr__(self): + return "{}({!r})".format(self.__class__.__name__, self.as_posix()) + + def __fspath__(self): + return str(self) + + def __bytes__(self): + """Return the bytes representation of the path. This is only + recommended to use under Unix.""" + return os.fsencode(self) + + @property + def _str_normcase(self): + # String with normalized case, for hashing and equality checks + try: + return self._str_normcase_cached + except AttributeError: + if self.parser is posixpath: + self._str_normcase_cached = str(self) + else: + self._str_normcase_cached = str(self).lower() + return self._str_normcase_cached + + def __hash__(self): + try: + return self._hash + except AttributeError: + self._hash = hash(self._str_normcase) + return self._hash + + def __eq__(self, other): + if not isinstance(other, PurePath): + return NotImplemented + return self._str_normcase == other._str_normcase and self.parser is other.parser + + @property + def _parts_normcase(self): + # Cached parts with normalized case, for comparisons. + try: + return self._parts_normcase_cached + except AttributeError: + self._parts_normcase_cached = self._str_normcase.split(self.parser.sep) + return self._parts_normcase_cached + + def __lt__(self, other): + if not isinstance(other, PurePath) or self.parser is not other.parser: + return NotImplemented + return self._parts_normcase < other._parts_normcase + + def __le__(self, other): + if not isinstance(other, PurePath) or self.parser is not other.parser: + return NotImplemented + return self._parts_normcase <= other._parts_normcase + + def __gt__(self, other): + if not isinstance(other, PurePath) or self.parser is not other.parser: + return NotImplemented + return self._parts_normcase > other._parts_normcase + + def __ge__(self, other): + if not isinstance(other, PurePath) or self.parser is not other.parser: + return NotImplemented + return self._parts_normcase >= other._parts_normcase + + def __str__(self): + """Return the string representation of the path, suitable for + passing to system calls.""" + try: + return self._str + except AttributeError: + self._str = self._format_parsed_parts(self.drive, self.root, + self._tail) or '.' + return self._str + + @classmethod + def _format_parsed_parts(cls, drv, root, tail): + if drv or root: + return drv + root + cls.parser.sep.join(tail) + elif tail and cls.parser.splitdrive(tail[0])[0]: + tail = ['.'] + tail + return cls.parser.sep.join(tail) + + def _from_parsed_parts(self, drv, root, tail): + path = self._from_parsed_string(self._format_parsed_parts(drv, root, tail)) + path._drv = drv + path._root = root + path._tail_cached = tail + return path + + def _from_parsed_string(self, path_str): + path = self.with_segments(path_str) + path._str = path_str or '.' + return path + + @classmethod + def _parse_path(cls, path): + if not path: + return '', '', [] + sep = cls.parser.sep + altsep = cls.parser.altsep + if altsep: + path = path.replace(altsep, sep) + drv, root, rel = cls.parser.splitroot(path) + if not root and drv.startswith(sep) and not drv.endswith(sep): + drv_parts = drv.split(sep) + if len(drv_parts) == 4 and drv_parts[2] not in '?.': + # e.g. //server/share + root = sep + elif len(drv_parts) == 6: + # e.g. //?/unc/server/share + root = sep + parsed = [sys.intern(str(x)) for x in rel.split(sep) if x and x != '.'] + return drv, root, parsed + + @property + def _raw_path(self): + """The joined but unnormalized path.""" + paths = self._raw_paths + if len(paths) == 0: + path = '' + elif len(paths) == 1: + path = paths[0] + else: + path = self.parser.join(*paths) + return path + + @property + def drive(self): + """The drive prefix (letter or UNC path), if any.""" + try: + return self._drv + except AttributeError: + self._drv, self._root, self._tail_cached = self._parse_path(self._raw_path) + return self._drv + + @property + def root(self): + """The root of the path, if any.""" + try: + return self._root + except AttributeError: + self._drv, self._root, self._tail_cached = self._parse_path(self._raw_path) + return self._root + + @property + def _tail(self): + try: + return self._tail_cached + except AttributeError: + self._drv, self._root, self._tail_cached = self._parse_path(self._raw_path) + return self._tail_cached + + @property + def anchor(self): + """The concatenation of the drive and root, or ''.""" + return self.drive + self.root + + @property + def parts(self): + """An object providing sequence-like access to the + components in the filesystem path.""" + if self.drive or self.root: + return (self.drive + self.root,) + tuple(self._tail) + else: + return tuple(self._tail) + + @property + def parent(self): + """The logical parent of the path.""" + drv = self.drive + root = self.root + tail = self._tail + if not tail: + return self + return self._from_parsed_parts(drv, root, tail[:-1]) + + @property + def parents(self): + """A sequence of this path's logical parents.""" + # The value of this property should not be cached on the path object, + # as doing so would introduce a reference cycle. + return _PathParents(self) + + @property + def name(self): + """The final path component, if any.""" + tail = self._tail + if not tail: + return '' + return tail[-1] + + def with_name(self, name): + """Return a new path with the file name changed.""" + p = self.parser + if not name or p.sep in name or (p.altsep and p.altsep in name) or name == '.': + raise ValueError(f"Invalid name {name!r}") + tail = self._tail.copy() + if not tail: + raise ValueError(f"{self!r} has an empty name") + tail[-1] = name + return self._from_parsed_parts(self.drive, self.root, tail) + + def relative_to(self, other, /, *_deprecated, walk_up=False): + """Return the relative path to another path identified by the passed + arguments. If the operation is not possible (because this is not + related to the other path), raise ValueError. + + The *walk_up* parameter controls whether `..` may be used to resolve + the path. + """ + if _deprecated: + msg = ("support for supplying more than one positional argument " + "to pathlib.PurePath.relative_to() is deprecated and " + "scheduled for removal in Python 3.14") + warnings.warn(msg, DeprecationWarning, stacklevel=2) + other = self.with_segments(other, *_deprecated) + elif not isinstance(other, PurePath): + other = self.with_segments(other) + for step, path in enumerate(chain([other], other.parents)): + if path == self or path in self.parents: + break + elif not walk_up: + raise ValueError(f"{str(self)!r} is not in the subpath of {str(other)!r}") + elif path.name == '..': + raise ValueError(f"'..' segment in {str(other)!r} cannot be walked") + else: + raise ValueError(f"{str(self)!r} and {str(other)!r} have different anchors") + parts = ['..'] * step + self._tail[len(path._tail):] + return self._from_parsed_parts('', '', parts) + + def is_relative_to(self, other, /, *_deprecated): + """Return True if the path is relative to another path or False. + """ + if _deprecated: + msg = ("support for supplying more than one argument to " + "pathlib.PurePath.is_relative_to() is deprecated and " + "scheduled for removal in Python 3.14") + warnings.warn(msg, DeprecationWarning, stacklevel=2) + other = self.with_segments(other, *_deprecated) + elif not isinstance(other, PurePath): + other = self.with_segments(other) + return other == self or other in self.parents + + def is_absolute(self): + """True if the path is absolute (has both a root and, if applicable, + a drive).""" + if self.parser is posixpath: + # Optimization: work with raw paths on POSIX. + for path in self._raw_paths: + if path.startswith('/'): + return True + return False + return self.parser.isabs(self) + + def is_reserved(self): + """Return True if the path contains one of the special names reserved + by the system, if any.""" + msg = ("pathlib.PurePath.is_reserved() is deprecated and scheduled " + "for removal in Python 3.15. Use os.path.isreserved() to " + "detect reserved paths on Windows.") + warnings.warn(msg, DeprecationWarning, stacklevel=2) + if self.parser is ntpath: + return self.parser.isreserved(self) + return False + + def as_uri(self): + """Return the path as a URI.""" + if not self.is_absolute(): + raise ValueError("relative path can't be expressed as a file URI") + + drive = self.drive + if len(drive) == 2 and drive[1] == ':': + # It's a path on a local drive => 'file:///c:/a/b' + prefix = 'file:///' + drive + path = self.as_posix()[2:] + elif drive: + # It's a path on a network drive => 'file://host/share/a/b' + prefix = 'file:' + path = self.as_posix() + else: + # It's a posix path => 'file:///etc/hosts' + prefix = 'file://' + path = str(self) + from urllib.parse import quote_from_bytes + return prefix + quote_from_bytes(os.fsencode(path)) + + @property + def _pattern_str(self): + """The path expressed as a string, for use in pattern-matching.""" + # The string representation of an empty path is a single dot ('.'). Empty + # paths shouldn't match wildcards, so we change it to the empty string. + path_str = str(self) + return '' if path_str == '.' else path_str + +# Subclassing os.PathLike makes isinstance() checks slower, +# which in turn makes Path construction slower. Register instead! +os.PathLike.register(PurePath) + + +class PurePosixPath(PurePath): + """PurePath subclass for non-Windows systems. + + On a POSIX system, instantiating a PurePath should return this object. + However, you can also instantiate it directly on any system. + """ + parser = posixpath + __slots__ = () + + +class PureWindowsPath(PurePath): + """PurePath subclass for Windows systems. + + On a Windows system, instantiating a PurePath should return this object. + However, you can also instantiate it directly on any system. + """ + parser = ntpath + __slots__ = () + + +class Path(PathBase, PurePath): + """PurePath subclass that can make system calls. + + Path represents a filesystem path but unlike PurePath, also offers + methods to do system calls on path objects. Depending on your system, + instantiating a Path will return either a PosixPath or a WindowsPath + object. You can also instantiate a PosixPath or WindowsPath directly, + but cannot instantiate a WindowsPath on a POSIX system or vice versa. + """ + __slots__ = () + as_uri = PurePath.as_uri + + @classmethod + def _unsupported_msg(cls, attribute): + return f"{cls.__name__}.{attribute} is unsupported on this system" + + def __init__(self, *args, **kwargs): + if kwargs: + msg = ("support for supplying keyword arguments to pathlib.PurePath " + "is deprecated and scheduled for removal in Python {remove}") + warnings._deprecated("pathlib.PurePath(**kwargs)", msg, remove=(3, 14)) + super().__init__(*args) + + def __new__(cls, *args, **kwargs): + if cls is Path: + cls = WindowsPath if os.name == 'nt' else PosixPath + return object.__new__(cls) + + def stat(self, *, follow_symlinks=True): + """ + Return the result of the stat() system call on this path, like + os.stat() does. + """ + return os.stat(self, follow_symlinks=follow_symlinks) + + def is_mount(self): + """ + Check if this path is a mount point + """ + return os.path.ismount(self) + + def is_junction(self): + """ + Whether this path is a junction. + """ + return os.path.isjunction(self) + + def open(self, mode='r', buffering=-1, encoding=None, + errors=None, newline=None): + """ + Open the file pointed to by this path and return a file object, as + the built-in open() function does. + """ + if "b" not in mode: + encoding = io.text_encoding(encoding) + return io.open(self, mode, buffering, encoding, errors, newline) + + def read_text(self, encoding=None, errors=None, newline=None): + """ + Open the file in text mode, read it, and close the file. + """ + # Call io.text_encoding() here to ensure any warning is raised at an + # appropriate stack level. + encoding = io.text_encoding(encoding) + return PathBase.read_text(self, encoding, errors, newline) + + def write_text(self, data, encoding=None, errors=None, newline=None): + """ + Open the file in text mode, write to it, and close the file. + """ + # Call io.text_encoding() here to ensure any warning is raised at an + # appropriate stack level. + encoding = io.text_encoding(encoding) + return PathBase.write_text(self, data, encoding, errors, newline) + + _remove_leading_dot = operator.itemgetter(slice(2, None)) + _remove_trailing_slash = operator.itemgetter(slice(-1)) + + def _filter_trailing_slash(self, paths): + sep = self.parser.sep + anchor_len = len(self.anchor) + for path_str in paths: + if len(path_str) > anchor_len and path_str[-1] == sep: + path_str = path_str[:-1] + yield path_str + + def iterdir(self): + """Yield path objects of the directory contents. + + The children are yielded in arbitrary order, and the + special entries '.' and '..' are not included. + """ + root_dir = str(self) + with os.scandir(root_dir) as scandir_it: + paths = [entry.path for entry in scandir_it] + if root_dir == '.': + paths = map(self._remove_leading_dot, paths) + return map(self._from_parsed_string, paths) + + def glob(self, pattern, *, case_sensitive=None, recurse_symlinks=False): + """Iterate over this subtree and yield all existing files (of any + kind, including directories) matching the given relative pattern. + """ + sys.audit("pathlib.Path.glob", self, pattern) + if not isinstance(pattern, PurePath): + pattern = self.with_segments(pattern) + if pattern.anchor: + raise NotImplementedError("Non-relative patterns are unsupported") + parts = pattern._tail.copy() + if not parts: + raise ValueError("Unacceptable pattern: {!r}".format(pattern)) + raw = pattern._raw_path + if raw[-1] in (self.parser.sep, self.parser.altsep): + # GH-65238: pathlib doesn't preserve trailing slash. Add it back. + parts.append('') + select = self._glob_selector(parts[::-1], case_sensitive, recurse_symlinks) + root = str(self) + paths = select(root) + + # Normalize results + if root == '.': + paths = map(self._remove_leading_dot, paths) + if parts[-1] == '': + paths = map(self._remove_trailing_slash, paths) + elif parts[-1] == '**': + paths = self._filter_trailing_slash(paths) + paths = map(self._from_parsed_string, paths) + return paths + + def rglob(self, pattern, *, case_sensitive=None, recurse_symlinks=False): + """Recursively yield all existing files (of any kind, including + directories) matching the given relative pattern, anywhere in + this subtree. + """ + sys.audit("pathlib.Path.rglob", self, pattern) + if not isinstance(pattern, PurePath): + pattern = self.with_segments(pattern) + pattern = '**' / pattern + return self.glob(pattern, case_sensitive=case_sensitive, recurse_symlinks=recurse_symlinks) + + def walk(self, top_down=True, on_error=None, follow_symlinks=False): + """Walk the directory tree from this directory, similar to os.walk().""" + sys.audit("pathlib.Path.walk", self, on_error, follow_symlinks) + root_dir = str(self) + if not follow_symlinks: + follow_symlinks = os._walk_symlinks_as_files + results = os.walk(root_dir, top_down, on_error, follow_symlinks) + for path_str, dirnames, filenames in results: + if root_dir == '.': + path_str = path_str[2:] + yield self._from_parsed_string(path_str), dirnames, filenames + + def absolute(self): + """Return an absolute version of this path + No normalization or symlink resolution is performed. + + Use resolve() to resolve symlinks and remove '..' segments. + """ + if self.is_absolute(): + return self + if self.root: + drive = os.path.splitroot(os.getcwd())[0] + return self._from_parsed_parts(drive, self.root, self._tail) + if self.drive: + # There is a CWD on each drive-letter drive. + cwd = os.path.abspath(self.drive) + else: + cwd = os.getcwd() + if not self._tail: + # Fast path for "empty" paths, e.g. Path("."), Path("") or Path(). + # We pass only one argument to with_segments() to avoid the cost + # of joining, and we exploit the fact that getcwd() returns a + # fully-normalized string by storing it in _str. This is used to + # implement Path.cwd(). + return self._from_parsed_string(cwd) + drive, root, rel = os.path.splitroot(cwd) + if not rel: + return self._from_parsed_parts(drive, root, self._tail) + tail = rel.split(self.parser.sep) + tail.extend(self._tail) + return self._from_parsed_parts(drive, root, tail) + + def resolve(self, strict=False): + """ + Make the path absolute, resolving all symlinks on the way and also + normalizing it. + """ + + return self.with_segments(os.path.realpath(self, strict=strict)) + + if pwd: + def owner(self, *, follow_symlinks=True): + """ + Return the login name of the file owner. + """ + uid = self.stat(follow_symlinks=follow_symlinks).st_uid + return pwd.getpwuid(uid).pw_name + + if grp: + def group(self, *, follow_symlinks=True): + """ + Return the group name of the file gid. + """ + gid = self.stat(follow_symlinks=follow_symlinks).st_gid + return grp.getgrgid(gid).gr_name + + if hasattr(os, "readlink"): + def readlink(self): + """ + Return the path to which the symbolic link points. + """ + return self.with_segments(os.readlink(self)) + + def touch(self, mode=0o666, exist_ok=True): + """ + Create this file with the given access mode, if it doesn't exist. + """ + + if exist_ok: + # First try to bump modification time + # Implementation note: GNU touch uses the UTIME_NOW option of + # the utimensat() / futimens() functions. + try: + os.utime(self, None) + except OSError: + # Avoid exception chaining + pass + else: + return + flags = os.O_CREAT | os.O_WRONLY + if not exist_ok: + flags |= os.O_EXCL + fd = os.open(self, flags, mode) + os.close(fd) + + def mkdir(self, mode=0o777, parents=False, exist_ok=False): + """ + Create a new directory at this given path. + """ + try: + os.mkdir(self, mode) + except FileNotFoundError: + if not parents or self.parent == self: + raise + self.parent.mkdir(parents=True, exist_ok=True) + self.mkdir(mode, parents=False, exist_ok=exist_ok) + except OSError: + # Cannot rely on checking for EEXIST, since the operating system + # could give priority to other errors like EACCES or EROFS + if not exist_ok or not self.is_dir(): + raise + + def chmod(self, mode, *, follow_symlinks=True): + """ + Change the permissions of the path, like os.chmod(). + """ + os.chmod(self, mode, follow_symlinks=follow_symlinks) + + def unlink(self, missing_ok=False): + """ + Remove this file or link. + If the path is a directory, use rmdir() instead. + """ + try: + os.unlink(self) + except FileNotFoundError: + if not missing_ok: + raise + + def rmdir(self): + """ + Remove this directory. The directory must be empty. + """ + os.rmdir(self) + + def rename(self, target): + """ + Rename this path to the target path. + + The target path may be absolute or relative. Relative paths are + interpreted relative to the current working directory, *not* the + directory of the Path object. + + Returns the new Path instance pointing to the target path. + """ + os.rename(self, target) + return self.with_segments(target) + + def replace(self, target): + """ + Rename this path to the target path, overwriting if that path exists. + + The target path may be absolute or relative. Relative paths are + interpreted relative to the current working directory, *not* the + directory of the Path object. + + Returns the new Path instance pointing to the target path. + """ + os.replace(self, target) + return self.with_segments(target) + + if hasattr(os, "symlink"): + def symlink_to(self, target, target_is_directory=False): + """ + Make this path a symlink pointing to the target path. + Note the order of arguments (link, target) is the reverse of os.symlink. + """ + os.symlink(target, self, target_is_directory) + + if hasattr(os, "link"): + def hardlink_to(self, target): + """ + Make this path a hard link pointing to the same file as *target*. + + Note the order of arguments (self, target) is the reverse of os.link's. + """ + os.link(target, self) + + def expanduser(self): + """ Return a new path with expanded ~ and ~user constructs + (as returned by os.path.expanduser) + """ + if (not (self.drive or self.root) and + self._tail and self._tail[0][:1] == '~'): + homedir = os.path.expanduser(self._tail[0]) + if homedir[:1] == "~": + raise RuntimeError("Could not determine home directory.") + drv, root, tail = self._parse_path(homedir) + return self._from_parsed_parts(drv, root, tail + self._tail[1:]) + + return self + + @classmethod + def from_uri(cls, uri): + """Return a new path from the given 'file' URI.""" + if not uri.startswith('file:'): + raise ValueError(f"URI does not start with 'file:': {uri!r}") + path = uri[5:] + if path[:3] == '///': + # Remove empty authority + path = path[2:] + elif path[:12] == '//localhost/': + # Remove 'localhost' authority + path = path[11:] + if path[:3] == '///' or (path[:1] == '/' and path[2:3] in ':|'): + # Remove slash before DOS device/UNC path + path = path[1:] + if path[1:2] == '|': + # Replace bar with colon in DOS drive + path = path[:1] + ':' + path[2:] + from urllib.parse import unquote_to_bytes + path = cls(os.fsdecode(unquote_to_bytes(path))) + if not path.is_absolute(): + raise ValueError(f"URI is not absolute: {uri!r}") + return path + + +class PosixPath(Path, PurePosixPath): + """Path subclass for non-Windows systems. + + On a POSIX system, instantiating a Path should return this object. + """ + __slots__ = () + + if os.name == 'nt': + def __new__(cls, *args, **kwargs): + raise UnsupportedOperation( + f"cannot instantiate {cls.__name__!r} on your system") + +class WindowsPath(Path, PureWindowsPath): + """Path subclass for Windows systems. + + On a Windows system, instantiating a Path should return this object. + """ + __slots__ = () + + if os.name != 'nt': + def __new__(cls, *args, **kwargs): + raise UnsupportedOperation( + f"cannot instantiate {cls.__name__!r} on your system") diff --git a/Lib/test/test_pathlib.py b/Lib/test/test_pathlib.py deleted file mode 100644 index c1696bb3737..00000000000 --- a/Lib/test/test_pathlib.py +++ /dev/null @@ -1,3328 +0,0 @@ -import contextlib -import collections.abc -import io -import os -import sys -import errno -import pathlib -import pickle -import socket -import stat -import tempfile -import unittest -from unittest import mock - -from test.support import import_helper -from test.support import set_recursion_limit -from test.support import is_emscripten, is_wasi -from test.support import os_helper -from test.support.os_helper import TESTFN, FakePath - -try: - import grp, pwd -except ImportError: - grp = pwd = None - - -# -# Tests for the pure classes. -# - -class _BasePurePathSubclass(object): - def __init__(self, *pathsegments, session_id): - super().__init__(*pathsegments) - self.session_id = session_id - - def with_segments(self, *pathsegments): - return type(self)(*pathsegments, session_id=self.session_id) - - -class _BasePurePathTest(object): - - # Keys are canonical paths, values are list of tuples of arguments - # supposed to produce equal paths. - equivalences = { - 'a/b': [ - ('a', 'b'), ('a/', 'b'), ('a', 'b/'), ('a/', 'b/'), - ('a/b/',), ('a//b',), ('a//b//',), - # Empty components get removed. - ('', 'a', 'b'), ('a', '', 'b'), ('a', 'b', ''), - ], - '/b/c/d': [ - ('a', '/b/c', 'd'), ('/a', '/b/c', 'd'), - # Empty components get removed. - ('/', 'b', '', 'c/d'), ('/', '', 'b/c/d'), ('', '/b/c/d'), - ], - } - - def setUp(self): - p = self.cls('a') - self.flavour = p._flavour - self.sep = self.flavour.sep - self.altsep = self.flavour.altsep - - def test_constructor_common(self): - P = self.cls - p = P('a') - self.assertIsInstance(p, P) - P('a', 'b', 'c') - P('/a', 'b', 'c') - P('a/b/c') - P('/a/b/c') - P(FakePath("a/b/c")) - self.assertEqual(P(P('a')), P('a')) - self.assertEqual(P(P('a'), 'b'), P('a/b')) - self.assertEqual(P(P('a'), P('b')), P('a/b')) - self.assertEqual(P(P('a'), P('b'), P('c')), P(FakePath("a/b/c"))) - self.assertEqual(P(P('./a:b')), P('./a:b')) - - def test_bytes(self): - P = self.cls - message = (r"argument should be a str or an os\.PathLike object " - r"where __fspath__ returns a str, not 'bytes'") - with self.assertRaisesRegex(TypeError, message): - P(b'a') - with self.assertRaisesRegex(TypeError, message): - P(b'a', 'b') - with self.assertRaisesRegex(TypeError, message): - P('a', b'b') - with self.assertRaises(TypeError): - P('a').joinpath(b'b') - with self.assertRaises(TypeError): - P('a') / b'b' - with self.assertRaises(TypeError): - b'a' / P('b') - with self.assertRaises(TypeError): - P('a').match(b'b') - with self.assertRaises(TypeError): - P('a').relative_to(b'b') - with self.assertRaises(TypeError): - P('a').with_name(b'b') - with self.assertRaises(TypeError): - P('a').with_stem(b'b') - with self.assertRaises(TypeError): - P('a').with_suffix(b'b') - - def _check_str_subclass(self, *args): - # Issue #21127: it should be possible to construct a PurePath object - # from a str subclass instance, and it then gets converted to - # a pure str object. - class StrSubclass(str): - pass - P = self.cls - p = P(*(StrSubclass(x) for x in args)) - self.assertEqual(p, P(*args)) - for part in p.parts: - self.assertIs(type(part), str) - - def test_str_subclass_common(self): - self._check_str_subclass('') - self._check_str_subclass('.') - self._check_str_subclass('a') - self._check_str_subclass('a/b.txt') - self._check_str_subclass('/a/b.txt') - - @unittest.skip("TODO: RUSTPYTHON; PyObject::set_slot index out of bounds") - def test_with_segments_common(self): - class P(_BasePurePathSubclass, self.cls): - pass - p = P('foo', 'bar', session_id=42) - self.assertEqual(42, (p / 'foo').session_id) - self.assertEqual(42, ('foo' / p).session_id) - self.assertEqual(42, p.joinpath('foo').session_id) - self.assertEqual(42, p.with_name('foo').session_id) - self.assertEqual(42, p.with_stem('foo').session_id) - self.assertEqual(42, p.with_suffix('.foo').session_id) - self.assertEqual(42, p.with_segments('foo').session_id) - self.assertEqual(42, p.relative_to('foo').session_id) - self.assertEqual(42, p.parent.session_id) - for parent in p.parents: - self.assertEqual(42, parent.session_id) - - def _get_drive_root_parts(self, parts): - path = self.cls(*parts) - return path.drive, path.root, path.parts - - def _check_drive_root_parts(self, arg, *expected): - sep = self.flavour.sep - actual = self._get_drive_root_parts([x.replace('/', sep) for x in arg]) - self.assertEqual(actual, expected) - if altsep := self.flavour.altsep: - actual = self._get_drive_root_parts([x.replace('/', altsep) for x in arg]) - self.assertEqual(actual, expected) - - def test_drive_root_parts_common(self): - check = self._check_drive_root_parts - sep = self.flavour.sep - # Unanchored parts. - check((), '', '', ()) - check(('a',), '', '', ('a',)) - check(('a/',), '', '', ('a',)) - check(('a', 'b'), '', '', ('a', 'b')) - # Expansion. - check(('a/b',), '', '', ('a', 'b')) - check(('a/b/',), '', '', ('a', 'b')) - check(('a', 'b/c', 'd'), '', '', ('a', 'b', 'c', 'd')) - # Collapsing and stripping excess slashes. - check(('a', 'b//c', 'd'), '', '', ('a', 'b', 'c', 'd')) - check(('a', 'b/c/', 'd'), '', '', ('a', 'b', 'c', 'd')) - # Eliminating standalone dots. - check(('.',), '', '', ()) - check(('.', '.', 'b'), '', '', ('b',)) - check(('a', '.', 'b'), '', '', ('a', 'b')) - check(('a', '.', '.'), '', '', ('a',)) - # The first part is anchored. - check(('/a/b',), '', sep, (sep, 'a', 'b')) - check(('/a', 'b'), '', sep, (sep, 'a', 'b')) - check(('/a/', 'b'), '', sep, (sep, 'a', 'b')) - # Ignoring parts before an anchored part. - check(('a', '/b', 'c'), '', sep, (sep, 'b', 'c')) - check(('a', '/b', '/c'), '', sep, (sep, 'c')) - - def test_join_common(self): - P = self.cls - p = P('a/b') - pp = p.joinpath('c') - self.assertEqual(pp, P('a/b/c')) - self.assertIs(type(pp), type(p)) - pp = p.joinpath('c', 'd') - self.assertEqual(pp, P('a/b/c/d')) - pp = p.joinpath(P('c')) - self.assertEqual(pp, P('a/b/c')) - pp = p.joinpath('/c') - self.assertEqual(pp, P('/c')) - - def test_div_common(self): - # Basically the same as joinpath(). - P = self.cls - p = P('a/b') - pp = p / 'c' - self.assertEqual(pp, P('a/b/c')) - self.assertIs(type(pp), type(p)) - pp = p / 'c/d' - self.assertEqual(pp, P('a/b/c/d')) - pp = p / 'c' / 'd' - self.assertEqual(pp, P('a/b/c/d')) - pp = 'c' / p / 'd' - self.assertEqual(pp, P('c/a/b/d')) - pp = p / P('c') - self.assertEqual(pp, P('a/b/c')) - pp = p/ '/c' - self.assertEqual(pp, P('/c')) - - def _check_str(self, expected, args): - p = self.cls(*args) - self.assertEqual(str(p), expected.replace('/', self.sep)) - - def test_str_common(self): - # Canonicalized paths roundtrip. - for pathstr in ('a', 'a/b', 'a/b/c', '/', '/a/b', '/a/b/c'): - self._check_str(pathstr, (pathstr,)) - # Special case for the empty path. - self._check_str('.', ('',)) - # Other tests for str() are in test_equivalences(). - - def test_as_posix_common(self): - P = self.cls - for pathstr in ('a', 'a/b', 'a/b/c', '/', '/a/b', '/a/b/c'): - self.assertEqual(P(pathstr).as_posix(), pathstr) - # Other tests for as_posix() are in test_equivalences(). - - def test_as_bytes_common(self): - sep = os.fsencode(self.sep) - P = self.cls - self.assertEqual(bytes(P('a/b')), b'a' + sep + b'b') - - def test_as_uri_common(self): - P = self.cls - with self.assertRaises(ValueError): - P('a').as_uri() - with self.assertRaises(ValueError): - P().as_uri() - - def test_repr_common(self): - for pathstr in ('a', 'a/b', 'a/b/c', '/', '/a/b', '/a/b/c'): - with self.subTest(pathstr=pathstr): - p = self.cls(pathstr) - clsname = p.__class__.__name__ - r = repr(p) - # The repr() is in the form ClassName("forward-slashes path"). - self.assertTrue(r.startswith(clsname + '('), r) - self.assertTrue(r.endswith(')'), r) - inner = r[len(clsname) + 1 : -1] - self.assertEqual(eval(inner), p.as_posix()) - - def test_repr_roundtrips(self): - for pathstr in ('a', 'a/b', 'a/b/c', '/', '/a/b', '/a/b/c'): - with self.subTest(pathstr=pathstr): - p = self.cls(pathstr) - r = repr(p) - # The repr() roundtrips. - q = eval(r, pathlib.__dict__) - self.assertIs(q.__class__, p.__class__) - self.assertEqual(q, p) - self.assertEqual(repr(q), r) - - def test_eq_common(self): - P = self.cls - self.assertEqual(P('a/b'), P('a/b')) - self.assertEqual(P('a/b'), P('a', 'b')) - self.assertNotEqual(P('a/b'), P('a')) - self.assertNotEqual(P('a/b'), P('/a/b')) - self.assertNotEqual(P('a/b'), P()) - self.assertNotEqual(P('/a/b'), P('/')) - self.assertNotEqual(P(), P('/')) - self.assertNotEqual(P(), "") - self.assertNotEqual(P(), {}) - self.assertNotEqual(P(), int) - - def test_match_common(self): - P = self.cls - self.assertRaises(ValueError, P('a').match, '') - self.assertRaises(ValueError, P('a').match, '.') - # Simple relative pattern. - self.assertTrue(P('b.py').match('b.py')) - self.assertTrue(P('a/b.py').match('b.py')) - self.assertTrue(P('/a/b.py').match('b.py')) - self.assertFalse(P('a.py').match('b.py')) - self.assertFalse(P('b/py').match('b.py')) - self.assertFalse(P('/a.py').match('b.py')) - self.assertFalse(P('b.py/c').match('b.py')) - # Wildcard relative pattern. - self.assertTrue(P('b.py').match('*.py')) - self.assertTrue(P('a/b.py').match('*.py')) - self.assertTrue(P('/a/b.py').match('*.py')) - self.assertFalse(P('b.pyc').match('*.py')) - self.assertFalse(P('b./py').match('*.py')) - self.assertFalse(P('b.py/c').match('*.py')) - # Multi-part relative pattern. - self.assertTrue(P('ab/c.py').match('a*/*.py')) - self.assertTrue(P('/d/ab/c.py').match('a*/*.py')) - self.assertFalse(P('a.py').match('a*/*.py')) - self.assertFalse(P('/dab/c.py').match('a*/*.py')) - self.assertFalse(P('ab/c.py/d').match('a*/*.py')) - # Absolute pattern. - self.assertTrue(P('/b.py').match('/*.py')) - self.assertFalse(P('b.py').match('/*.py')) - self.assertFalse(P('a/b.py').match('/*.py')) - self.assertFalse(P('/a/b.py').match('/*.py')) - # Multi-part absolute pattern. - self.assertTrue(P('/a/b.py').match('/a/*.py')) - self.assertFalse(P('/ab.py').match('/a/*.py')) - self.assertFalse(P('/a/b/c.py').match('/a/*.py')) - # Multi-part glob-style pattern. - self.assertFalse(P('/a/b/c.py').match('/**/*.py')) - self.assertTrue(P('/a/b/c.py').match('/a/**/*.py')) - # Case-sensitive flag - self.assertFalse(P('A.py').match('a.PY', case_sensitive=True)) - self.assertTrue(P('A.py').match('a.PY', case_sensitive=False)) - self.assertFalse(P('c:/a/B.Py').match('C:/A/*.pY', case_sensitive=True)) - self.assertTrue(P('/a/b/c.py').match('/A/*/*.Py', case_sensitive=False)) - # Matching against empty path - self.assertFalse(P().match('*')) - self.assertTrue(P().match('**')) - self.assertFalse(P().match('**/*')) - - def test_ordering_common(self): - # Ordering is tuple-alike. - def assertLess(a, b): - self.assertLess(a, b) - self.assertGreater(b, a) - P = self.cls - a = P('a') - b = P('a/b') - c = P('abc') - d = P('b') - assertLess(a, b) - assertLess(a, c) - assertLess(a, d) - assertLess(b, c) - assertLess(c, d) - P = self.cls - a = P('/a') - b = P('/a/b') - c = P('/abc') - d = P('/b') - assertLess(a, b) - assertLess(a, c) - assertLess(a, d) - assertLess(b, c) - assertLess(c, d) - with self.assertRaises(TypeError): - P() < {} - - def test_parts_common(self): - # `parts` returns a tuple. - sep = self.sep - P = self.cls - p = P('a/b') - parts = p.parts - self.assertEqual(parts, ('a', 'b')) - # When the path is absolute, the anchor is a separate part. - p = P('/a/b') - parts = p.parts - self.assertEqual(parts, (sep, 'a', 'b')) - - def test_fspath_common(self): - P = self.cls - p = P('a/b') - self._check_str(p.__fspath__(), ('a/b',)) - self._check_str(os.fspath(p), ('a/b',)) - - def test_equivalences(self): - for k, tuples in self.equivalences.items(): - canon = k.replace('/', self.sep) - posix = k.replace(self.sep, '/') - if canon != posix: - tuples = tuples + [ - tuple(part.replace('/', self.sep) for part in t) - for t in tuples - ] - tuples.append((posix, )) - pcanon = self.cls(canon) - for t in tuples: - p = self.cls(*t) - self.assertEqual(p, pcanon, "failed with args {}".format(t)) - self.assertEqual(hash(p), hash(pcanon)) - self.assertEqual(str(p), canon) - self.assertEqual(p.as_posix(), posix) - - def test_parent_common(self): - # Relative - P = self.cls - p = P('a/b/c') - self.assertEqual(p.parent, P('a/b')) - self.assertEqual(p.parent.parent, P('a')) - self.assertEqual(p.parent.parent.parent, P()) - self.assertEqual(p.parent.parent.parent.parent, P()) - # Anchored - p = P('/a/b/c') - self.assertEqual(p.parent, P('/a/b')) - self.assertEqual(p.parent.parent, P('/a')) - self.assertEqual(p.parent.parent.parent, P('/')) - self.assertEqual(p.parent.parent.parent.parent, P('/')) - - def test_parents_common(self): - # Relative - P = self.cls - p = P('a/b/c') - par = p.parents - self.assertEqual(len(par), 3) - self.assertEqual(par[0], P('a/b')) - self.assertEqual(par[1], P('a')) - self.assertEqual(par[2], P('.')) - self.assertEqual(par[-1], P('.')) - self.assertEqual(par[-2], P('a')) - self.assertEqual(par[-3], P('a/b')) - self.assertEqual(par[0:1], (P('a/b'),)) - self.assertEqual(par[:2], (P('a/b'), P('a'))) - self.assertEqual(par[:-1], (P('a/b'), P('a'))) - self.assertEqual(par[1:], (P('a'), P('.'))) - self.assertEqual(par[::2], (P('a/b'), P('.'))) - self.assertEqual(par[::-1], (P('.'), P('a'), P('a/b'))) - self.assertEqual(list(par), [P('a/b'), P('a'), P('.')]) - with self.assertRaises(IndexError): - par[-4] - with self.assertRaises(IndexError): - par[3] - with self.assertRaises(TypeError): - par[0] = p - # Anchored - p = P('/a/b/c') - par = p.parents - self.assertEqual(len(par), 3) - self.assertEqual(par[0], P('/a/b')) - self.assertEqual(par[1], P('/a')) - self.assertEqual(par[2], P('/')) - self.assertEqual(par[-1], P('/')) - self.assertEqual(par[-2], P('/a')) - self.assertEqual(par[-3], P('/a/b')) - self.assertEqual(par[0:1], (P('/a/b'),)) - self.assertEqual(par[:2], (P('/a/b'), P('/a'))) - self.assertEqual(par[:-1], (P('/a/b'), P('/a'))) - self.assertEqual(par[1:], (P('/a'), P('/'))) - self.assertEqual(par[::2], (P('/a/b'), P('/'))) - self.assertEqual(par[::-1], (P('/'), P('/a'), P('/a/b'))) - self.assertEqual(list(par), [P('/a/b'), P('/a'), P('/')]) - with self.assertRaises(IndexError): - par[-4] - with self.assertRaises(IndexError): - par[3] - - def test_drive_common(self): - P = self.cls - self.assertEqual(P('a/b').drive, '') - self.assertEqual(P('/a/b').drive, '') - self.assertEqual(P('').drive, '') - - def test_root_common(self): - P = self.cls - sep = self.sep - self.assertEqual(P('').root, '') - self.assertEqual(P('a/b').root, '') - self.assertEqual(P('/').root, sep) - self.assertEqual(P('/a/b').root, sep) - - def test_anchor_common(self): - P = self.cls - sep = self.sep - self.assertEqual(P('').anchor, '') - self.assertEqual(P('a/b').anchor, '') - self.assertEqual(P('/').anchor, sep) - self.assertEqual(P('/a/b').anchor, sep) - - def test_name_common(self): - P = self.cls - self.assertEqual(P('').name, '') - self.assertEqual(P('.').name, '') - self.assertEqual(P('/').name, '') - self.assertEqual(P('a/b').name, 'b') - self.assertEqual(P('/a/b').name, 'b') - self.assertEqual(P('/a/b/.').name, 'b') - self.assertEqual(P('a/b.py').name, 'b.py') - self.assertEqual(P('/a/b.py').name, 'b.py') - - def test_suffix_common(self): - P = self.cls - self.assertEqual(P('').suffix, '') - self.assertEqual(P('.').suffix, '') - self.assertEqual(P('..').suffix, '') - self.assertEqual(P('/').suffix, '') - self.assertEqual(P('a/b').suffix, '') - self.assertEqual(P('/a/b').suffix, '') - self.assertEqual(P('/a/b/.').suffix, '') - self.assertEqual(P('a/b.py').suffix, '.py') - self.assertEqual(P('/a/b.py').suffix, '.py') - self.assertEqual(P('a/.hgrc').suffix, '') - self.assertEqual(P('/a/.hgrc').suffix, '') - self.assertEqual(P('a/.hg.rc').suffix, '.rc') - self.assertEqual(P('/a/.hg.rc').suffix, '.rc') - self.assertEqual(P('a/b.tar.gz').suffix, '.gz') - self.assertEqual(P('/a/b.tar.gz').suffix, '.gz') - self.assertEqual(P('a/Some name. Ending with a dot.').suffix, '') - self.assertEqual(P('/a/Some name. Ending with a dot.').suffix, '') - - def test_suffixes_common(self): - P = self.cls - self.assertEqual(P('').suffixes, []) - self.assertEqual(P('.').suffixes, []) - self.assertEqual(P('/').suffixes, []) - self.assertEqual(P('a/b').suffixes, []) - self.assertEqual(P('/a/b').suffixes, []) - self.assertEqual(P('/a/b/.').suffixes, []) - self.assertEqual(P('a/b.py').suffixes, ['.py']) - self.assertEqual(P('/a/b.py').suffixes, ['.py']) - self.assertEqual(P('a/.hgrc').suffixes, []) - self.assertEqual(P('/a/.hgrc').suffixes, []) - self.assertEqual(P('a/.hg.rc').suffixes, ['.rc']) - self.assertEqual(P('/a/.hg.rc').suffixes, ['.rc']) - self.assertEqual(P('a/b.tar.gz').suffixes, ['.tar', '.gz']) - self.assertEqual(P('/a/b.tar.gz').suffixes, ['.tar', '.gz']) - self.assertEqual(P('a/Some name. Ending with a dot.').suffixes, []) - self.assertEqual(P('/a/Some name. Ending with a dot.').suffixes, []) - - def test_stem_common(self): - P = self.cls - self.assertEqual(P('').stem, '') - self.assertEqual(P('.').stem, '') - self.assertEqual(P('..').stem, '..') - self.assertEqual(P('/').stem, '') - self.assertEqual(P('a/b').stem, 'b') - self.assertEqual(P('a/b.py').stem, 'b') - self.assertEqual(P('a/.hgrc').stem, '.hgrc') - self.assertEqual(P('a/.hg.rc').stem, '.hg') - self.assertEqual(P('a/b.tar.gz').stem, 'b.tar') - self.assertEqual(P('a/Some name. Ending with a dot.').stem, - 'Some name. Ending with a dot.') - - def test_with_name_common(self): - P = self.cls - self.assertEqual(P('a/b').with_name('d.xml'), P('a/d.xml')) - self.assertEqual(P('/a/b').with_name('d.xml'), P('/a/d.xml')) - self.assertEqual(P('a/b.py').with_name('d.xml'), P('a/d.xml')) - self.assertEqual(P('/a/b.py').with_name('d.xml'), P('/a/d.xml')) - self.assertEqual(P('a/Dot ending.').with_name('d.xml'), P('a/d.xml')) - self.assertEqual(P('/a/Dot ending.').with_name('d.xml'), P('/a/d.xml')) - self.assertRaises(ValueError, P('').with_name, 'd.xml') - self.assertRaises(ValueError, P('.').with_name, 'd.xml') - self.assertRaises(ValueError, P('/').with_name, 'd.xml') - self.assertRaises(ValueError, P('a/b').with_name, '') - self.assertRaises(ValueError, P('a/b').with_name, '.') - self.assertRaises(ValueError, P('a/b').with_name, '/c') - self.assertRaises(ValueError, P('a/b').with_name, 'c/') - self.assertRaises(ValueError, P('a/b').with_name, 'c/d') - - def test_with_stem_common(self): - P = self.cls - self.assertEqual(P('a/b').with_stem('d'), P('a/d')) - self.assertEqual(P('/a/b').with_stem('d'), P('/a/d')) - self.assertEqual(P('a/b.py').with_stem('d'), P('a/d.py')) - self.assertEqual(P('/a/b.py').with_stem('d'), P('/a/d.py')) - self.assertEqual(P('/a/b.tar.gz').with_stem('d'), P('/a/d.gz')) - self.assertEqual(P('a/Dot ending.').with_stem('d'), P('a/d')) - self.assertEqual(P('/a/Dot ending.').with_stem('d'), P('/a/d')) - self.assertRaises(ValueError, P('').with_stem, 'd') - self.assertRaises(ValueError, P('.').with_stem, 'd') - self.assertRaises(ValueError, P('/').with_stem, 'd') - self.assertRaises(ValueError, P('a/b').with_stem, '') - self.assertRaises(ValueError, P('a/b').with_stem, '.') - self.assertRaises(ValueError, P('a/b').with_stem, '/c') - self.assertRaises(ValueError, P('a/b').with_stem, 'c/') - self.assertRaises(ValueError, P('a/b').with_stem, 'c/d') - - def test_with_suffix_common(self): - P = self.cls - self.assertEqual(P('a/b').with_suffix('.gz'), P('a/b.gz')) - self.assertEqual(P('/a/b').with_suffix('.gz'), P('/a/b.gz')) - self.assertEqual(P('a/b.py').with_suffix('.gz'), P('a/b.gz')) - self.assertEqual(P('/a/b.py').with_suffix('.gz'), P('/a/b.gz')) - # Stripping suffix. - self.assertEqual(P('a/b.py').with_suffix(''), P('a/b')) - self.assertEqual(P('/a/b').with_suffix(''), P('/a/b')) - # Path doesn't have a "filename" component. - self.assertRaises(ValueError, P('').with_suffix, '.gz') - self.assertRaises(ValueError, P('.').with_suffix, '.gz') - self.assertRaises(ValueError, P('/').with_suffix, '.gz') - # Invalid suffix. - self.assertRaises(ValueError, P('a/b').with_suffix, 'gz') - self.assertRaises(ValueError, P('a/b').with_suffix, '/') - self.assertRaises(ValueError, P('a/b').with_suffix, '.') - self.assertRaises(ValueError, P('a/b').with_suffix, '/.gz') - self.assertRaises(ValueError, P('a/b').with_suffix, 'c/d') - self.assertRaises(ValueError, P('a/b').with_suffix, '.c/.d') - self.assertRaises(ValueError, P('a/b').with_suffix, './.d') - self.assertRaises(ValueError, P('a/b').with_suffix, '.d/.') - self.assertRaises(ValueError, P('a/b').with_suffix, - (self.flavour.sep, 'd')) - - def test_relative_to_common(self): - P = self.cls - p = P('a/b') - self.assertRaises(TypeError, p.relative_to) - self.assertRaises(TypeError, p.relative_to, b'a') - self.assertEqual(p.relative_to(P()), P('a/b')) - self.assertEqual(p.relative_to(''), P('a/b')) - self.assertEqual(p.relative_to(P('a')), P('b')) - self.assertEqual(p.relative_to('a'), P('b')) - self.assertEqual(p.relative_to('a/'), P('b')) - self.assertEqual(p.relative_to(P('a/b')), P()) - self.assertEqual(p.relative_to('a/b'), P()) - self.assertEqual(p.relative_to(P(), walk_up=True), P('a/b')) - self.assertEqual(p.relative_to('', walk_up=True), P('a/b')) - self.assertEqual(p.relative_to(P('a'), walk_up=True), P('b')) - self.assertEqual(p.relative_to('a', walk_up=True), P('b')) - self.assertEqual(p.relative_to('a/', walk_up=True), P('b')) - self.assertEqual(p.relative_to(P('a/b'), walk_up=True), P()) - self.assertEqual(p.relative_to('a/b', walk_up=True), P()) - self.assertEqual(p.relative_to(P('a/c'), walk_up=True), P('../b')) - self.assertEqual(p.relative_to('a/c', walk_up=True), P('../b')) - self.assertEqual(p.relative_to(P('a/b/c'), walk_up=True), P('..')) - self.assertEqual(p.relative_to('a/b/c', walk_up=True), P('..')) - self.assertEqual(p.relative_to(P('c'), walk_up=True), P('../a/b')) - self.assertEqual(p.relative_to('c', walk_up=True), P('../a/b')) - # With several args. - with self.assertWarns(DeprecationWarning): - p.relative_to('a', 'b') - p.relative_to('a', 'b', walk_up=True) - # Unrelated paths. - self.assertRaises(ValueError, p.relative_to, P('c')) - self.assertRaises(ValueError, p.relative_to, P('a/b/c')) - self.assertRaises(ValueError, p.relative_to, P('a/c')) - self.assertRaises(ValueError, p.relative_to, P('/a')) - self.assertRaises(ValueError, p.relative_to, P("../a")) - self.assertRaises(ValueError, p.relative_to, P("a/..")) - self.assertRaises(ValueError, p.relative_to, P("/a/..")) - self.assertRaises(ValueError, p.relative_to, P('/'), walk_up=True) - self.assertRaises(ValueError, p.relative_to, P('/a'), walk_up=True) - self.assertRaises(ValueError, p.relative_to, P("../a"), walk_up=True) - self.assertRaises(ValueError, p.relative_to, P("a/.."), walk_up=True) - self.assertRaises(ValueError, p.relative_to, P("/a/.."), walk_up=True) - p = P('/a/b') - self.assertEqual(p.relative_to(P('/')), P('a/b')) - self.assertEqual(p.relative_to('/'), P('a/b')) - self.assertEqual(p.relative_to(P('/a')), P('b')) - self.assertEqual(p.relative_to('/a'), P('b')) - self.assertEqual(p.relative_to('/a/'), P('b')) - self.assertEqual(p.relative_to(P('/a/b')), P()) - self.assertEqual(p.relative_to('/a/b'), P()) - self.assertEqual(p.relative_to(P('/'), walk_up=True), P('a/b')) - self.assertEqual(p.relative_to('/', walk_up=True), P('a/b')) - self.assertEqual(p.relative_to(P('/a'), walk_up=True), P('b')) - self.assertEqual(p.relative_to('/a', walk_up=True), P('b')) - self.assertEqual(p.relative_to('/a/', walk_up=True), P('b')) - self.assertEqual(p.relative_to(P('/a/b'), walk_up=True), P()) - self.assertEqual(p.relative_to('/a/b', walk_up=True), P()) - self.assertEqual(p.relative_to(P('/a/c'), walk_up=True), P('../b')) - self.assertEqual(p.relative_to('/a/c', walk_up=True), P('../b')) - self.assertEqual(p.relative_to(P('/a/b/c'), walk_up=True), P('..')) - self.assertEqual(p.relative_to('/a/b/c', walk_up=True), P('..')) - self.assertEqual(p.relative_to(P('/c'), walk_up=True), P('../a/b')) - self.assertEqual(p.relative_to('/c', walk_up=True), P('../a/b')) - # Unrelated paths. - self.assertRaises(ValueError, p.relative_to, P('/c')) - self.assertRaises(ValueError, p.relative_to, P('/a/b/c')) - self.assertRaises(ValueError, p.relative_to, P('/a/c')) - self.assertRaises(ValueError, p.relative_to, P()) - self.assertRaises(ValueError, p.relative_to, '') - self.assertRaises(ValueError, p.relative_to, P('a')) - self.assertRaises(ValueError, p.relative_to, P("../a")) - self.assertRaises(ValueError, p.relative_to, P("a/..")) - self.assertRaises(ValueError, p.relative_to, P("/a/..")) - self.assertRaises(ValueError, p.relative_to, P(''), walk_up=True) - self.assertRaises(ValueError, p.relative_to, P('a'), walk_up=True) - self.assertRaises(ValueError, p.relative_to, P("../a"), walk_up=True) - self.assertRaises(ValueError, p.relative_to, P("a/.."), walk_up=True) - self.assertRaises(ValueError, p.relative_to, P("/a/.."), walk_up=True) - - def test_is_relative_to_common(self): - P = self.cls - p = P('a/b') - self.assertRaises(TypeError, p.is_relative_to) - self.assertRaises(TypeError, p.is_relative_to, b'a') - self.assertTrue(p.is_relative_to(P())) - self.assertTrue(p.is_relative_to('')) - self.assertTrue(p.is_relative_to(P('a'))) - self.assertTrue(p.is_relative_to('a/')) - self.assertTrue(p.is_relative_to(P('a/b'))) - self.assertTrue(p.is_relative_to('a/b')) - # With several args. - with self.assertWarns(DeprecationWarning): - p.is_relative_to('a', 'b') - # Unrelated paths. - self.assertFalse(p.is_relative_to(P('c'))) - self.assertFalse(p.is_relative_to(P('a/b/c'))) - self.assertFalse(p.is_relative_to(P('a/c'))) - self.assertFalse(p.is_relative_to(P('/a'))) - p = P('/a/b') - self.assertTrue(p.is_relative_to(P('/'))) - self.assertTrue(p.is_relative_to('/')) - self.assertTrue(p.is_relative_to(P('/a'))) - self.assertTrue(p.is_relative_to('/a')) - self.assertTrue(p.is_relative_to('/a/')) - self.assertTrue(p.is_relative_to(P('/a/b'))) - self.assertTrue(p.is_relative_to('/a/b')) - # Unrelated paths. - self.assertFalse(p.is_relative_to(P('/c'))) - self.assertFalse(p.is_relative_to(P('/a/b/c'))) - self.assertFalse(p.is_relative_to(P('/a/c'))) - self.assertFalse(p.is_relative_to(P())) - self.assertFalse(p.is_relative_to('')) - self.assertFalse(p.is_relative_to(P('a'))) - - def test_pickling_common(self): - P = self.cls - p = P('/a/b') - for proto in range(0, pickle.HIGHEST_PROTOCOL + 1): - dumped = pickle.dumps(p, proto) - pp = pickle.loads(dumped) - self.assertIs(pp.__class__, p.__class__) - self.assertEqual(pp, p) - self.assertEqual(hash(pp), hash(p)) - self.assertEqual(str(pp), str(p)) - - -class PurePosixPathTest(_BasePurePathTest, unittest.TestCase): - cls = pathlib.PurePosixPath - - def test_drive_root_parts(self): - check = self._check_drive_root_parts - # Collapsing of excess leading slashes, except for the double-slash - # special case. - check(('//a', 'b'), '', '//', ('//', 'a', 'b')) - check(('///a', 'b'), '', '/', ('/', 'a', 'b')) - check(('////a', 'b'), '', '/', ('/', 'a', 'b')) - # Paths which look like NT paths aren't treated specially. - check(('c:a',), '', '', ('c:a',)) - check(('c:\\a',), '', '', ('c:\\a',)) - check(('\\a',), '', '', ('\\a',)) - - def test_root(self): - P = self.cls - self.assertEqual(P('/a/b').root, '/') - self.assertEqual(P('///a/b').root, '/') - # POSIX special case for two leading slashes. - self.assertEqual(P('//a/b').root, '//') - - def test_eq(self): - P = self.cls - self.assertNotEqual(P('a/b'), P('A/b')) - self.assertEqual(P('/a'), P('///a')) - self.assertNotEqual(P('/a'), P('//a')) - - def test_as_uri(self): - P = self.cls - self.assertEqual(P('/').as_uri(), 'file:///') - self.assertEqual(P('/a/b.c').as_uri(), 'file:///a/b.c') - self.assertEqual(P('/a/b%#c').as_uri(), 'file:///a/b%25%23c') - - def test_as_uri_non_ascii(self): - from urllib.parse import quote_from_bytes - P = self.cls - try: - os.fsencode('\xe9') - except UnicodeEncodeError: - self.skipTest("\\xe9 cannot be encoded to the filesystem encoding") - self.assertEqual(P('/a/b\xe9').as_uri(), - 'file:///a/b' + quote_from_bytes(os.fsencode('\xe9'))) - - def test_match(self): - P = self.cls - self.assertFalse(P('A.py').match('a.PY')) - - def test_is_absolute(self): - P = self.cls - self.assertFalse(P().is_absolute()) - self.assertFalse(P('a').is_absolute()) - self.assertFalse(P('a/b/').is_absolute()) - self.assertTrue(P('/').is_absolute()) - self.assertTrue(P('/a').is_absolute()) - self.assertTrue(P('/a/b/').is_absolute()) - self.assertTrue(P('//a').is_absolute()) - self.assertTrue(P('//a/b').is_absolute()) - - def test_is_reserved(self): - P = self.cls - self.assertIs(False, P('').is_reserved()) - self.assertIs(False, P('/').is_reserved()) - self.assertIs(False, P('/foo/bar').is_reserved()) - self.assertIs(False, P('/dev/con/PRN/NUL').is_reserved()) - - def test_join(self): - P = self.cls - p = P('//a') - pp = p.joinpath('b') - self.assertEqual(pp, P('//a/b')) - pp = P('/a').joinpath('//c') - self.assertEqual(pp, P('//c')) - pp = P('//a').joinpath('/c') - self.assertEqual(pp, P('/c')) - - def test_div(self): - # Basically the same as joinpath(). - P = self.cls - p = P('//a') - pp = p / 'b' - self.assertEqual(pp, P('//a/b')) - pp = P('/a') / '//c' - self.assertEqual(pp, P('//c')) - pp = P('//a') / '/c' - self.assertEqual(pp, P('/c')) - - def test_parse_windows_path(self): - P = self.cls - p = P('c:', 'a', 'b') - pp = P(pathlib.PureWindowsPath('c:\\a\\b')) - self.assertEqual(p, pp) - - -class PureWindowsPathTest(_BasePurePathTest, unittest.TestCase): - cls = pathlib.PureWindowsPath - - equivalences = _BasePurePathTest.equivalences.copy() - equivalences.update({ - './a:b': [ ('./a:b',) ], - 'c:a': [ ('c:', 'a'), ('c:', 'a/'), ('.', 'c:', 'a') ], - 'c:/a': [ - ('c:/', 'a'), ('c:', '/', 'a'), ('c:', '/a'), - ('/z', 'c:/', 'a'), ('//x/y', 'c:/', 'a'), - ], - '//a/b/': [ ('//a/b',) ], - '//a/b/c': [ - ('//a/b', 'c'), ('//a/b/', 'c'), - ], - }) - - def test_drive_root_parts(self): - check = self._check_drive_root_parts - # First part is anchored. - check(('c:',), 'c:', '', ('c:',)) - check(('c:/',), 'c:', '\\', ('c:\\',)) - check(('/',), '', '\\', ('\\',)) - check(('c:a',), 'c:', '', ('c:', 'a')) - check(('c:/a',), 'c:', '\\', ('c:\\', 'a')) - check(('/a',), '', '\\', ('\\', 'a')) - # UNC paths. - check(('//',), '\\\\', '', ('\\\\',)) - check(('//a',), '\\\\a', '', ('\\\\a',)) - check(('//a/',), '\\\\a\\', '', ('\\\\a\\',)) - check(('//a/b',), '\\\\a\\b', '\\', ('\\\\a\\b\\',)) - check(('//a/b/',), '\\\\a\\b', '\\', ('\\\\a\\b\\',)) - check(('//a/b/c',), '\\\\a\\b', '\\', ('\\\\a\\b\\', 'c')) - # Second part is anchored, so that the first part is ignored. - check(('a', 'Z:b', 'c'), 'Z:', '', ('Z:', 'b', 'c')) - check(('a', 'Z:/b', 'c'), 'Z:', '\\', ('Z:\\', 'b', 'c')) - # UNC paths. - check(('a', '//b/c', 'd'), '\\\\b\\c', '\\', ('\\\\b\\c\\', 'd')) - # Collapsing and stripping excess slashes. - check(('a', 'Z://b//c/', 'd/'), 'Z:', '\\', ('Z:\\', 'b', 'c', 'd')) - # UNC paths. - check(('a', '//b/c//', 'd'), '\\\\b\\c', '\\', ('\\\\b\\c\\', 'd')) - # Extended paths. - check(('//./c:',), '\\\\.\\c:', '', ('\\\\.\\c:',)) - check(('//?/c:/',), '\\\\?\\c:', '\\', ('\\\\?\\c:\\',)) - check(('//?/c:/a',), '\\\\?\\c:', '\\', ('\\\\?\\c:\\', 'a')) - check(('//?/c:/a', '/b'), '\\\\?\\c:', '\\', ('\\\\?\\c:\\', 'b')) - # Extended UNC paths (format is "\\?\UNC\server\share"). - check(('//?',), '\\\\?', '', ('\\\\?',)) - check(('//?/',), '\\\\?\\', '', ('\\\\?\\',)) - check(('//?/UNC',), '\\\\?\\UNC', '', ('\\\\?\\UNC',)) - check(('//?/UNC/',), '\\\\?\\UNC\\', '', ('\\\\?\\UNC\\',)) - check(('//?/UNC/b',), '\\\\?\\UNC\\b', '', ('\\\\?\\UNC\\b',)) - check(('//?/UNC/b/',), '\\\\?\\UNC\\b\\', '', ('\\\\?\\UNC\\b\\',)) - check(('//?/UNC/b/c',), '\\\\?\\UNC\\b\\c', '\\', ('\\\\?\\UNC\\b\\c\\',)) - check(('//?/UNC/b/c/',), '\\\\?\\UNC\\b\\c', '\\', ('\\\\?\\UNC\\b\\c\\',)) - check(('//?/UNC/b/c/d',), '\\\\?\\UNC\\b\\c', '\\', ('\\\\?\\UNC\\b\\c\\', 'd')) - # UNC device paths - check(('//./BootPartition/',), '\\\\.\\BootPartition', '\\', ('\\\\.\\BootPartition\\',)) - check(('//?/BootPartition/',), '\\\\?\\BootPartition', '\\', ('\\\\?\\BootPartition\\',)) - check(('//./PhysicalDrive0',), '\\\\.\\PhysicalDrive0', '', ('\\\\.\\PhysicalDrive0',)) - check(('//?/Volume{}/',), '\\\\?\\Volume{}', '\\', ('\\\\?\\Volume{}\\',)) - check(('//./nul',), '\\\\.\\nul', '', ('\\\\.\\nul',)) - # Second part has a root but not drive. - check(('a', '/b', 'c'), '', '\\', ('\\', 'b', 'c')) - check(('Z:/a', '/b', 'c'), 'Z:', '\\', ('Z:\\', 'b', 'c')) - check(('//?/Z:/a', '/b', 'c'), '\\\\?\\Z:', '\\', ('\\\\?\\Z:\\', 'b', 'c')) - # Joining with the same drive => the first path is appended to if - # the second path is relative. - check(('c:/a/b', 'c:x/y'), 'c:', '\\', ('c:\\', 'a', 'b', 'x', 'y')) - check(('c:/a/b', 'c:/x/y'), 'c:', '\\', ('c:\\', 'x', 'y')) - # Paths to files with NTFS alternate data streams - check(('./c:s',), '', '', ('c:s',)) - check(('cc:s',), '', '', ('cc:s',)) - check(('C:c:s',), 'C:', '', ('C:', 'c:s')) - check(('C:/c:s',), 'C:', '\\', ('C:\\', 'c:s')) - check(('D:a', './c:b'), 'D:', '', ('D:', 'a', 'c:b')) - check(('D:/a', './c:b'), 'D:', '\\', ('D:\\', 'a', 'c:b')) - - def test_str(self): - p = self.cls('a/b/c') - self.assertEqual(str(p), 'a\\b\\c') - p = self.cls('c:/a/b/c') - self.assertEqual(str(p), 'c:\\a\\b\\c') - p = self.cls('//a/b') - self.assertEqual(str(p), '\\\\a\\b\\') - p = self.cls('//a/b/c') - self.assertEqual(str(p), '\\\\a\\b\\c') - p = self.cls('//a/b/c/d') - self.assertEqual(str(p), '\\\\a\\b\\c\\d') - - def test_str_subclass(self): - self._check_str_subclass('.\\a:b') - self._check_str_subclass('c:') - self._check_str_subclass('c:a') - self._check_str_subclass('c:a\\b.txt') - self._check_str_subclass('c:\\') - self._check_str_subclass('c:\\a') - self._check_str_subclass('c:\\a\\b.txt') - self._check_str_subclass('\\\\some\\share') - self._check_str_subclass('\\\\some\\share\\a') - self._check_str_subclass('\\\\some\\share\\a\\b.txt') - - def test_eq(self): - P = self.cls - self.assertEqual(P('c:a/b'), P('c:a/b')) - self.assertEqual(P('c:a/b'), P('c:', 'a', 'b')) - self.assertNotEqual(P('c:a/b'), P('d:a/b')) - self.assertNotEqual(P('c:a/b'), P('c:/a/b')) - self.assertNotEqual(P('/a/b'), P('c:/a/b')) - # Case-insensitivity. - self.assertEqual(P('a/B'), P('A/b')) - self.assertEqual(P('C:a/B'), P('c:A/b')) - self.assertEqual(P('//Some/SHARE/a/B'), P('//somE/share/A/b')) - self.assertEqual(P('\u0130'), P('i\u0307')) - - def test_as_uri(self): - P = self.cls - with self.assertRaises(ValueError): - P('/a/b').as_uri() - with self.assertRaises(ValueError): - P('c:a/b').as_uri() - self.assertEqual(P('c:/').as_uri(), 'file:///c:/') - self.assertEqual(P('c:/a/b.c').as_uri(), 'file:///c:/a/b.c') - self.assertEqual(P('c:/a/b%#c').as_uri(), 'file:///c:/a/b%25%23c') - self.assertEqual(P('c:/a/b\xe9').as_uri(), 'file:///c:/a/b%C3%A9') - self.assertEqual(P('//some/share/').as_uri(), 'file://some/share/') - self.assertEqual(P('//some/share/a/b.c').as_uri(), - 'file://some/share/a/b.c') - self.assertEqual(P('//some/share/a/b%#c\xe9').as_uri(), - 'file://some/share/a/b%25%23c%C3%A9') - - def test_match(self): - P = self.cls - # Absolute patterns. - self.assertTrue(P('c:/b.py').match('*:/*.py')) - self.assertTrue(P('c:/b.py').match('c:/*.py')) - self.assertFalse(P('d:/b.py').match('c:/*.py')) # wrong drive - self.assertFalse(P('b.py').match('/*.py')) - self.assertFalse(P('b.py').match('c:*.py')) - self.assertFalse(P('b.py').match('c:/*.py')) - self.assertFalse(P('c:b.py').match('/*.py')) - self.assertFalse(P('c:b.py').match('c:/*.py')) - self.assertFalse(P('/b.py').match('c:*.py')) - self.assertFalse(P('/b.py').match('c:/*.py')) - # UNC patterns. - self.assertTrue(P('//some/share/a.py').match('//*/*/*.py')) - self.assertTrue(P('//some/share/a.py').match('//some/share/*.py')) - self.assertFalse(P('//other/share/a.py').match('//some/share/*.py')) - self.assertFalse(P('//some/share/a/b.py').match('//some/share/*.py')) - # Case-insensitivity. - self.assertTrue(P('B.py').match('b.PY')) - self.assertTrue(P('c:/a/B.Py').match('C:/A/*.pY')) - self.assertTrue(P('//Some/Share/B.Py').match('//somE/sharE/*.pY')) - # Path anchor doesn't match pattern anchor - self.assertFalse(P('c:/b.py').match('/*.py')) # 'c:/' vs '/' - self.assertFalse(P('c:/b.py').match('c:*.py')) # 'c:/' vs 'c:' - self.assertFalse(P('//some/share/a.py').match('/*.py')) # '//some/share/' vs '/' - - def test_ordering_common(self): - # Case-insensitivity. - def assertOrderedEqual(a, b): - self.assertLessEqual(a, b) - self.assertGreaterEqual(b, a) - P = self.cls - p = P('c:A/b') - q = P('C:a/B') - assertOrderedEqual(p, q) - self.assertFalse(p < q) - self.assertFalse(p > q) - p = P('//some/Share/A/b') - q = P('//Some/SHARE/a/B') - assertOrderedEqual(p, q) - self.assertFalse(p < q) - self.assertFalse(p > q) - - def test_parts(self): - P = self.cls - p = P('c:a/b') - parts = p.parts - self.assertEqual(parts, ('c:', 'a', 'b')) - p = P('c:/a/b') - parts = p.parts - self.assertEqual(parts, ('c:\\', 'a', 'b')) - p = P('//a/b/c/d') - parts = p.parts - self.assertEqual(parts, ('\\\\a\\b\\', 'c', 'd')) - - def test_parent(self): - # Anchored - P = self.cls - p = P('z:a/b/c') - self.assertEqual(p.parent, P('z:a/b')) - self.assertEqual(p.parent.parent, P('z:a')) - self.assertEqual(p.parent.parent.parent, P('z:')) - self.assertEqual(p.parent.parent.parent.parent, P('z:')) - p = P('z:/a/b/c') - self.assertEqual(p.parent, P('z:/a/b')) - self.assertEqual(p.parent.parent, P('z:/a')) - self.assertEqual(p.parent.parent.parent, P('z:/')) - self.assertEqual(p.parent.parent.parent.parent, P('z:/')) - p = P('//a/b/c/d') - self.assertEqual(p.parent, P('//a/b/c')) - self.assertEqual(p.parent.parent, P('//a/b')) - self.assertEqual(p.parent.parent.parent, P('//a/b')) - - def test_parents(self): - # Anchored - P = self.cls - p = P('z:a/b/') - par = p.parents - self.assertEqual(len(par), 2) - self.assertEqual(par[0], P('z:a')) - self.assertEqual(par[1], P('z:')) - self.assertEqual(par[0:1], (P('z:a'),)) - self.assertEqual(par[:-1], (P('z:a'),)) - self.assertEqual(par[:2], (P('z:a'), P('z:'))) - self.assertEqual(par[1:], (P('z:'),)) - self.assertEqual(par[::2], (P('z:a'),)) - self.assertEqual(par[::-1], (P('z:'), P('z:a'))) - self.assertEqual(list(par), [P('z:a'), P('z:')]) - with self.assertRaises(IndexError): - par[2] - p = P('z:/a/b/') - par = p.parents - self.assertEqual(len(par), 2) - self.assertEqual(par[0], P('z:/a')) - self.assertEqual(par[1], P('z:/')) - self.assertEqual(par[0:1], (P('z:/a'),)) - self.assertEqual(par[0:-1], (P('z:/a'),)) - self.assertEqual(par[:2], (P('z:/a'), P('z:/'))) - self.assertEqual(par[1:], (P('z:/'),)) - self.assertEqual(par[::2], (P('z:/a'),)) - self.assertEqual(par[::-1], (P('z:/'), P('z:/a'),)) - self.assertEqual(list(par), [P('z:/a'), P('z:/')]) - with self.assertRaises(IndexError): - par[2] - p = P('//a/b/c/d') - par = p.parents - self.assertEqual(len(par), 2) - self.assertEqual(par[0], P('//a/b/c')) - self.assertEqual(par[1], P('//a/b')) - self.assertEqual(par[0:1], (P('//a/b/c'),)) - self.assertEqual(par[0:-1], (P('//a/b/c'),)) - self.assertEqual(par[:2], (P('//a/b/c'), P('//a/b'))) - self.assertEqual(par[1:], (P('//a/b'),)) - self.assertEqual(par[::2], (P('//a/b/c'),)) - self.assertEqual(par[::-1], (P('//a/b'), P('//a/b/c'))) - self.assertEqual(list(par), [P('//a/b/c'), P('//a/b')]) - with self.assertRaises(IndexError): - par[2] - - def test_drive(self): - P = self.cls - self.assertEqual(P('c:').drive, 'c:') - self.assertEqual(P('c:a/b').drive, 'c:') - self.assertEqual(P('c:/').drive, 'c:') - self.assertEqual(P('c:/a/b/').drive, 'c:') - self.assertEqual(P('//a/b').drive, '\\\\a\\b') - self.assertEqual(P('//a/b/').drive, '\\\\a\\b') - self.assertEqual(P('//a/b/c/d').drive, '\\\\a\\b') - self.assertEqual(P('./c:a').drive, '') - - def test_root(self): - P = self.cls - self.assertEqual(P('c:').root, '') - self.assertEqual(P('c:a/b').root, '') - self.assertEqual(P('c:/').root, '\\') - self.assertEqual(P('c:/a/b/').root, '\\') - self.assertEqual(P('//a/b').root, '\\') - self.assertEqual(P('//a/b/').root, '\\') - self.assertEqual(P('//a/b/c/d').root, '\\') - - def test_anchor(self): - P = self.cls - self.assertEqual(P('c:').anchor, 'c:') - self.assertEqual(P('c:a/b').anchor, 'c:') - self.assertEqual(P('c:/').anchor, 'c:\\') - self.assertEqual(P('c:/a/b/').anchor, 'c:\\') - self.assertEqual(P('//a/b').anchor, '\\\\a\\b\\') - self.assertEqual(P('//a/b/').anchor, '\\\\a\\b\\') - self.assertEqual(P('//a/b/c/d').anchor, '\\\\a\\b\\') - - def test_name(self): - P = self.cls - self.assertEqual(P('c:').name, '') - self.assertEqual(P('c:/').name, '') - self.assertEqual(P('c:a/b').name, 'b') - self.assertEqual(P('c:/a/b').name, 'b') - self.assertEqual(P('c:a/b.py').name, 'b.py') - self.assertEqual(P('c:/a/b.py').name, 'b.py') - self.assertEqual(P('//My.py/Share.php').name, '') - self.assertEqual(P('//My.py/Share.php/a/b').name, 'b') - - def test_suffix(self): - P = self.cls - self.assertEqual(P('c:').suffix, '') - self.assertEqual(P('c:/').suffix, '') - self.assertEqual(P('c:a/b').suffix, '') - self.assertEqual(P('c:/a/b').suffix, '') - self.assertEqual(P('c:a/b.py').suffix, '.py') - self.assertEqual(P('c:/a/b.py').suffix, '.py') - self.assertEqual(P('c:a/.hgrc').suffix, '') - self.assertEqual(P('c:/a/.hgrc').suffix, '') - self.assertEqual(P('c:a/.hg.rc').suffix, '.rc') - self.assertEqual(P('c:/a/.hg.rc').suffix, '.rc') - self.assertEqual(P('c:a/b.tar.gz').suffix, '.gz') - self.assertEqual(P('c:/a/b.tar.gz').suffix, '.gz') - self.assertEqual(P('c:a/Some name. Ending with a dot.').suffix, '') - self.assertEqual(P('c:/a/Some name. Ending with a dot.').suffix, '') - self.assertEqual(P('//My.py/Share.php').suffix, '') - self.assertEqual(P('//My.py/Share.php/a/b').suffix, '') - - def test_suffixes(self): - P = self.cls - self.assertEqual(P('c:').suffixes, []) - self.assertEqual(P('c:/').suffixes, []) - self.assertEqual(P('c:a/b').suffixes, []) - self.assertEqual(P('c:/a/b').suffixes, []) - self.assertEqual(P('c:a/b.py').suffixes, ['.py']) - self.assertEqual(P('c:/a/b.py').suffixes, ['.py']) - self.assertEqual(P('c:a/.hgrc').suffixes, []) - self.assertEqual(P('c:/a/.hgrc').suffixes, []) - self.assertEqual(P('c:a/.hg.rc').suffixes, ['.rc']) - self.assertEqual(P('c:/a/.hg.rc').suffixes, ['.rc']) - self.assertEqual(P('c:a/b.tar.gz').suffixes, ['.tar', '.gz']) - self.assertEqual(P('c:/a/b.tar.gz').suffixes, ['.tar', '.gz']) - self.assertEqual(P('//My.py/Share.php').suffixes, []) - self.assertEqual(P('//My.py/Share.php/a/b').suffixes, []) - self.assertEqual(P('c:a/Some name. Ending with a dot.').suffixes, []) - self.assertEqual(P('c:/a/Some name. Ending with a dot.').suffixes, []) - - def test_stem(self): - P = self.cls - self.assertEqual(P('c:').stem, '') - self.assertEqual(P('c:.').stem, '') - self.assertEqual(P('c:..').stem, '..') - self.assertEqual(P('c:/').stem, '') - self.assertEqual(P('c:a/b').stem, 'b') - self.assertEqual(P('c:a/b.py').stem, 'b') - self.assertEqual(P('c:a/.hgrc').stem, '.hgrc') - self.assertEqual(P('c:a/.hg.rc').stem, '.hg') - self.assertEqual(P('c:a/b.tar.gz').stem, 'b.tar') - self.assertEqual(P('c:a/Some name. Ending with a dot.').stem, - 'Some name. Ending with a dot.') - - def test_with_name(self): - P = self.cls - self.assertEqual(P('c:a/b').with_name('d.xml'), P('c:a/d.xml')) - self.assertEqual(P('c:/a/b').with_name('d.xml'), P('c:/a/d.xml')) - self.assertEqual(P('c:a/Dot ending.').with_name('d.xml'), P('c:a/d.xml')) - self.assertEqual(P('c:/a/Dot ending.').with_name('d.xml'), P('c:/a/d.xml')) - self.assertRaises(ValueError, P('c:').with_name, 'd.xml') - self.assertRaises(ValueError, P('c:/').with_name, 'd.xml') - self.assertRaises(ValueError, P('//My/Share').with_name, 'd.xml') - self.assertEqual(str(P('a').with_name('d:')), '.\\d:') - self.assertEqual(str(P('a').with_name('d:e')), '.\\d:e') - self.assertEqual(P('c:a/b').with_name('d:'), P('c:a/d:')) - self.assertEqual(P('c:a/b').with_name('d:e'), P('c:a/d:e')) - self.assertRaises(ValueError, P('c:a/b').with_name, 'd:/e') - self.assertRaises(ValueError, P('c:a/b').with_name, '//My/Share') - - def test_with_stem(self): - P = self.cls - self.assertEqual(P('c:a/b').with_stem('d'), P('c:a/d')) - self.assertEqual(P('c:/a/b').with_stem('d'), P('c:/a/d')) - self.assertEqual(P('c:a/Dot ending.').with_stem('d'), P('c:a/d')) - self.assertEqual(P('c:/a/Dot ending.').with_stem('d'), P('c:/a/d')) - self.assertRaises(ValueError, P('c:').with_stem, 'd') - self.assertRaises(ValueError, P('c:/').with_stem, 'd') - self.assertRaises(ValueError, P('//My/Share').with_stem, 'd') - self.assertEqual(str(P('a').with_stem('d:')), '.\\d:') - self.assertEqual(str(P('a').with_stem('d:e')), '.\\d:e') - self.assertEqual(P('c:a/b').with_stem('d:'), P('c:a/d:')) - self.assertEqual(P('c:a/b').with_stem('d:e'), P('c:a/d:e')) - self.assertRaises(ValueError, P('c:a/b').with_stem, 'd:/e') - self.assertRaises(ValueError, P('c:a/b').with_stem, '//My/Share') - - def test_with_suffix(self): - P = self.cls - self.assertEqual(P('c:a/b').with_suffix('.gz'), P('c:a/b.gz')) - self.assertEqual(P('c:/a/b').with_suffix('.gz'), P('c:/a/b.gz')) - self.assertEqual(P('c:a/b.py').with_suffix('.gz'), P('c:a/b.gz')) - self.assertEqual(P('c:/a/b.py').with_suffix('.gz'), P('c:/a/b.gz')) - # Path doesn't have a "filename" component. - self.assertRaises(ValueError, P('').with_suffix, '.gz') - self.assertRaises(ValueError, P('.').with_suffix, '.gz') - self.assertRaises(ValueError, P('/').with_suffix, '.gz') - self.assertRaises(ValueError, P('//My/Share').with_suffix, '.gz') - # Invalid suffix. - self.assertRaises(ValueError, P('c:a/b').with_suffix, 'gz') - self.assertRaises(ValueError, P('c:a/b').with_suffix, '/') - self.assertRaises(ValueError, P('c:a/b').with_suffix, '\\') - self.assertRaises(ValueError, P('c:a/b').with_suffix, 'c:') - self.assertRaises(ValueError, P('c:a/b').with_suffix, '/.gz') - self.assertRaises(ValueError, P('c:a/b').with_suffix, '\\.gz') - self.assertRaises(ValueError, P('c:a/b').with_suffix, 'c:.gz') - self.assertRaises(ValueError, P('c:a/b').with_suffix, 'c/d') - self.assertRaises(ValueError, P('c:a/b').with_suffix, 'c\\d') - self.assertRaises(ValueError, P('c:a/b').with_suffix, '.c/d') - self.assertRaises(ValueError, P('c:a/b').with_suffix, '.c\\d') - - def test_relative_to(self): - P = self.cls - p = P('C:Foo/Bar') - self.assertEqual(p.relative_to(P('c:')), P('Foo/Bar')) - self.assertEqual(p.relative_to('c:'), P('Foo/Bar')) - self.assertEqual(p.relative_to(P('c:foO')), P('Bar')) - self.assertEqual(p.relative_to('c:foO'), P('Bar')) - self.assertEqual(p.relative_to('c:foO/'), P('Bar')) - self.assertEqual(p.relative_to(P('c:foO/baR')), P()) - self.assertEqual(p.relative_to('c:foO/baR'), P()) - self.assertEqual(p.relative_to(P('c:'), walk_up=True), P('Foo/Bar')) - self.assertEqual(p.relative_to('c:', walk_up=True), P('Foo/Bar')) - self.assertEqual(p.relative_to(P('c:foO'), walk_up=True), P('Bar')) - self.assertEqual(p.relative_to('c:foO', walk_up=True), P('Bar')) - self.assertEqual(p.relative_to('c:foO/', walk_up=True), P('Bar')) - self.assertEqual(p.relative_to(P('c:foO/baR'), walk_up=True), P()) - self.assertEqual(p.relative_to('c:foO/baR', walk_up=True), P()) - self.assertEqual(p.relative_to(P('C:Foo/Bar/Baz'), walk_up=True), P('..')) - self.assertEqual(p.relative_to(P('C:Foo/Baz'), walk_up=True), P('../Bar')) - self.assertEqual(p.relative_to(P('C:Baz/Bar'), walk_up=True), P('../../Foo/Bar')) - # Unrelated paths. - self.assertRaises(ValueError, p.relative_to, P()) - self.assertRaises(ValueError, p.relative_to, '') - self.assertRaises(ValueError, p.relative_to, P('d:')) - self.assertRaises(ValueError, p.relative_to, P('/')) - self.assertRaises(ValueError, p.relative_to, P('Foo')) - self.assertRaises(ValueError, p.relative_to, P('/Foo')) - self.assertRaises(ValueError, p.relative_to, P('C:/Foo')) - self.assertRaises(ValueError, p.relative_to, P('C:Foo/Bar/Baz')) - self.assertRaises(ValueError, p.relative_to, P('C:Foo/Baz')) - self.assertRaises(ValueError, p.relative_to, P(), walk_up=True) - self.assertRaises(ValueError, p.relative_to, '', walk_up=True) - self.assertRaises(ValueError, p.relative_to, P('d:'), walk_up=True) - self.assertRaises(ValueError, p.relative_to, P('/'), walk_up=True) - self.assertRaises(ValueError, p.relative_to, P('Foo'), walk_up=True) - self.assertRaises(ValueError, p.relative_to, P('/Foo'), walk_up=True) - self.assertRaises(ValueError, p.relative_to, P('C:/Foo'), walk_up=True) - p = P('C:/Foo/Bar') - self.assertEqual(p.relative_to(P('c:/')), P('Foo/Bar')) - self.assertEqual(p.relative_to('c:/'), P('Foo/Bar')) - self.assertEqual(p.relative_to(P('c:/foO')), P('Bar')) - self.assertEqual(p.relative_to('c:/foO'), P('Bar')) - self.assertEqual(p.relative_to('c:/foO/'), P('Bar')) - self.assertEqual(p.relative_to(P('c:/foO/baR')), P()) - self.assertEqual(p.relative_to('c:/foO/baR'), P()) - self.assertEqual(p.relative_to(P('c:/'), walk_up=True), P('Foo/Bar')) - self.assertEqual(p.relative_to('c:/', walk_up=True), P('Foo/Bar')) - self.assertEqual(p.relative_to(P('c:/foO'), walk_up=True), P('Bar')) - self.assertEqual(p.relative_to('c:/foO', walk_up=True), P('Bar')) - self.assertEqual(p.relative_to('c:/foO/', walk_up=True), P('Bar')) - self.assertEqual(p.relative_to(P('c:/foO/baR'), walk_up=True), P()) - self.assertEqual(p.relative_to('c:/foO/baR', walk_up=True), P()) - self.assertEqual(p.relative_to('C:/Baz', walk_up=True), P('../Foo/Bar')) - self.assertEqual(p.relative_to('C:/Foo/Bar/Baz', walk_up=True), P('..')) - self.assertEqual(p.relative_to('C:/Foo/Baz', walk_up=True), P('../Bar')) - # Unrelated paths. - self.assertRaises(ValueError, p.relative_to, 'c:') - self.assertRaises(ValueError, p.relative_to, P('c:')) - self.assertRaises(ValueError, p.relative_to, P('C:/Baz')) - self.assertRaises(ValueError, p.relative_to, P('C:/Foo/Bar/Baz')) - self.assertRaises(ValueError, p.relative_to, P('C:/Foo/Baz')) - self.assertRaises(ValueError, p.relative_to, P('C:Foo')) - self.assertRaises(ValueError, p.relative_to, P('d:')) - self.assertRaises(ValueError, p.relative_to, P('d:/')) - self.assertRaises(ValueError, p.relative_to, P('/')) - self.assertRaises(ValueError, p.relative_to, P('/Foo')) - self.assertRaises(ValueError, p.relative_to, P('//C/Foo')) - self.assertRaises(ValueError, p.relative_to, 'c:', walk_up=True) - self.assertRaises(ValueError, p.relative_to, P('c:'), walk_up=True) - self.assertRaises(ValueError, p.relative_to, P('C:Foo'), walk_up=True) - self.assertRaises(ValueError, p.relative_to, P('d:'), walk_up=True) - self.assertRaises(ValueError, p.relative_to, P('d:/'), walk_up=True) - self.assertRaises(ValueError, p.relative_to, P('/'), walk_up=True) - self.assertRaises(ValueError, p.relative_to, P('/Foo'), walk_up=True) - self.assertRaises(ValueError, p.relative_to, P('//C/Foo'), walk_up=True) - # UNC paths. - p = P('//Server/Share/Foo/Bar') - self.assertEqual(p.relative_to(P('//sErver/sHare')), P('Foo/Bar')) - self.assertEqual(p.relative_to('//sErver/sHare'), P('Foo/Bar')) - self.assertEqual(p.relative_to('//sErver/sHare/'), P('Foo/Bar')) - self.assertEqual(p.relative_to(P('//sErver/sHare/Foo')), P('Bar')) - self.assertEqual(p.relative_to('//sErver/sHare/Foo'), P('Bar')) - self.assertEqual(p.relative_to('//sErver/sHare/Foo/'), P('Bar')) - self.assertEqual(p.relative_to(P('//sErver/sHare/Foo/Bar')), P()) - self.assertEqual(p.relative_to('//sErver/sHare/Foo/Bar'), P()) - self.assertEqual(p.relative_to(P('//sErver/sHare'), walk_up=True), P('Foo/Bar')) - self.assertEqual(p.relative_to('//sErver/sHare', walk_up=True), P('Foo/Bar')) - self.assertEqual(p.relative_to('//sErver/sHare/', walk_up=True), P('Foo/Bar')) - self.assertEqual(p.relative_to(P('//sErver/sHare/Foo'), walk_up=True), P('Bar')) - self.assertEqual(p.relative_to('//sErver/sHare/Foo', walk_up=True), P('Bar')) - self.assertEqual(p.relative_to('//sErver/sHare/Foo/', walk_up=True), P('Bar')) - self.assertEqual(p.relative_to(P('//sErver/sHare/Foo/Bar'), walk_up=True), P()) - self.assertEqual(p.relative_to('//sErver/sHare/Foo/Bar', walk_up=True), P()) - self.assertEqual(p.relative_to(P('//sErver/sHare/bar'), walk_up=True), P('../Foo/Bar')) - self.assertEqual(p.relative_to('//sErver/sHare/bar', walk_up=True), P('../Foo/Bar')) - # Unrelated paths. - self.assertRaises(ValueError, p.relative_to, P('/Server/Share/Foo')) - self.assertRaises(ValueError, p.relative_to, P('c:/Server/Share/Foo')) - self.assertRaises(ValueError, p.relative_to, P('//z/Share/Foo')) - self.assertRaises(ValueError, p.relative_to, P('//Server/z/Foo')) - self.assertRaises(ValueError, p.relative_to, P('/Server/Share/Foo'), walk_up=True) - self.assertRaises(ValueError, p.relative_to, P('c:/Server/Share/Foo'), walk_up=True) - self.assertRaises(ValueError, p.relative_to, P('//z/Share/Foo'), walk_up=True) - self.assertRaises(ValueError, p.relative_to, P('//Server/z/Foo'), walk_up=True) - - def test_is_relative_to(self): - P = self.cls - p = P('C:Foo/Bar') - self.assertTrue(p.is_relative_to(P('c:'))) - self.assertTrue(p.is_relative_to('c:')) - self.assertTrue(p.is_relative_to(P('c:foO'))) - self.assertTrue(p.is_relative_to('c:foO')) - self.assertTrue(p.is_relative_to('c:foO/')) - self.assertTrue(p.is_relative_to(P('c:foO/baR'))) - self.assertTrue(p.is_relative_to('c:foO/baR')) - # Unrelated paths. - self.assertFalse(p.is_relative_to(P())) - self.assertFalse(p.is_relative_to('')) - self.assertFalse(p.is_relative_to(P('d:'))) - self.assertFalse(p.is_relative_to(P('/'))) - self.assertFalse(p.is_relative_to(P('Foo'))) - self.assertFalse(p.is_relative_to(P('/Foo'))) - self.assertFalse(p.is_relative_to(P('C:/Foo'))) - self.assertFalse(p.is_relative_to(P('C:Foo/Bar/Baz'))) - self.assertFalse(p.is_relative_to(P('C:Foo/Baz'))) - p = P('C:/Foo/Bar') - self.assertTrue(p.is_relative_to(P('c:/'))) - self.assertTrue(p.is_relative_to(P('c:/foO'))) - self.assertTrue(p.is_relative_to('c:/foO/')) - self.assertTrue(p.is_relative_to(P('c:/foO/baR'))) - self.assertTrue(p.is_relative_to('c:/foO/baR')) - # Unrelated paths. - self.assertFalse(p.is_relative_to('c:')) - self.assertFalse(p.is_relative_to(P('C:/Baz'))) - self.assertFalse(p.is_relative_to(P('C:/Foo/Bar/Baz'))) - self.assertFalse(p.is_relative_to(P('C:/Foo/Baz'))) - self.assertFalse(p.is_relative_to(P('C:Foo'))) - self.assertFalse(p.is_relative_to(P('d:'))) - self.assertFalse(p.is_relative_to(P('d:/'))) - self.assertFalse(p.is_relative_to(P('/'))) - self.assertFalse(p.is_relative_to(P('/Foo'))) - self.assertFalse(p.is_relative_to(P('//C/Foo'))) - # UNC paths. - p = P('//Server/Share/Foo/Bar') - self.assertTrue(p.is_relative_to(P('//sErver/sHare'))) - self.assertTrue(p.is_relative_to('//sErver/sHare')) - self.assertTrue(p.is_relative_to('//sErver/sHare/')) - self.assertTrue(p.is_relative_to(P('//sErver/sHare/Foo'))) - self.assertTrue(p.is_relative_to('//sErver/sHare/Foo')) - self.assertTrue(p.is_relative_to('//sErver/sHare/Foo/')) - self.assertTrue(p.is_relative_to(P('//sErver/sHare/Foo/Bar'))) - self.assertTrue(p.is_relative_to('//sErver/sHare/Foo/Bar')) - # Unrelated paths. - self.assertFalse(p.is_relative_to(P('/Server/Share/Foo'))) - self.assertFalse(p.is_relative_to(P('c:/Server/Share/Foo'))) - self.assertFalse(p.is_relative_to(P('//z/Share/Foo'))) - self.assertFalse(p.is_relative_to(P('//Server/z/Foo'))) - - def test_is_absolute(self): - P = self.cls - # Under NT, only paths with both a drive and a root are absolute. - self.assertFalse(P().is_absolute()) - self.assertFalse(P('a').is_absolute()) - self.assertFalse(P('a/b/').is_absolute()) - self.assertFalse(P('/').is_absolute()) - self.assertFalse(P('/a').is_absolute()) - self.assertFalse(P('/a/b/').is_absolute()) - self.assertFalse(P('c:').is_absolute()) - self.assertFalse(P('c:a').is_absolute()) - self.assertFalse(P('c:a/b/').is_absolute()) - self.assertTrue(P('c:/').is_absolute()) - self.assertTrue(P('c:/a').is_absolute()) - self.assertTrue(P('c:/a/b/').is_absolute()) - # UNC paths are absolute by definition. - self.assertTrue(P('//a/b').is_absolute()) - self.assertTrue(P('//a/b/').is_absolute()) - self.assertTrue(P('//a/b/c').is_absolute()) - self.assertTrue(P('//a/b/c/d').is_absolute()) - - def test_join(self): - P = self.cls - p = P('C:/a/b') - pp = p.joinpath('x/y') - self.assertEqual(pp, P('C:/a/b/x/y')) - pp = p.joinpath('/x/y') - self.assertEqual(pp, P('C:/x/y')) - # Joining with a different drive => the first path is ignored, even - # if the second path is relative. - pp = p.joinpath('D:x/y') - self.assertEqual(pp, P('D:x/y')) - pp = p.joinpath('D:/x/y') - self.assertEqual(pp, P('D:/x/y')) - pp = p.joinpath('//host/share/x/y') - self.assertEqual(pp, P('//host/share/x/y')) - # Joining with the same drive => the first path is appended to if - # the second path is relative. - pp = p.joinpath('c:x/y') - self.assertEqual(pp, P('C:/a/b/x/y')) - pp = p.joinpath('c:/x/y') - self.assertEqual(pp, P('C:/x/y')) - # Joining with files with NTFS data streams => the filename should - # not be parsed as a drive letter - pp = p.joinpath(P('./d:s')) - self.assertEqual(pp, P('C:/a/b/d:s')) - pp = p.joinpath(P('./dd:s')) - self.assertEqual(pp, P('C:/a/b/dd:s')) - pp = p.joinpath(P('E:d:s')) - self.assertEqual(pp, P('E:d:s')) - # Joining onto a UNC path with no root - pp = P('//').joinpath('server') - self.assertEqual(pp, P('//server')) - pp = P('//server').joinpath('share') - self.assertEqual(pp, P('//server/share')) - pp = P('//./BootPartition').joinpath('Windows') - self.assertEqual(pp, P('//./BootPartition/Windows')) - - def test_div(self): - # Basically the same as joinpath(). - P = self.cls - p = P('C:/a/b') - self.assertEqual(p / 'x/y', P('C:/a/b/x/y')) - self.assertEqual(p / 'x' / 'y', P('C:/a/b/x/y')) - self.assertEqual(p / '/x/y', P('C:/x/y')) - self.assertEqual(p / '/x' / 'y', P('C:/x/y')) - # Joining with a different drive => the first path is ignored, even - # if the second path is relative. - self.assertEqual(p / 'D:x/y', P('D:x/y')) - self.assertEqual(p / 'D:' / 'x/y', P('D:x/y')) - self.assertEqual(p / 'D:/x/y', P('D:/x/y')) - self.assertEqual(p / 'D:' / '/x/y', P('D:/x/y')) - self.assertEqual(p / '//host/share/x/y', P('//host/share/x/y')) - # Joining with the same drive => the first path is appended to if - # the second path is relative. - self.assertEqual(p / 'c:x/y', P('C:/a/b/x/y')) - self.assertEqual(p / 'c:/x/y', P('C:/x/y')) - # Joining with files with NTFS data streams => the filename should - # not be parsed as a drive letter - self.assertEqual(p / P('./d:s'), P('C:/a/b/d:s')) - self.assertEqual(p / P('./dd:s'), P('C:/a/b/dd:s')) - self.assertEqual(p / P('E:d:s'), P('E:d:s')) - - def test_is_reserved(self): - P = self.cls - self.assertIs(False, P('').is_reserved()) - self.assertIs(False, P('/').is_reserved()) - self.assertIs(False, P('/foo/bar').is_reserved()) - # UNC paths are never reserved. - self.assertIs(False, P('//my/share/nul/con/aux').is_reserved()) - # Case-insensitive DOS-device names are reserved. - self.assertIs(True, P('nul').is_reserved()) - self.assertIs(True, P('aux').is_reserved()) - self.assertIs(True, P('prn').is_reserved()) - self.assertIs(True, P('con').is_reserved()) - self.assertIs(True, P('conin$').is_reserved()) - self.assertIs(True, P('conout$').is_reserved()) - # COM/LPT + 1-9 or + superscript 1-3 are reserved. - self.assertIs(True, P('COM1').is_reserved()) - self.assertIs(True, P('LPT9').is_reserved()) - self.assertIs(True, P('com\xb9').is_reserved()) - self.assertIs(True, P('com\xb2').is_reserved()) - self.assertIs(True, P('lpt\xb3').is_reserved()) - # DOS-device name mataching ignores characters after a dot or - # a colon and also ignores trailing spaces. - self.assertIs(True, P('NUL.txt').is_reserved()) - self.assertIs(True, P('PRN ').is_reserved()) - self.assertIs(True, P('AUX .txt').is_reserved()) - self.assertIs(True, P('COM1:bar').is_reserved()) - self.assertIs(True, P('LPT9 :bar').is_reserved()) - # DOS-device names are only matched at the beginning - # of a path component. - self.assertIs(False, P('bar.com9').is_reserved()) - self.assertIs(False, P('bar.lpt9').is_reserved()) - # Only the last path component matters. - self.assertIs(True, P('c:/baz/con/NUL').is_reserved()) - self.assertIs(False, P('c:/NUL/con/baz').is_reserved()) - -class PurePathTest(_BasePurePathTest, unittest.TestCase): - cls = pathlib.PurePath - - def test_concrete_class(self): - p = self.cls('a') - self.assertIs(type(p), - pathlib.PureWindowsPath if os.name == 'nt' else pathlib.PurePosixPath) - - def test_different_flavours_unequal(self): - p = pathlib.PurePosixPath('a') - q = pathlib.PureWindowsPath('a') - self.assertNotEqual(p, q) - - def test_different_flavours_unordered(self): - p = pathlib.PurePosixPath('a') - q = pathlib.PureWindowsPath('a') - with self.assertRaises(TypeError): - p < q - with self.assertRaises(TypeError): - p <= q - with self.assertRaises(TypeError): - p > q - with self.assertRaises(TypeError): - p >= q - - -# -# Tests for the concrete classes. -# - -# Make sure any symbolic links in the base test path are resolved. -BASE = os.path.realpath(TESTFN) -join = lambda *x: os.path.join(BASE, *x) -rel_join = lambda *x: os.path.join(TESTFN, *x) - -only_nt = unittest.skipIf(os.name != 'nt', - 'test requires a Windows-compatible system') -only_posix = unittest.skipIf(os.name == 'nt', - 'test requires a POSIX-compatible system') - -@only_posix -class PosixPathAsPureTest(PurePosixPathTest): - cls = pathlib.PosixPath - -@only_nt -class WindowsPathAsPureTest(PureWindowsPathTest): - cls = pathlib.WindowsPath - - def test_owner(self): - P = self.cls - with self.assertRaises(NotImplementedError): - P('c:/').owner() - - def test_group(self): - P = self.cls - with self.assertRaises(NotImplementedError): - P('c:/').group() - - -class _BasePathTest(object): - """Tests for the FS-accessing functionalities of the Path classes.""" - - # (BASE) - # | - # |-- brokenLink -> non-existing - # |-- dirA - # | `-- linkC -> ../dirB - # |-- dirB - # | |-- fileB - # | `-- linkD -> ../dirB - # |-- dirC - # | |-- dirD - # | | `-- fileD - # | `-- fileC - # | `-- novel.txt - # |-- dirE # No permissions - # |-- fileA - # |-- linkA -> fileA - # |-- linkB -> dirB - # `-- brokenLinkLoop -> brokenLinkLoop - # - - def setUp(self): - def cleanup(): - os.chmod(join('dirE'), 0o777) - os_helper.rmtree(BASE) - self.addCleanup(cleanup) - os.mkdir(BASE) - os.mkdir(join('dirA')) - os.mkdir(join('dirB')) - os.mkdir(join('dirC')) - os.mkdir(join('dirC', 'dirD')) - os.mkdir(join('dirE')) - with open(join('fileA'), 'wb') as f: - f.write(b"this is file A\n") - with open(join('dirB', 'fileB'), 'wb') as f: - f.write(b"this is file B\n") - with open(join('dirC', 'fileC'), 'wb') as f: - f.write(b"this is file C\n") - with open(join('dirC', 'novel.txt'), 'wb') as f: - f.write(b"this is a novel\n") - with open(join('dirC', 'dirD', 'fileD'), 'wb') as f: - f.write(b"this is file D\n") - os.chmod(join('dirE'), 0) - if os_helper.can_symlink(): - # Relative symlinks. - os.symlink('fileA', join('linkA')) - os.symlink('non-existing', join('brokenLink')) - self.dirlink('dirB', join('linkB')) - self.dirlink(os.path.join('..', 'dirB'), join('dirA', 'linkC')) - # This one goes upwards, creating a loop. - self.dirlink(os.path.join('..', 'dirB'), join('dirB', 'linkD')) - # Broken symlink (pointing to itself). - os.symlink('brokenLinkLoop', join('brokenLinkLoop')) - - if os.name == 'nt': - # Workaround for http://bugs.python.org/issue13772. - def dirlink(self, src, dest): - os.symlink(src, dest, target_is_directory=True) - else: - def dirlink(self, src, dest): - os.symlink(src, dest) - - def assertSame(self, path_a, path_b): - self.assertTrue(os.path.samefile(str(path_a), str(path_b)), - "%r and %r don't point to the same file" % - (path_a, path_b)) - - def assertFileNotFound(self, func, *args, **kwargs): - with self.assertRaises(FileNotFoundError) as cm: - func(*args, **kwargs) - self.assertEqual(cm.exception.errno, errno.ENOENT) - - def assertEqualNormCase(self, path_a, path_b): - self.assertEqual(os.path.normcase(path_a), os.path.normcase(path_b)) - - def _test_cwd(self, p): - q = self.cls(os.getcwd()) - self.assertEqual(p, q) - self.assertEqualNormCase(str(p), str(q)) - self.assertIs(type(p), type(q)) - self.assertTrue(p.is_absolute()) - - def test_cwd(self): - p = self.cls.cwd() - self._test_cwd(p) - - def test_absolute_common(self): - P = self.cls - - with mock.patch("os.getcwd") as getcwd: - getcwd.return_value = BASE - - # Simple relative paths. - self.assertEqual(str(P().absolute()), BASE) - self.assertEqual(str(P('.').absolute()), BASE) - self.assertEqual(str(P('a').absolute()), os.path.join(BASE, 'a')) - self.assertEqual(str(P('a', 'b', 'c').absolute()), os.path.join(BASE, 'a', 'b', 'c')) - - # Symlinks should not be resolved. - self.assertEqual(str(P('linkB', 'fileB').absolute()), os.path.join(BASE, 'linkB', 'fileB')) - self.assertEqual(str(P('brokenLink').absolute()), os.path.join(BASE, 'brokenLink')) - self.assertEqual(str(P('brokenLinkLoop').absolute()), os.path.join(BASE, 'brokenLinkLoop')) - - # '..' entries should be preserved and not normalised. - self.assertEqual(str(P('..').absolute()), os.path.join(BASE, '..')) - self.assertEqual(str(P('a', '..').absolute()), os.path.join(BASE, 'a', '..')) - self.assertEqual(str(P('..', 'b').absolute()), os.path.join(BASE, '..', 'b')) - - def _test_home(self, p): - q = self.cls(os.path.expanduser('~')) - self.assertEqual(p, q) - self.assertEqualNormCase(str(p), str(q)) - self.assertIs(type(p), type(q)) - self.assertTrue(p.is_absolute()) - - @unittest.skipIf( - pwd is None, reason="Test requires pwd module to get homedir." - ) - def test_home(self): - with os_helper.EnvironmentVarGuard() as env: - self._test_home(self.cls.home()) - - env.clear() - env['USERPROFILE'] = os.path.join(BASE, 'userprofile') - self._test_home(self.cls.home()) - - # bpo-38883: ignore `HOME` when set on windows - env['HOME'] = os.path.join(BASE, 'home') - self._test_home(self.cls.home()) - - @unittest.skip("TODO: RUSTPYTHON; PyObject::set_slot index out of bounds") - def test_with_segments(self): - class P(_BasePurePathSubclass, self.cls): - pass - p = P(BASE, session_id=42) - self.assertEqual(42, p.absolute().session_id) - self.assertEqual(42, p.resolve().session_id) - if not is_wasi: # WASI has no user accounts. - self.assertEqual(42, p.with_segments('~').expanduser().session_id) - self.assertEqual(42, (p / 'fileA').rename(p / 'fileB').session_id) - self.assertEqual(42, (p / 'fileB').replace(p / 'fileA').session_id) - if os_helper.can_symlink(): - self.assertEqual(42, (p / 'linkA').readlink().session_id) - for path in p.iterdir(): - self.assertEqual(42, path.session_id) - for path in p.glob('*'): - self.assertEqual(42, path.session_id) - for path in p.rglob('*'): - self.assertEqual(42, path.session_id) - for dirpath, dirnames, filenames in p.walk(): - self.assertEqual(42, dirpath.session_id) - - def test_samefile(self): - fileA_path = os.path.join(BASE, 'fileA') - fileB_path = os.path.join(BASE, 'dirB', 'fileB') - p = self.cls(fileA_path) - pp = self.cls(fileA_path) - q = self.cls(fileB_path) - self.assertTrue(p.samefile(fileA_path)) - self.assertTrue(p.samefile(pp)) - self.assertFalse(p.samefile(fileB_path)) - self.assertFalse(p.samefile(q)) - # Test the non-existent file case - non_existent = os.path.join(BASE, 'foo') - r = self.cls(non_existent) - self.assertRaises(FileNotFoundError, p.samefile, r) - self.assertRaises(FileNotFoundError, p.samefile, non_existent) - self.assertRaises(FileNotFoundError, r.samefile, p) - self.assertRaises(FileNotFoundError, r.samefile, non_existent) - self.assertRaises(FileNotFoundError, r.samefile, r) - self.assertRaises(FileNotFoundError, r.samefile, non_existent) - - def test_empty_path(self): - # The empty path points to '.' - p = self.cls('') - self.assertEqual(p.stat(), os.stat('.')) - - @unittest.skipIf(is_wasi, "WASI has no user accounts.") - def test_expanduser_common(self): - P = self.cls - p = P('~') - self.assertEqual(p.expanduser(), P(os.path.expanduser('~'))) - p = P('foo') - self.assertEqual(p.expanduser(), p) - p = P('/~') - self.assertEqual(p.expanduser(), p) - p = P('../~') - self.assertEqual(p.expanduser(), p) - p = P(P('').absolute().anchor) / '~' - self.assertEqual(p.expanduser(), p) - p = P('~/a:b') - self.assertEqual(p.expanduser(), P(os.path.expanduser('~'), './a:b')) - - def test_exists(self): - P = self.cls - p = P(BASE) - self.assertIs(True, p.exists()) - self.assertIs(True, (p / 'dirA').exists()) - self.assertIs(True, (p / 'fileA').exists()) - self.assertIs(False, (p / 'fileA' / 'bah').exists()) - if os_helper.can_symlink(): - self.assertIs(True, (p / 'linkA').exists()) - self.assertIs(True, (p / 'linkB').exists()) - self.assertIs(True, (p / 'linkB' / 'fileB').exists()) - self.assertIs(False, (p / 'linkA' / 'bah').exists()) - self.assertIs(False, (p / 'brokenLink').exists()) - self.assertIs(True, (p / 'brokenLink').exists(follow_symlinks=False)) - self.assertIs(False, (p / 'foo').exists()) - self.assertIs(False, P('/xyzzy').exists()) - self.assertIs(False, P(BASE + '\udfff').exists()) - self.assertIs(False, P(BASE + '\x00').exists()) - - def test_open_common(self): - p = self.cls(BASE) - with (p / 'fileA').open('r') as f: - self.assertIsInstance(f, io.TextIOBase) - self.assertEqual(f.read(), "this is file A\n") - with (p / 'fileA').open('rb') as f: - self.assertIsInstance(f, io.BufferedIOBase) - self.assertEqual(f.read().strip(), b"this is file A") - with (p / 'fileA').open('rb', buffering=0) as f: - self.assertIsInstance(f, io.RawIOBase) - self.assertEqual(f.read().strip(), b"this is file A") - - def test_read_write_bytes(self): - p = self.cls(BASE) - (p / 'fileA').write_bytes(b'abcdefg') - self.assertEqual((p / 'fileA').read_bytes(), b'abcdefg') - # Check that trying to write str does not truncate the file. - self.assertRaises(TypeError, (p / 'fileA').write_bytes, 'somestr') - self.assertEqual((p / 'fileA').read_bytes(), b'abcdefg') - - def test_read_write_text(self): - p = self.cls(BASE) - (p / 'fileA').write_text('äbcdefg', encoding='latin-1') - self.assertEqual((p / 'fileA').read_text( - encoding='utf-8', errors='ignore'), 'bcdefg') - # Check that trying to write bytes does not truncate the file. - self.assertRaises(TypeError, (p / 'fileA').write_text, b'somebytes') - self.assertEqual((p / 'fileA').read_text(encoding='latin-1'), 'äbcdefg') - - def test_write_text_with_newlines(self): - p = self.cls(BASE) - # Check that `\n` character change nothing - (p / 'fileA').write_text('abcde\r\nfghlk\n\rmnopq', newline='\n') - self.assertEqual((p / 'fileA').read_bytes(), - b'abcde\r\nfghlk\n\rmnopq') - # Check that `\r` character replaces `\n` - (p / 'fileA').write_text('abcde\r\nfghlk\n\rmnopq', newline='\r') - self.assertEqual((p / 'fileA').read_bytes(), - b'abcde\r\rfghlk\r\rmnopq') - # Check that `\r\n` character replaces `\n` - (p / 'fileA').write_text('abcde\r\nfghlk\n\rmnopq', newline='\r\n') - self.assertEqual((p / 'fileA').read_bytes(), - b'abcde\r\r\nfghlk\r\n\rmnopq') - # Check that no argument passed will change `\n` to `os.linesep` - os_linesep_byte = bytes(os.linesep, encoding='ascii') - (p / 'fileA').write_text('abcde\nfghlk\n\rmnopq') - self.assertEqual((p / 'fileA').read_bytes(), - b'abcde' + os_linesep_byte + b'fghlk' + os_linesep_byte + b'\rmnopq') - - def test_iterdir(self): - P = self.cls - p = P(BASE) - it = p.iterdir() - paths = set(it) - expected = ['dirA', 'dirB', 'dirC', 'dirE', 'fileA'] - if os_helper.can_symlink(): - expected += ['linkA', 'linkB', 'brokenLink', 'brokenLinkLoop'] - self.assertEqual(paths, { P(BASE, q) for q in expected }) - - @os_helper.skip_unless_symlink - def test_iterdir_symlink(self): - # __iter__ on a symlink to a directory. - P = self.cls - p = P(BASE, 'linkB') - paths = set(p.iterdir()) - expected = { P(BASE, 'linkB', q) for q in ['fileB', 'linkD'] } - self.assertEqual(paths, expected) - - def test_iterdir_nodir(self): - # __iter__ on something that is not a directory. - p = self.cls(BASE, 'fileA') - with self.assertRaises(OSError) as cm: - next(p.iterdir()) - # ENOENT or EINVAL under Windows, ENOTDIR otherwise - # (see issue #12802). - self.assertIn(cm.exception.errno, (errno.ENOTDIR, - errno.ENOENT, errno.EINVAL)) - - def test_glob_common(self): - def _check(glob, expected): - self.assertEqual(set(glob), { P(BASE, q) for q in expected }) - P = self.cls - p = P(BASE) - it = p.glob("fileA") - self.assertIsInstance(it, collections.abc.Iterator) - _check(it, ["fileA"]) - _check(p.glob("fileB"), []) - _check(p.glob("dir*/file*"), ["dirB/fileB", "dirC/fileC"]) - if not os_helper.can_symlink(): - _check(p.glob("*A"), ['dirA', 'fileA']) - else: - _check(p.glob("*A"), ['dirA', 'fileA', 'linkA']) - if not os_helper.can_symlink(): - _check(p.glob("*B/*"), ['dirB/fileB']) - else: - _check(p.glob("*B/*"), ['dirB/fileB', 'dirB/linkD', - 'linkB/fileB', 'linkB/linkD']) - if not os_helper.can_symlink(): - _check(p.glob("*/fileB"), ['dirB/fileB']) - else: - _check(p.glob("*/fileB"), ['dirB/fileB', 'linkB/fileB']) - if os_helper.can_symlink(): - _check(p.glob("brokenLink"), ['brokenLink']) - - if not os_helper.can_symlink(): - _check(p.glob("*/"), ["dirA", "dirB", "dirC", "dirE"]) - else: - _check(p.glob("*/"), ["dirA", "dirB", "dirC", "dirE", "linkB"]) - - def test_glob_case_sensitive(self): - P = self.cls - def _check(path, pattern, case_sensitive, expected): - actual = {str(q) for q in path.glob(pattern, case_sensitive=case_sensitive)} - expected = {str(P(BASE, q)) for q in expected} - self.assertEqual(actual, expected) - path = P(BASE) - _check(path, "DIRB/FILE*", True, []) - _check(path, "DIRB/FILE*", False, ["dirB/fileB"]) - _check(path, "dirb/file*", True, []) - _check(path, "dirb/file*", False, ["dirB/fileB"]) - - def test_rglob_common(self): - def _check(glob, expected): - self.assertEqual(sorted(glob), sorted(P(BASE, q) for q in expected)) - P = self.cls - p = P(BASE) - it = p.rglob("fileA") - self.assertIsInstance(it, collections.abc.Iterator) - _check(it, ["fileA"]) - _check(p.rglob("fileB"), ["dirB/fileB"]) - _check(p.rglob("**/fileB"), ["dirB/fileB"]) - _check(p.rglob("*/fileA"), []) - if not os_helper.can_symlink(): - _check(p.rglob("*/fileB"), ["dirB/fileB"]) - else: - _check(p.rglob("*/fileB"), ["dirB/fileB", "dirB/linkD/fileB", - "linkB/fileB", "dirA/linkC/fileB"]) - _check(p.rglob("file*"), ["fileA", "dirB/fileB", - "dirC/fileC", "dirC/dirD/fileD"]) - if not os_helper.can_symlink(): - _check(p.rglob("*/"), [ - "dirA", "dirB", "dirC", "dirC/dirD", "dirE", - ]) - else: - _check(p.rglob("*/"), [ - "dirA", "dirA/linkC", "dirB", "dirB/linkD", "dirC", - "dirC/dirD", "dirE", "linkB", - ]) - _check(p.rglob(""), ["", "dirA", "dirB", "dirC", "dirE", "dirC/dirD"]) - - p = P(BASE, "dirC") - _check(p.rglob("*"), ["dirC/fileC", "dirC/novel.txt", - "dirC/dirD", "dirC/dirD/fileD"]) - _check(p.rglob("file*"), ["dirC/fileC", "dirC/dirD/fileD"]) - _check(p.rglob("**/file*"), ["dirC/fileC", "dirC/dirD/fileD"]) - _check(p.rglob("dir*/**"), ["dirC/dirD"]) - _check(p.rglob("*/*"), ["dirC/dirD/fileD"]) - _check(p.rglob("*/"), ["dirC/dirD"]) - _check(p.rglob(""), ["dirC", "dirC/dirD"]) - _check(p.rglob("**"), ["dirC", "dirC/dirD"]) - # gh-91616, a re module regression - _check(p.rglob("*.txt"), ["dirC/novel.txt"]) - _check(p.rglob("*.*"), ["dirC/novel.txt"]) - - @os_helper.skip_unless_symlink - def test_rglob_symlink_loop(self): - # Don't get fooled by symlink loops (Issue #26012). - P = self.cls - p = P(BASE) - given = set(p.rglob('*')) - expect = {'brokenLink', - 'dirA', 'dirA/linkC', - 'dirB', 'dirB/fileB', 'dirB/linkD', - 'dirC', 'dirC/dirD', 'dirC/dirD/fileD', - 'dirC/fileC', 'dirC/novel.txt', - 'dirE', - 'fileA', - 'linkA', - 'linkB', - 'brokenLinkLoop', - } - self.assertEqual(given, {p / x for x in expect}) - - def test_glob_many_open_files(self): - depth = 30 - P = self.cls - base = P(BASE) / 'deep' - p = P(base, *(['d']*depth)) - p.mkdir(parents=True) - pattern = '/'.join(['*'] * depth) - iters = [base.glob(pattern) for j in range(100)] - for it in iters: - self.assertEqual(next(it), p) - iters = [base.rglob('d') for j in range(100)] - p = base - for i in range(depth): - p = p / 'd' - for it in iters: - self.assertEqual(next(it), p) - - def test_glob_dotdot(self): - # ".." is not special in globs. - P = self.cls - p = P(BASE) - self.assertEqual(set(p.glob("..")), { P(BASE, "..") }) - self.assertEqual(set(p.glob("../..")), { P(BASE, "..", "..") }) - self.assertEqual(set(p.glob("dirA/..")), { P(BASE, "dirA", "..") }) - self.assertEqual(set(p.glob("dirA/../file*")), { P(BASE, "dirA/../fileA") }) - self.assertEqual(set(p.glob("dirA/../file*/..")), set()) - self.assertEqual(set(p.glob("../xyzzy")), set()) - self.assertEqual(set(p.glob("xyzzy/..")), set()) - self.assertEqual(set(p.glob("/".join([".."] * 50))), { P(BASE, *[".."] * 50)}) - - @os_helper.skip_unless_symlink - def test_glob_permissions(self): - # See bpo-38894 - P = self.cls - base = P(BASE) / 'permissions' - base.mkdir() - self.addCleanup(os_helper.rmtree, base) - - for i in range(100): - link = base / f"link{i}" - if i % 2: - link.symlink_to(P(BASE, "dirE", "nonexistent")) - else: - link.symlink_to(P(BASE, "dirC")) - - self.assertEqual(len(set(base.glob("*"))), 100) - self.assertEqual(len(set(base.glob("*/"))), 50) - self.assertEqual(len(set(base.glob("*/fileC"))), 50) - self.assertEqual(len(set(base.glob("*/file*"))), 50) - - @os_helper.skip_unless_symlink - def test_glob_long_symlink(self): - # See gh-87695 - base = self.cls(BASE) / 'long_symlink' - base.mkdir() - bad_link = base / 'bad_link' - bad_link.symlink_to("bad" * 200) - self.assertEqual(sorted(base.glob('**/*')), [bad_link]) - - def test_glob_above_recursion_limit(self): - recursion_limit = 50 - # directory_depth > recursion_limit - directory_depth = recursion_limit + 10 - base = pathlib.Path(os_helper.TESTFN, 'deep') - path = pathlib.Path(base, *(['d'] * directory_depth)) - path.mkdir(parents=True) - - with set_recursion_limit(recursion_limit): - list(base.glob('**')) - - def _check_resolve(self, p, expected, strict=True): - q = p.resolve(strict) - self.assertEqual(q, expected) - - # This can be used to check both relative and absolute resolutions. - _check_resolve_relative = _check_resolve_absolute = _check_resolve - - @os_helper.skip_unless_symlink - def test_resolve_common(self): - P = self.cls - p = P(BASE, 'foo') - with self.assertRaises(OSError) as cm: - p.resolve(strict=True) - self.assertEqual(cm.exception.errno, errno.ENOENT) - # Non-strict - self.assertEqualNormCase(str(p.resolve(strict=False)), - os.path.join(BASE, 'foo')) - p = P(BASE, 'foo', 'in', 'spam') - self.assertEqualNormCase(str(p.resolve(strict=False)), - os.path.join(BASE, 'foo', 'in', 'spam')) - p = P(BASE, '..', 'foo', 'in', 'spam') - self.assertEqualNormCase(str(p.resolve(strict=False)), - os.path.abspath(os.path.join('foo', 'in', 'spam'))) - # These are all relative symlinks. - p = P(BASE, 'dirB', 'fileB') - self._check_resolve_relative(p, p) - p = P(BASE, 'linkA') - self._check_resolve_relative(p, P(BASE, 'fileA')) - p = P(BASE, 'dirA', 'linkC', 'fileB') - self._check_resolve_relative(p, P(BASE, 'dirB', 'fileB')) - p = P(BASE, 'dirB', 'linkD', 'fileB') - self._check_resolve_relative(p, P(BASE, 'dirB', 'fileB')) - # Non-strict - p = P(BASE, 'dirA', 'linkC', 'fileB', 'foo', 'in', 'spam') - self._check_resolve_relative(p, P(BASE, 'dirB', 'fileB', 'foo', 'in', - 'spam'), False) - p = P(BASE, 'dirA', 'linkC', '..', 'foo', 'in', 'spam') - if os.name == 'nt': - # In Windows, if linkY points to dirB, 'dirA\linkY\..' - # resolves to 'dirA' without resolving linkY first. - self._check_resolve_relative(p, P(BASE, 'dirA', 'foo', 'in', - 'spam'), False) - else: - # In Posix, if linkY points to dirB, 'dirA/linkY/..' - # resolves to 'dirB/..' first before resolving to parent of dirB. - self._check_resolve_relative(p, P(BASE, 'foo', 'in', 'spam'), False) - # Now create absolute symlinks. - d = os_helper._longpath(tempfile.mkdtemp(suffix='-dirD', - dir=os.getcwd())) - self.addCleanup(os_helper.rmtree, d) - os.symlink(os.path.join(d), join('dirA', 'linkX')) - os.symlink(join('dirB'), os.path.join(d, 'linkY')) - p = P(BASE, 'dirA', 'linkX', 'linkY', 'fileB') - self._check_resolve_absolute(p, P(BASE, 'dirB', 'fileB')) - # Non-strict - p = P(BASE, 'dirA', 'linkX', 'linkY', 'foo', 'in', 'spam') - self._check_resolve_relative(p, P(BASE, 'dirB', 'foo', 'in', 'spam'), - False) - p = P(BASE, 'dirA', 'linkX', 'linkY', '..', 'foo', 'in', 'spam') - if os.name == 'nt': - # In Windows, if linkY points to dirB, 'dirA\linkY\..' - # resolves to 'dirA' without resolving linkY first. - self._check_resolve_relative(p, P(d, 'foo', 'in', 'spam'), False) - else: - # In Posix, if linkY points to dirB, 'dirA/linkY/..' - # resolves to 'dirB/..' first before resolving to parent of dirB. - self._check_resolve_relative(p, P(BASE, 'foo', 'in', 'spam'), False) - - @os_helper.skip_unless_symlink - def test_resolve_dot(self): - # See http://web.archive.org/web/20200623062557/https://bitbucket.org/pitrou/pathlib/issues/9/ - p = self.cls(BASE) - self.dirlink('.', join('0')) - self.dirlink(os.path.join('0', '0'), join('1')) - self.dirlink(os.path.join('1', '1'), join('2')) - q = p / '2' - self.assertEqual(q.resolve(strict=True), p) - r = q / '3' / '4' - self.assertRaises(FileNotFoundError, r.resolve, strict=True) - # Non-strict - self.assertEqual(r.resolve(strict=False), p / '3' / '4') - - def test_resolve_nonexist_relative_issue38671(self): - p = self.cls('non', 'exist') - - old_cwd = os.getcwd() - os.chdir(BASE) - try: - self.assertEqual(p.resolve(), self.cls(BASE, p)) - finally: - os.chdir(old_cwd) - - def test_with(self): - p = self.cls(BASE) - it = p.iterdir() - it2 = p.iterdir() - next(it2) - # bpo-46556: path context managers are deprecated in Python 3.11. - with self.assertWarns(DeprecationWarning): - with p: - pass - # Using a path as a context manager is a no-op, thus the following - # operations should still succeed after the context manage exits. - next(it) - next(it2) - p.exists() - p.resolve() - p.absolute() - with self.assertWarns(DeprecationWarning): - with p: - pass - - @os_helper.skip_unless_working_chmod - def test_chmod(self): - p = self.cls(BASE) / 'fileA' - mode = p.stat().st_mode - # Clear writable bit. - new_mode = mode & ~0o222 - p.chmod(new_mode) - self.assertEqual(p.stat().st_mode, new_mode) - # Set writable bit. - new_mode = mode | 0o222 - p.chmod(new_mode) - self.assertEqual(p.stat().st_mode, new_mode) - - # On Windows, os.chmod does not follow symlinks (issue #15411) - @only_posix - @os_helper.skip_unless_working_chmod - def test_chmod_follow_symlinks_true(self): - p = self.cls(BASE) / 'linkA' - q = p.resolve() - mode = q.stat().st_mode - # Clear writable bit. - new_mode = mode & ~0o222 - p.chmod(new_mode, follow_symlinks=True) - self.assertEqual(q.stat().st_mode, new_mode) - # Set writable bit - new_mode = mode | 0o222 - p.chmod(new_mode, follow_symlinks=True) - self.assertEqual(q.stat().st_mode, new_mode) - - # XXX also need a test for lchmod. - - @os_helper.skip_unless_working_chmod - def test_stat(self): - p = self.cls(BASE) / 'fileA' - st = p.stat() - self.assertEqual(p.stat(), st) - # Change file mode by flipping write bit. - p.chmod(st.st_mode ^ 0o222) - self.addCleanup(p.chmod, st.st_mode) - self.assertNotEqual(p.stat(), st) - - @os_helper.skip_unless_symlink - def test_stat_no_follow_symlinks(self): - p = self.cls(BASE) / 'linkA' - st = p.stat() - self.assertNotEqual(st, p.stat(follow_symlinks=False)) - - def test_stat_no_follow_symlinks_nosymlink(self): - p = self.cls(BASE) / 'fileA' - st = p.stat() - self.assertEqual(st, p.stat(follow_symlinks=False)) - - @os_helper.skip_unless_symlink - def test_lstat(self): - p = self.cls(BASE)/ 'linkA' - st = p.stat() - self.assertNotEqual(st, p.lstat()) - - def test_lstat_nosymlink(self): - p = self.cls(BASE) / 'fileA' - st = p.stat() - self.assertEqual(st, p.lstat()) - - @unittest.skipUnless(pwd, "the pwd module is needed for this test") - def test_owner(self): - p = self.cls(BASE) / 'fileA' - uid = p.stat().st_uid - try: - name = pwd.getpwuid(uid).pw_name - except KeyError: - self.skipTest( - "user %d doesn't have an entry in the system database" % uid) - self.assertEqual(name, p.owner()) - - @unittest.skipUnless(grp, "the grp module is needed for this test") - def test_group(self): - p = self.cls(BASE) / 'fileA' - gid = p.stat().st_gid - try: - name = grp.getgrgid(gid).gr_name - except KeyError: - self.skipTest( - "group %d doesn't have an entry in the system database" % gid) - self.assertEqual(name, p.group()) - - def test_unlink(self): - p = self.cls(BASE) / 'fileA' - p.unlink() - self.assertFileNotFound(p.stat) - self.assertFileNotFound(p.unlink) - - def test_unlink_missing_ok(self): - p = self.cls(BASE) / 'fileAAA' - self.assertFileNotFound(p.unlink) - p.unlink(missing_ok=True) - - def test_rmdir(self): - p = self.cls(BASE) / 'dirA' - for q in p.iterdir(): - q.unlink() - p.rmdir() - self.assertFileNotFound(p.stat) - self.assertFileNotFound(p.unlink) - - @unittest.skipUnless(hasattr(os, "link"), "os.link() is not present") - def test_hardlink_to(self): - P = self.cls(BASE) - target = P / 'fileA' - size = target.stat().st_size - # linking to another path. - link = P / 'dirA' / 'fileAA' - link.hardlink_to(target) - self.assertEqual(link.stat().st_size, size) - self.assertTrue(os.path.samefile(target, link)) - self.assertTrue(target.exists()) - # Linking to a str of a relative path. - link2 = P / 'dirA' / 'fileAAA' - target2 = rel_join('fileA') - link2.hardlink_to(target2) - self.assertEqual(os.stat(target2).st_size, size) - self.assertTrue(link2.exists()) - - @unittest.skipIf(hasattr(os, "link"), "os.link() is present") - def test_link_to_not_implemented(self): - P = self.cls(BASE) - p = P / 'fileA' - # linking to another path. - q = P / 'dirA' / 'fileAA' - with self.assertRaises(NotImplementedError): - q.hardlink_to(p) - - def test_rename(self): - P = self.cls(BASE) - p = P / 'fileA' - size = p.stat().st_size - # Renaming to another path. - q = P / 'dirA' / 'fileAA' - renamed_p = p.rename(q) - self.assertEqual(renamed_p, q) - self.assertEqual(q.stat().st_size, size) - self.assertFileNotFound(p.stat) - # Renaming to a str of a relative path. - r = rel_join('fileAAA') - renamed_q = q.rename(r) - self.assertEqual(renamed_q, self.cls(r)) - self.assertEqual(os.stat(r).st_size, size) - self.assertFileNotFound(q.stat) - - def test_replace(self): - P = self.cls(BASE) - p = P / 'fileA' - size = p.stat().st_size - # Replacing a non-existing path. - q = P / 'dirA' / 'fileAA' - replaced_p = p.replace(q) - self.assertEqual(replaced_p, q) - self.assertEqual(q.stat().st_size, size) - self.assertFileNotFound(p.stat) - # Replacing another (existing) path. - r = rel_join('dirB', 'fileB') - replaced_q = q.replace(r) - self.assertEqual(replaced_q, self.cls(r)) - self.assertEqual(os.stat(r).st_size, size) - self.assertFileNotFound(q.stat) - - @os_helper.skip_unless_symlink - def test_readlink(self): - P = self.cls(BASE) - self.assertEqual((P / 'linkA').readlink(), self.cls('fileA')) - self.assertEqual((P / 'brokenLink').readlink(), - self.cls('non-existing')) - self.assertEqual((P / 'linkB').readlink(), self.cls('dirB')) - with self.assertRaises(OSError): - (P / 'fileA').readlink() - - def test_touch_common(self): - P = self.cls(BASE) - p = P / 'newfileA' - self.assertFalse(p.exists()) - p.touch() - self.assertTrue(p.exists()) - st = p.stat() - old_mtime = st.st_mtime - old_mtime_ns = st.st_mtime_ns - # Rewind the mtime sufficiently far in the past to work around - # filesystem-specific timestamp granularity. - os.utime(str(p), (old_mtime - 10, old_mtime - 10)) - # The file mtime should be refreshed by calling touch() again. - p.touch() - st = p.stat() - self.assertGreaterEqual(st.st_mtime_ns, old_mtime_ns) - self.assertGreaterEqual(st.st_mtime, old_mtime) - # Now with exist_ok=False. - p = P / 'newfileB' - self.assertFalse(p.exists()) - p.touch(mode=0o700, exist_ok=False) - self.assertTrue(p.exists()) - self.assertRaises(OSError, p.touch, exist_ok=False) - - def test_touch_nochange(self): - P = self.cls(BASE) - p = P / 'fileA' - p.touch() - with p.open('rb') as f: - self.assertEqual(f.read().strip(), b"this is file A") - - def test_mkdir(self): - P = self.cls(BASE) - p = P / 'newdirA' - self.assertFalse(p.exists()) - p.mkdir() - self.assertTrue(p.exists()) - self.assertTrue(p.is_dir()) - with self.assertRaises(OSError) as cm: - p.mkdir() - self.assertEqual(cm.exception.errno, errno.EEXIST) - - def test_mkdir_parents(self): - # Creating a chain of directories. - p = self.cls(BASE, 'newdirB', 'newdirC') - self.assertFalse(p.exists()) - with self.assertRaises(OSError) as cm: - p.mkdir() - self.assertEqual(cm.exception.errno, errno.ENOENT) - p.mkdir(parents=True) - self.assertTrue(p.exists()) - self.assertTrue(p.is_dir()) - with self.assertRaises(OSError) as cm: - p.mkdir(parents=True) - self.assertEqual(cm.exception.errno, errno.EEXIST) - # Test `mode` arg. - mode = stat.S_IMODE(p.stat().st_mode) # Default mode. - p = self.cls(BASE, 'newdirD', 'newdirE') - p.mkdir(0o555, parents=True) - self.assertTrue(p.exists()) - self.assertTrue(p.is_dir()) - if os.name != 'nt': - # The directory's permissions follow the mode argument. - self.assertEqual(stat.S_IMODE(p.stat().st_mode), 0o7555 & mode) - # The parent's permissions follow the default process settings. - self.assertEqual(stat.S_IMODE(p.parent.stat().st_mode), mode) - - def test_mkdir_exist_ok(self): - p = self.cls(BASE, 'dirB') - st_ctime_first = p.stat().st_ctime - self.assertTrue(p.exists()) - self.assertTrue(p.is_dir()) - with self.assertRaises(FileExistsError) as cm: - p.mkdir() - self.assertEqual(cm.exception.errno, errno.EEXIST) - p.mkdir(exist_ok=True) - self.assertTrue(p.exists()) - self.assertEqual(p.stat().st_ctime, st_ctime_first) - - def test_mkdir_exist_ok_with_parent(self): - p = self.cls(BASE, 'dirC') - self.assertTrue(p.exists()) - with self.assertRaises(FileExistsError) as cm: - p.mkdir() - self.assertEqual(cm.exception.errno, errno.EEXIST) - p = p / 'newdirC' - p.mkdir(parents=True) - st_ctime_first = p.stat().st_ctime - self.assertTrue(p.exists()) - with self.assertRaises(FileExistsError) as cm: - p.mkdir(parents=True) - self.assertEqual(cm.exception.errno, errno.EEXIST) - p.mkdir(parents=True, exist_ok=True) - self.assertTrue(p.exists()) - self.assertEqual(p.stat().st_ctime, st_ctime_first) - - @unittest.skipIf(is_emscripten, "FS root cannot be modified on Emscripten.") - def test_mkdir_exist_ok_root(self): - # Issue #25803: A drive root could raise PermissionError on Windows. - self.cls('/').resolve().mkdir(exist_ok=True) - self.cls('/').resolve().mkdir(parents=True, exist_ok=True) - - @only_nt # XXX: not sure how to test this on POSIX. - def test_mkdir_with_unknown_drive(self): - for d in 'ZYXWVUTSRQPONMLKJIHGFEDCBA': - p = self.cls(d + ':\\') - if not p.is_dir(): - break - else: - self.skipTest("cannot find a drive that doesn't exist") - with self.assertRaises(OSError): - (p / 'child' / 'path').mkdir(parents=True) - - def test_mkdir_with_child_file(self): - p = self.cls(BASE, 'dirB', 'fileB') - self.assertTrue(p.exists()) - # An exception is raised when the last path component is an existing - # regular file, regardless of whether exist_ok is true or not. - with self.assertRaises(FileExistsError) as cm: - p.mkdir(parents=True) - self.assertEqual(cm.exception.errno, errno.EEXIST) - with self.assertRaises(FileExistsError) as cm: - p.mkdir(parents=True, exist_ok=True) - self.assertEqual(cm.exception.errno, errno.EEXIST) - - def test_mkdir_no_parents_file(self): - p = self.cls(BASE, 'fileA') - self.assertTrue(p.exists()) - # An exception is raised when the last path component is an existing - # regular file, regardless of whether exist_ok is true or not. - with self.assertRaises(FileExistsError) as cm: - p.mkdir() - self.assertEqual(cm.exception.errno, errno.EEXIST) - with self.assertRaises(FileExistsError) as cm: - p.mkdir(exist_ok=True) - self.assertEqual(cm.exception.errno, errno.EEXIST) - - def test_mkdir_concurrent_parent_creation(self): - for pattern_num in range(32): - p = self.cls(BASE, 'dirCPC%d' % pattern_num) - self.assertFalse(p.exists()) - - real_mkdir = os.mkdir - def my_mkdir(path, mode=0o777): - path = str(path) - # Emulate another process that would create the directory - # just before we try to create it ourselves. We do it - # in all possible pattern combinations, assuming that this - # function is called at most 5 times (dirCPC/dir1/dir2, - # dirCPC/dir1, dirCPC, dirCPC/dir1, dirCPC/dir1/dir2). - if pattern.pop(): - real_mkdir(path, mode) # From another process. - concurrently_created.add(path) - real_mkdir(path, mode) # Our real call. - - pattern = [bool(pattern_num & (1 << n)) for n in range(5)] - concurrently_created = set() - p12 = p / 'dir1' / 'dir2' - try: - with mock.patch("os.mkdir", my_mkdir): - p12.mkdir(parents=True, exist_ok=False) - except FileExistsError: - self.assertIn(str(p12), concurrently_created) - else: - self.assertNotIn(str(p12), concurrently_created) - self.assertTrue(p.exists()) - - @os_helper.skip_unless_symlink - def test_symlink_to(self): - P = self.cls(BASE) - target = P / 'fileA' - # Symlinking a path target. - link = P / 'dirA' / 'linkAA' - link.symlink_to(target) - self.assertEqual(link.stat(), target.stat()) - self.assertNotEqual(link.lstat(), target.stat()) - # Symlinking a str target. - link = P / 'dirA' / 'linkAAA' - link.symlink_to(str(target)) - self.assertEqual(link.stat(), target.stat()) - self.assertNotEqual(link.lstat(), target.stat()) - self.assertFalse(link.is_dir()) - # Symlinking to a directory. - target = P / 'dirB' - link = P / 'dirA' / 'linkAAAA' - link.symlink_to(target, target_is_directory=True) - self.assertEqual(link.stat(), target.stat()) - self.assertNotEqual(link.lstat(), target.stat()) - self.assertTrue(link.is_dir()) - self.assertTrue(list(link.iterdir())) - - def test_is_dir(self): - P = self.cls(BASE) - self.assertTrue((P / 'dirA').is_dir()) - self.assertFalse((P / 'fileA').is_dir()) - self.assertFalse((P / 'non-existing').is_dir()) - self.assertFalse((P / 'fileA' / 'bah').is_dir()) - if os_helper.can_symlink(): - self.assertFalse((P / 'linkA').is_dir()) - self.assertTrue((P / 'linkB').is_dir()) - self.assertFalse((P/ 'brokenLink').is_dir(), False) - self.assertIs((P / 'dirA\udfff').is_dir(), False) - self.assertIs((P / 'dirA\x00').is_dir(), False) - - def test_is_file(self): - P = self.cls(BASE) - self.assertTrue((P / 'fileA').is_file()) - self.assertFalse((P / 'dirA').is_file()) - self.assertFalse((P / 'non-existing').is_file()) - self.assertFalse((P / 'fileA' / 'bah').is_file()) - if os_helper.can_symlink(): - self.assertTrue((P / 'linkA').is_file()) - self.assertFalse((P / 'linkB').is_file()) - self.assertFalse((P/ 'brokenLink').is_file()) - self.assertIs((P / 'fileA\udfff').is_file(), False) - self.assertIs((P / 'fileA\x00').is_file(), False) - - def test_is_mount(self): - P = self.cls(BASE) - if os.name == 'nt': - R = self.cls('c:\\') - else: - R = self.cls('/') - self.assertFalse((P / 'fileA').is_mount()) - self.assertFalse((P / 'dirA').is_mount()) - self.assertFalse((P / 'non-existing').is_mount()) - self.assertFalse((P / 'fileA' / 'bah').is_mount()) - self.assertTrue(R.is_mount()) - if os_helper.can_symlink(): - self.assertFalse((P / 'linkA').is_mount()) - self.assertIs((R / '\udfff').is_mount(), False) - - def test_is_symlink(self): - P = self.cls(BASE) - self.assertFalse((P / 'fileA').is_symlink()) - self.assertFalse((P / 'dirA').is_symlink()) - self.assertFalse((P / 'non-existing').is_symlink()) - self.assertFalse((P / 'fileA' / 'bah').is_symlink()) - if os_helper.can_symlink(): - self.assertTrue((P / 'linkA').is_symlink()) - self.assertTrue((P / 'linkB').is_symlink()) - self.assertTrue((P/ 'brokenLink').is_symlink()) - self.assertIs((P / 'fileA\udfff').is_file(), False) - self.assertIs((P / 'fileA\x00').is_file(), False) - if os_helper.can_symlink(): - self.assertIs((P / 'linkA\udfff').is_file(), False) - self.assertIs((P / 'linkA\x00').is_file(), False) - - def test_is_junction(self): - P = self.cls(BASE) - - with mock.patch.object(P._flavour, 'isjunction'): - self.assertEqual(P.is_junction(), P._flavour.isjunction.return_value) - P._flavour.isjunction.assert_called_once_with(P) - - def test_is_fifo_false(self): - P = self.cls(BASE) - self.assertFalse((P / 'fileA').is_fifo()) - self.assertFalse((P / 'dirA').is_fifo()) - self.assertFalse((P / 'non-existing').is_fifo()) - self.assertFalse((P / 'fileA' / 'bah').is_fifo()) - self.assertIs((P / 'fileA\udfff').is_fifo(), False) - self.assertIs((P / 'fileA\x00').is_fifo(), False) - - @unittest.skipUnless(hasattr(os, "mkfifo"), "os.mkfifo() required") - @unittest.skipIf(sys.platform == "vxworks", - "fifo requires special path on VxWorks") - def test_is_fifo_true(self): - P = self.cls(BASE, 'myfifo') - try: - os.mkfifo(str(P)) - except PermissionError as e: - self.skipTest('os.mkfifo(): %s' % e) - self.assertTrue(P.is_fifo()) - self.assertFalse(P.is_socket()) - self.assertFalse(P.is_file()) - self.assertIs(self.cls(BASE, 'myfifo\udfff').is_fifo(), False) - self.assertIs(self.cls(BASE, 'myfifo\x00').is_fifo(), False) - - def test_is_socket_false(self): - P = self.cls(BASE) - self.assertFalse((P / 'fileA').is_socket()) - self.assertFalse((P / 'dirA').is_socket()) - self.assertFalse((P / 'non-existing').is_socket()) - self.assertFalse((P / 'fileA' / 'bah').is_socket()) - self.assertIs((P / 'fileA\udfff').is_socket(), False) - self.assertIs((P / 'fileA\x00').is_socket(), False) - - @unittest.skipUnless(hasattr(socket, "AF_UNIX"), "Unix sockets required") - @unittest.skipIf( - is_emscripten, "Unix sockets are not implemented on Emscripten." - ) - @unittest.skipIf( - is_wasi, "Cannot create socket on WASI." - ) - def test_is_socket_true(self): - P = self.cls(BASE, 'mysock') - sock = socket.socket(socket.AF_UNIX, socket.SOCK_STREAM) - self.addCleanup(sock.close) - try: - sock.bind(str(P)) - except OSError as e: - if (isinstance(e, PermissionError) or - "AF_UNIX path too long" in str(e)): - self.skipTest("cannot bind Unix socket: " + str(e)) - self.assertTrue(P.is_socket()) - self.assertFalse(P.is_fifo()) - self.assertFalse(P.is_file()) - self.assertIs(self.cls(BASE, 'mysock\udfff').is_socket(), False) - self.assertIs(self.cls(BASE, 'mysock\x00').is_socket(), False) - - def test_is_block_device_false(self): - P = self.cls(BASE) - self.assertFalse((P / 'fileA').is_block_device()) - self.assertFalse((P / 'dirA').is_block_device()) - self.assertFalse((P / 'non-existing').is_block_device()) - self.assertFalse((P / 'fileA' / 'bah').is_block_device()) - self.assertIs((P / 'fileA\udfff').is_block_device(), False) - self.assertIs((P / 'fileA\x00').is_block_device(), False) - - def test_is_char_device_false(self): - P = self.cls(BASE) - self.assertFalse((P / 'fileA').is_char_device()) - self.assertFalse((P / 'dirA').is_char_device()) - self.assertFalse((P / 'non-existing').is_char_device()) - self.assertFalse((P / 'fileA' / 'bah').is_char_device()) - self.assertIs((P / 'fileA\udfff').is_char_device(), False) - self.assertIs((P / 'fileA\x00').is_char_device(), False) - - def test_is_char_device_true(self): - # os.devnull should generally be a char device. - P = self.cls(os.devnull) - if not P.exists(): - self.skipTest("null device required") - self.assertTrue(P.is_char_device()) - self.assertFalse(P.is_block_device()) - self.assertFalse(P.is_file()) - self.assertIs(self.cls(f'{os.devnull}\udfff').is_char_device(), False) - self.assertIs(self.cls(f'{os.devnull}\x00').is_char_device(), False) - - def test_pickling_common(self): - p = self.cls(BASE, 'fileA') - for proto in range(0, pickle.HIGHEST_PROTOCOL + 1): - dumped = pickle.dumps(p, proto) - pp = pickle.loads(dumped) - self.assertEqual(pp.stat(), p.stat()) - - def test_parts_interning(self): - P = self.cls - p = P('/usr/bin/foo') - q = P('/usr/local/bin') - # 'usr' - self.assertIs(p.parts[1], q.parts[1]) - # 'bin' - self.assertIs(p.parts[2], q.parts[3]) - - def _check_complex_symlinks(self, link0_target): - # Test solving a non-looping chain of symlinks (issue #19887). - P = self.cls(BASE) - self.dirlink(os.path.join('link0', 'link0'), join('link1')) - self.dirlink(os.path.join('link1', 'link1'), join('link2')) - self.dirlink(os.path.join('link2', 'link2'), join('link3')) - self.dirlink(link0_target, join('link0')) - - # Resolve absolute paths. - p = (P / 'link0').resolve() - self.assertEqual(p, P) - self.assertEqualNormCase(str(p), BASE) - p = (P / 'link1').resolve() - self.assertEqual(p, P) - self.assertEqualNormCase(str(p), BASE) - p = (P / 'link2').resolve() - self.assertEqual(p, P) - self.assertEqualNormCase(str(p), BASE) - p = (P / 'link3').resolve() - self.assertEqual(p, P) - self.assertEqualNormCase(str(p), BASE) - - # Resolve relative paths. - old_path = os.getcwd() - os.chdir(BASE) - try: - p = self.cls('link0').resolve() - self.assertEqual(p, P) - self.assertEqualNormCase(str(p), BASE) - p = self.cls('link1').resolve() - self.assertEqual(p, P) - self.assertEqualNormCase(str(p), BASE) - p = self.cls('link2').resolve() - self.assertEqual(p, P) - self.assertEqualNormCase(str(p), BASE) - p = self.cls('link3').resolve() - self.assertEqual(p, P) - self.assertEqualNormCase(str(p), BASE) - finally: - os.chdir(old_path) - - @os_helper.skip_unless_symlink - def test_complex_symlinks_absolute(self): - self._check_complex_symlinks(BASE) - - @os_helper.skip_unless_symlink - def test_complex_symlinks_relative(self): - self._check_complex_symlinks('.') - - @os_helper.skip_unless_symlink - def test_complex_symlinks_relative_dot_dot(self): - self._check_complex_symlinks(os.path.join('dirA', '..')) - - def test_passing_kwargs_deprecated(self): - with self.assertWarns(DeprecationWarning): - self.cls(foo="bar") - - -class WalkTests(unittest.TestCase): - - def setUp(self): - self.addCleanup(os_helper.rmtree, os_helper.TESTFN) - - # Build: - # TESTFN/ - # TEST1/ a file kid and two directory kids - # tmp1 - # SUB1/ a file kid and a directory kid - # tmp2 - # SUB11/ no kids - # SUB2/ a file kid and a dirsymlink kid - # tmp3 - # SUB21/ not readable - # tmp5 - # link/ a symlink to TEST2 - # broken_link - # broken_link2 - # broken_link3 - # TEST2/ - # tmp4 a lone file - self.walk_path = pathlib.Path(os_helper.TESTFN, "TEST1") - self.sub1_path = self.walk_path / "SUB1" - self.sub11_path = self.sub1_path / "SUB11" - self.sub2_path = self.walk_path / "SUB2" - sub21_path= self.sub2_path / "SUB21" - tmp1_path = self.walk_path / "tmp1" - tmp2_path = self.sub1_path / "tmp2" - tmp3_path = self.sub2_path / "tmp3" - tmp5_path = sub21_path / "tmp3" - self.link_path = self.sub2_path / "link" - t2_path = pathlib.Path(os_helper.TESTFN, "TEST2") - tmp4_path = pathlib.Path(os_helper.TESTFN, "TEST2", "tmp4") - broken_link_path = self.sub2_path / "broken_link" - broken_link2_path = self.sub2_path / "broken_link2" - broken_link3_path = self.sub2_path / "broken_link3" - - os.makedirs(self.sub11_path) - os.makedirs(self.sub2_path) - os.makedirs(sub21_path) - os.makedirs(t2_path) - - for path in tmp1_path, tmp2_path, tmp3_path, tmp4_path, tmp5_path: - with open(path, "x", encoding='utf-8') as f: - f.write(f"I'm {path} and proud of it. Blame test_pathlib.\n") - - if os_helper.can_symlink(): - os.symlink(os.path.abspath(t2_path), self.link_path) - os.symlink('broken', broken_link_path, True) - os.symlink(pathlib.Path('tmp3', 'broken'), broken_link2_path, True) - os.symlink(pathlib.Path('SUB21', 'tmp5'), broken_link3_path, True) - self.sub2_tree = (self.sub2_path, ["SUB21"], - ["broken_link", "broken_link2", "broken_link3", - "link", "tmp3"]) - else: - self.sub2_tree = (self.sub2_path, ["SUB21"], ["tmp3"]) - - if not is_emscripten: - # Emscripten fails with inaccessible directories. - os.chmod(sub21_path, 0) - try: - os.listdir(sub21_path) - except PermissionError: - self.addCleanup(os.chmod, sub21_path, stat.S_IRWXU) - else: - os.chmod(sub21_path, stat.S_IRWXU) - os.unlink(tmp5_path) - os.rmdir(sub21_path) - del self.sub2_tree[1][:1] - - def test_walk_topdown(self): - walker = self.walk_path.walk() - entry = next(walker) - entry[1].sort() # Ensure we visit SUB1 before SUB2 - self.assertEqual(entry, (self.walk_path, ["SUB1", "SUB2"], ["tmp1"])) - entry = next(walker) - self.assertEqual(entry, (self.sub1_path, ["SUB11"], ["tmp2"])) - entry = next(walker) - self.assertEqual(entry, (self.sub11_path, [], [])) - entry = next(walker) - entry[1].sort() - entry[2].sort() - self.assertEqual(entry, self.sub2_tree) - with self.assertRaises(StopIteration): - next(walker) - - def test_walk_prune(self, walk_path=None): - if walk_path is None: - walk_path = self.walk_path - # Prune the search. - all = [] - for root, dirs, files in walk_path.walk(): - all.append((root, dirs, files)) - if 'SUB1' in dirs: - # Note that this also mutates the dirs we appended to all! - dirs.remove('SUB1') - - self.assertEqual(len(all), 2) - self.assertEqual(all[0], (self.walk_path, ["SUB2"], ["tmp1"])) - - all[1][-1].sort() - all[1][1].sort() - self.assertEqual(all[1], self.sub2_tree) - - def test_file_like_path(self): - self.test_walk_prune(FakePath(self.walk_path).__fspath__()) - - def test_walk_bottom_up(self): - seen_testfn = seen_sub1 = seen_sub11 = seen_sub2 = False - for path, dirnames, filenames in self.walk_path.walk(top_down=False): - if path == self.walk_path: - self.assertFalse(seen_testfn) - self.assertTrue(seen_sub1) - self.assertTrue(seen_sub2) - self.assertEqual(sorted(dirnames), ["SUB1", "SUB2"]) - self.assertEqual(filenames, ["tmp1"]) - seen_testfn = True - elif path == self.sub1_path: - self.assertFalse(seen_testfn) - self.assertFalse(seen_sub1) - self.assertTrue(seen_sub11) - self.assertEqual(dirnames, ["SUB11"]) - self.assertEqual(filenames, ["tmp2"]) - seen_sub1 = True - elif path == self.sub11_path: - self.assertFalse(seen_sub1) - self.assertFalse(seen_sub11) - self.assertEqual(dirnames, []) - self.assertEqual(filenames, []) - seen_sub11 = True - elif path == self.sub2_path: - self.assertFalse(seen_testfn) - self.assertFalse(seen_sub2) - self.assertEqual(sorted(dirnames), sorted(self.sub2_tree[1])) - self.assertEqual(sorted(filenames), sorted(self.sub2_tree[2])) - seen_sub2 = True - else: - raise AssertionError(f"Unexpected path: {path}") - self.assertTrue(seen_testfn) - - @os_helper.skip_unless_symlink - def test_walk_follow_symlinks(self): - walk_it = self.walk_path.walk(follow_symlinks=True) - for root, dirs, files in walk_it: - if root == self.link_path: - self.assertEqual(dirs, []) - self.assertEqual(files, ["tmp4"]) - break - else: - self.fail("Didn't follow symlink with follow_symlinks=True") - - @os_helper.skip_unless_symlink - def test_walk_symlink_location(self): - # Tests whether symlinks end up in filenames or dirnames depending - # on the `follow_symlinks` argument. - walk_it = self.walk_path.walk(follow_symlinks=False) - for root, dirs, files in walk_it: - if root == self.sub2_path: - self.assertIn("link", files) - break - else: - self.fail("symlink not found") - - walk_it = self.walk_path.walk(follow_symlinks=True) - for root, dirs, files in walk_it: - if root == self.sub2_path: - self.assertIn("link", dirs) - break - - def test_walk_bad_dir(self): - errors = [] - walk_it = self.walk_path.walk(on_error=errors.append) - root, dirs, files = next(walk_it) - self.assertEqual(errors, []) - dir1 = 'SUB1' - path1 = root / dir1 - path1new = (root / dir1).with_suffix(".new") - path1.rename(path1new) - try: - roots = [r for r, _, _ in walk_it] - self.assertTrue(errors) - self.assertNotIn(path1, roots) - self.assertNotIn(path1new, roots) - for dir2 in dirs: - if dir2 != dir1: - self.assertIn(root / dir2, roots) - finally: - path1new.rename(path1) - - def test_walk_many_open_files(self): - depth = 30 - base = pathlib.Path(os_helper.TESTFN, 'deep') - path = pathlib.Path(base, *(['d']*depth)) - path.mkdir(parents=True) - - iters = [base.walk(top_down=False) for _ in range(100)] - for i in range(depth + 1): - expected = (path, ['d'] if i else [], []) - for it in iters: - self.assertEqual(next(it), expected) - path = path.parent - - iters = [base.walk(top_down=True) for _ in range(100)] - path = base - for i in range(depth + 1): - expected = (path, ['d'] if i < depth else [], []) - for it in iters: - self.assertEqual(next(it), expected) - path = path / 'd' - - def test_walk_above_recursion_limit(self): - recursion_limit = 40 - # directory_depth > recursion_limit - directory_depth = recursion_limit + 10 - base = pathlib.Path(os_helper.TESTFN, 'deep') - path = pathlib.Path(base, *(['d'] * directory_depth)) - path.mkdir(parents=True) - - with set_recursion_limit(recursion_limit): - list(base.walk()) - list(base.walk(top_down=False)) - - -class PathTest(_BasePathTest, unittest.TestCase): - cls = pathlib.Path - - def test_concrete_class(self): - p = self.cls('a') - self.assertIs(type(p), - pathlib.WindowsPath if os.name == 'nt' else pathlib.PosixPath) - - def test_unsupported_flavour(self): - if os.name == 'nt': - self.assertRaises(NotImplementedError, pathlib.PosixPath) - else: - self.assertRaises(NotImplementedError, pathlib.WindowsPath) - - def test_glob_empty_pattern(self): - p = self.cls() - with self.assertRaisesRegex(ValueError, 'Unacceptable pattern'): - list(p.glob('')) - - -@only_posix -class PosixPathTest(_BasePathTest, unittest.TestCase): - cls = pathlib.PosixPath - - def test_absolute(self): - P = self.cls - self.assertEqual(str(P('/').absolute()), '/') - self.assertEqual(str(P('/a').absolute()), '/a') - self.assertEqual(str(P('/a/b').absolute()), '/a/b') - - # '//'-prefixed absolute path (supported by POSIX). - self.assertEqual(str(P('//').absolute()), '//') - self.assertEqual(str(P('//a').absolute()), '//a') - self.assertEqual(str(P('//a/b').absolute()), '//a/b') - - def _check_symlink_loop(self, *args, strict=True): - path = self.cls(*args) - with self.assertRaises(RuntimeError): - print(path.resolve(strict)) - - @unittest.skipIf( - is_emscripten or is_wasi, - "umask is not implemented on Emscripten/WASI." - ) - def test_open_mode(self): - old_mask = os.umask(0) - self.addCleanup(os.umask, old_mask) - p = self.cls(BASE) - with (p / 'new_file').open('wb'): - pass - st = os.stat(join('new_file')) - self.assertEqual(stat.S_IMODE(st.st_mode), 0o666) - os.umask(0o022) - with (p / 'other_new_file').open('wb'): - pass - st = os.stat(join('other_new_file')) - self.assertEqual(stat.S_IMODE(st.st_mode), 0o644) - - def test_resolve_root(self): - current_directory = os.getcwd() - try: - os.chdir('/') - p = self.cls('spam') - self.assertEqual(str(p.resolve()), '/spam') - finally: - os.chdir(current_directory) - - @unittest.skipIf( - is_emscripten or is_wasi, - "umask is not implemented on Emscripten/WASI." - ) - def test_touch_mode(self): - old_mask = os.umask(0) - self.addCleanup(os.umask, old_mask) - p = self.cls(BASE) - (p / 'new_file').touch() - st = os.stat(join('new_file')) - self.assertEqual(stat.S_IMODE(st.st_mode), 0o666) - os.umask(0o022) - (p / 'other_new_file').touch() - st = os.stat(join('other_new_file')) - self.assertEqual(stat.S_IMODE(st.st_mode), 0o644) - (p / 'masked_new_file').touch(mode=0o750) - st = os.stat(join('masked_new_file')) - self.assertEqual(stat.S_IMODE(st.st_mode), 0o750) - - @os_helper.skip_unless_symlink - def test_resolve_loop(self): - # Loops with relative symlinks. - os.symlink('linkX/inside', join('linkX')) - self._check_symlink_loop(BASE, 'linkX') - os.symlink('linkY', join('linkY')) - self._check_symlink_loop(BASE, 'linkY') - os.symlink('linkZ/../linkZ', join('linkZ')) - self._check_symlink_loop(BASE, 'linkZ') - # Non-strict - self._check_symlink_loop(BASE, 'linkZ', 'foo', strict=False) - # Loops with absolute symlinks. - os.symlink(join('linkU/inside'), join('linkU')) - self._check_symlink_loop(BASE, 'linkU') - os.symlink(join('linkV'), join('linkV')) - self._check_symlink_loop(BASE, 'linkV') - os.symlink(join('linkW/../linkW'), join('linkW')) - self._check_symlink_loop(BASE, 'linkW') - # Non-strict - self._check_symlink_loop(BASE, 'linkW', 'foo', strict=False) - - def test_glob(self): - P = self.cls - p = P(BASE) - given = set(p.glob("FILEa")) - expect = set() if not os_helper.fs_is_case_insensitive(BASE) else given - self.assertEqual(given, expect) - self.assertEqual(set(p.glob("FILEa*")), set()) - - def test_rglob(self): - P = self.cls - p = P(BASE, "dirC") - given = set(p.rglob("FILEd")) - expect = set() if not os_helper.fs_is_case_insensitive(BASE) else given - self.assertEqual(given, expect) - self.assertEqual(set(p.rglob("FILEd*")), set()) - - @unittest.skipUnless(hasattr(pwd, 'getpwall'), - 'pwd module does not expose getpwall()') - @unittest.skipIf(sys.platform == "vxworks", - "no home directory on VxWorks") - def test_expanduser(self): - P = self.cls - import_helper.import_module('pwd') - import pwd - pwdent = pwd.getpwuid(os.getuid()) - username = pwdent.pw_name - userhome = pwdent.pw_dir.rstrip('/') or '/' - # Find arbitrary different user (if exists). - for pwdent in pwd.getpwall(): - othername = pwdent.pw_name - otherhome = pwdent.pw_dir.rstrip('/') - if othername != username and otherhome: - break - else: - othername = username - otherhome = userhome - - fakename = 'fakeuser' - # This user can theoretically exist on a test runner. Create unique name: - try: - while pwd.getpwnam(fakename): - fakename += '1' - except KeyError: - pass # Non-existent name found - - p1 = P('~/Documents') - p2 = P(f'~{username}/Documents') - p3 = P(f'~{othername}/Documents') - p4 = P(f'../~{username}/Documents') - p5 = P(f'/~{username}/Documents') - p6 = P('') - p7 = P(f'~{fakename}/Documents') - - with os_helper.EnvironmentVarGuard() as env: - env.pop('HOME', None) - - self.assertEqual(p1.expanduser(), P(userhome) / 'Documents') - self.assertEqual(p2.expanduser(), P(userhome) / 'Documents') - self.assertEqual(p3.expanduser(), P(otherhome) / 'Documents') - self.assertEqual(p4.expanduser(), p4) - self.assertEqual(p5.expanduser(), p5) - self.assertEqual(p6.expanduser(), p6) - self.assertRaises(RuntimeError, p7.expanduser) - - env['HOME'] = '/tmp' - self.assertEqual(p1.expanduser(), P('/tmp/Documents')) - self.assertEqual(p2.expanduser(), P(userhome) / 'Documents') - self.assertEqual(p3.expanduser(), P(otherhome) / 'Documents') - self.assertEqual(p4.expanduser(), p4) - self.assertEqual(p5.expanduser(), p5) - self.assertEqual(p6.expanduser(), p6) - self.assertRaises(RuntimeError, p7.expanduser) - - @unittest.skipIf(sys.platform != "darwin", - "Bad file descriptor in /dev/fd affects only macOS") - def test_handling_bad_descriptor(self): - try: - file_descriptors = list(pathlib.Path('/dev/fd').rglob("*"))[3:] - if not file_descriptors: - self.skipTest("no file descriptors - issue was not reproduced") - # Checking all file descriptors because there is no guarantee - # which one will fail. - for f in file_descriptors: - f.exists() - f.is_dir() - f.is_file() - f.is_symlink() - f.is_block_device() - f.is_char_device() - f.is_fifo() - f.is_socket() - except OSError as e: - if e.errno == errno.EBADF: - self.fail("Bad file descriptor not handled.") - raise - - -@only_nt -class WindowsPathTest(_BasePathTest, unittest.TestCase): - cls = pathlib.WindowsPath - - def test_absolute(self): - P = self.cls - - # Simple absolute paths. - self.assertEqual(str(P('c:\\').absolute()), 'c:\\') - self.assertEqual(str(P('c:\\a').absolute()), 'c:\\a') - self.assertEqual(str(P('c:\\a\\b').absolute()), 'c:\\a\\b') - - # UNC absolute paths. - share = '\\\\server\\share\\' - self.assertEqual(str(P(share).absolute()), share) - self.assertEqual(str(P(share + 'a').absolute()), share + 'a') - self.assertEqual(str(P(share + 'a\\b').absolute()), share + 'a\\b') - - # UNC relative paths. - with mock.patch("os.getcwd") as getcwd: - getcwd.return_value = share - - self.assertEqual(str(P().absolute()), share) - self.assertEqual(str(P('.').absolute()), share) - self.assertEqual(str(P('a').absolute()), os.path.join(share, 'a')) - self.assertEqual(str(P('a', 'b', 'c').absolute()), - os.path.join(share, 'a', 'b', 'c')) - - drive = os.path.splitdrive(BASE)[0] - with os_helper.change_cwd(BASE): - # Relative path with root - self.assertEqual(str(P('\\').absolute()), drive + '\\') - self.assertEqual(str(P('\\foo').absolute()), drive + '\\foo') - - # Relative path on current drive - self.assertEqual(str(P(drive).absolute()), BASE) - self.assertEqual(str(P(drive + 'foo').absolute()), os.path.join(BASE, 'foo')) - - with os_helper.subst_drive(BASE) as other_drive: - # Set the working directory on the substitute drive - saved_cwd = os.getcwd() - other_cwd = f'{other_drive}\\dirA' - os.chdir(other_cwd) - os.chdir(saved_cwd) - - # Relative path on another drive - self.assertEqual(str(P(other_drive).absolute()), other_cwd) - self.assertEqual(str(P(other_drive + 'foo').absolute()), other_cwd + '\\foo') - - def test_glob(self): - P = self.cls - p = P(BASE) - self.assertEqual(set(p.glob("FILEa")), { P(BASE, "fileA") }) - self.assertEqual(set(p.glob("*a\\")), { P(BASE, "dirA") }) - self.assertEqual(set(p.glob("F*a")), { P(BASE, "fileA") }) - self.assertEqual(set(map(str, p.glob("FILEa"))), {f"{p}\\fileA"}) - self.assertEqual(set(map(str, p.glob("F*a"))), {f"{p}\\fileA"}) - - def test_rglob(self): - P = self.cls - p = P(BASE, "dirC") - self.assertEqual(set(p.rglob("FILEd")), { P(BASE, "dirC/dirD/fileD") }) - self.assertEqual(set(p.rglob("*\\")), { P(BASE, "dirC/dirD") }) - self.assertEqual(set(map(str, p.rglob("FILEd"))), {f"{p}\\dirD\\fileD"}) - - def test_expanduser(self): - P = self.cls - with os_helper.EnvironmentVarGuard() as env: - env.pop('HOME', None) - env.pop('USERPROFILE', None) - env.pop('HOMEPATH', None) - env.pop('HOMEDRIVE', None) - env['USERNAME'] = 'alice' - - # test that the path returns unchanged - p1 = P('~/My Documents') - p2 = P('~alice/My Documents') - p3 = P('~bob/My Documents') - p4 = P('/~/My Documents') - p5 = P('d:~/My Documents') - p6 = P('') - self.assertRaises(RuntimeError, p1.expanduser) - self.assertRaises(RuntimeError, p2.expanduser) - self.assertRaises(RuntimeError, p3.expanduser) - self.assertEqual(p4.expanduser(), p4) - self.assertEqual(p5.expanduser(), p5) - self.assertEqual(p6.expanduser(), p6) - - def check(): - env.pop('USERNAME', None) - self.assertEqual(p1.expanduser(), - P('C:/Users/alice/My Documents')) - self.assertRaises(RuntimeError, p2.expanduser) - env['USERNAME'] = 'alice' - self.assertEqual(p2.expanduser(), - P('C:/Users/alice/My Documents')) - self.assertEqual(p3.expanduser(), - P('C:/Users/bob/My Documents')) - self.assertEqual(p4.expanduser(), p4) - self.assertEqual(p5.expanduser(), p5) - self.assertEqual(p6.expanduser(), p6) - - env['HOMEPATH'] = 'C:\\Users\\alice' - check() - - env['HOMEDRIVE'] = 'C:\\' - env['HOMEPATH'] = 'Users\\alice' - check() - - env.pop('HOMEDRIVE', None) - env.pop('HOMEPATH', None) - env['USERPROFILE'] = 'C:\\Users\\alice' - check() - - # bpo-38883: ignore `HOME` when set on windows - env['HOME'] = 'C:\\Users\\eve' - check() - - -class PurePathSubclassTest(_BasePurePathTest, unittest.TestCase): - class cls(pathlib.PurePath): - pass - - # repr() roundtripping is not supported in custom subclass. - test_repr_roundtrips = None - - -class PathSubclassTest(_BasePathTest, unittest.TestCase): - class cls(pathlib.Path): - pass - - # repr() roundtripping is not supported in custom subclass. - test_repr_roundtrips = None - - -class CompatiblePathTest(unittest.TestCase): - """ - Test that a type can be made compatible with PurePath - derivatives by implementing division operator overloads. - """ - - class CompatPath: - """ - Minimum viable class to test PurePath compatibility. - Simply uses the division operator to join a given - string and the string value of another object with - a forward slash. - """ - def __init__(self, string): - self.string = string - - def __truediv__(self, other): - return type(self)(f"{self.string}/{other}") - - def __rtruediv__(self, other): - return type(self)(f"{other}/{self.string}") - - def test_truediv(self): - result = pathlib.PurePath("test") / self.CompatPath("right") - self.assertIsInstance(result, self.CompatPath) - self.assertEqual(result.string, "test/right") - - with self.assertRaises(TypeError): - # Verify improper operations still raise a TypeError - pathlib.PurePath("test") / 10 - - def test_rtruediv(self): - result = self.CompatPath("left") / pathlib.PurePath("test") - self.assertIsInstance(result, self.CompatPath) - self.assertEqual(result.string, "left/test") - - with self.assertRaises(TypeError): - # Verify improper operations still raise a TypeError - 10 / pathlib.PurePath("test") - - -if __name__ == "__main__": - unittest.main() diff --git a/Lib/test/test_pathlib/__init__.py b/Lib/test/test_pathlib/__init__.py new file mode 100644 index 00000000000..4b16ecc3115 --- /dev/null +++ b/Lib/test/test_pathlib/__init__.py @@ -0,0 +1,5 @@ +import os +from test.support import load_package_tests + +def load_tests(*args): + return load_package_tests(os.path.dirname(__file__), *args) diff --git a/Lib/test/test_pathlib/test_pathlib.py b/Lib/test/test_pathlib/test_pathlib.py new file mode 100644 index 00000000000..6cbc15d7675 --- /dev/null +++ b/Lib/test/test_pathlib/test_pathlib.py @@ -0,0 +1,1660 @@ +# TODO: RUSTPYTHON +# Has not been tested with Windows style paths +import io +import os +import sys +import errno +import ntpath +import pathlib +import pickle +import posixpath +import socket +import stat +import tempfile +import unittest +from unittest import mock +from urllib.request import pathname2url + +from test.support import import_helper +from test.support import is_emscripten, is_wasi +from test.support import infinite_recursion +from test.support import os_helper +from test.support.os_helper import TESTFN, FS_NONASCII, FakePath +from test.test_pathlib import test_pathlib_abc +from test.test_pathlib.test_pathlib_abc import needs_posix, needs_windows, needs_symlinks + +try: + import grp, pwd +except ImportError: + grp = pwd = None + + +root_in_posix = False +if hasattr(os, 'geteuid'): + root_in_posix = (os.geteuid() == 0) + +# +# Tests for the pure classes. +# + +class PurePathTest(test_pathlib_abc.DummyPurePathTest): + cls = pathlib.PurePath + + # Make sure any symbolic links in the base test path are resolved. + base = os.path.realpath(TESTFN) + + # Keys are canonical paths, values are list of tuples of arguments + # supposed to produce equal paths. + equivalences = { + 'a/b': [ + ('a', 'b'), ('a/', 'b'), ('a', 'b/'), ('a/', 'b/'), + ('a/b/',), ('a//b',), ('a//b//',), + # Empty components get removed. + ('', 'a', 'b'), ('a', '', 'b'), ('a', 'b', ''), + ], + '/b/c/d': [ + ('a', '/b/c', 'd'), ('/a', '/b/c', 'd'), + # Empty components get removed. + ('/', 'b', '', 'c/d'), ('/', '', 'b/c/d'), ('', '/b/c/d'), + ], + } + + def test_concrete_class(self): + if self.cls is pathlib.PurePath: + expected = pathlib.PureWindowsPath if os.name == 'nt' else pathlib.PurePosixPath + else: + expected = self.cls + p = self.cls('a') + self.assertIs(type(p), expected) + + def test_concrete_parser(self): + if self.cls is pathlib.PurePosixPath: + expected = posixpath + elif self.cls is pathlib.PureWindowsPath: + expected = ntpath + else: + expected = os.path + p = self.cls('a') + self.assertIs(p.parser, expected) + + def test_different_parsers_unequal(self): + p = self.cls('a') + if p.parser is posixpath: + q = pathlib.PureWindowsPath('a') + else: + q = pathlib.PurePosixPath('a') + self.assertNotEqual(p, q) + + def test_different_parsers_unordered(self): + p = self.cls('a') + if p.parser is posixpath: + q = pathlib.PureWindowsPath('a') + else: + q = pathlib.PurePosixPath('a') + with self.assertRaises(TypeError): + p < q + with self.assertRaises(TypeError): + p <= q + with self.assertRaises(TypeError): + p > q + with self.assertRaises(TypeError): + p >= q + + def test_constructor_nested(self): + P = self.cls + P(FakePath("a/b/c")) + self.assertEqual(P(P('a')), P('a')) + self.assertEqual(P(P('a'), 'b'), P('a/b')) + self.assertEqual(P(P('a'), P('b')), P('a/b')) + self.assertEqual(P(P('a'), P('b'), P('c')), P(FakePath("a/b/c"))) + self.assertEqual(P(P('./a:b')), P('./a:b')) + + @needs_windows + def test_constructor_nested_foreign_flavour(self): + # See GH-125069. + p1 = pathlib.PurePosixPath('b/c:\\d') + p2 = pathlib.PurePosixPath('b/', 'c:\\d') + self.assertEqual(p1, p2) + self.assertEqual(self.cls(p1), self.cls('b/c:/d')) + self.assertEqual(self.cls(p2), self.cls('b/c:/d')) + + def _check_parse_path(self, raw_path, *expected): + sep = self.parser.sep + actual = self.cls._parse_path(raw_path.replace('/', sep)) + self.assertEqual(actual, expected) + if altsep := self.parser.altsep: + actual = self.cls._parse_path(raw_path.replace('/', altsep)) + self.assertEqual(actual, expected) + + def test_parse_path_common(self): + check = self._check_parse_path + sep = self.parser.sep + check('', '', '', []) + check('a', '', '', ['a']) + check('a/', '', '', ['a']) + check('a/b', '', '', ['a', 'b']) + check('a/b/', '', '', ['a', 'b']) + check('a/b/c/d', '', '', ['a', 'b', 'c', 'd']) + check('a/b//c/d', '', '', ['a', 'b', 'c', 'd']) + check('a/b/c/d', '', '', ['a', 'b', 'c', 'd']) + check('.', '', '', []) + check('././b', '', '', ['b']) + check('a/./b', '', '', ['a', 'b']) + check('a/./.', '', '', ['a']) + check('/a/b', '', sep, ['a', 'b']) + + def test_empty_path(self): + # The empty path points to '.' + p = self.cls('') + self.assertEqual(str(p), '.') + # Special case for the empty path. + self._check_str('.', ('',)) + + def test_parts_interning(self): + P = self.cls + p = P('/usr/bin/foo') + q = P('/usr/local/bin') + # 'usr' + self.assertIs(p.parts[1], q.parts[1]) + # 'bin' + self.assertIs(p.parts[2], q.parts[3]) + + def test_join_nested(self): + P = self.cls + p = P('a/b').joinpath(P('c')) + self.assertEqual(p, P('a/b/c')) + + def test_div_nested(self): + P = self.cls + p = P('a/b') / P('c') + self.assertEqual(p, P('a/b/c')) + + def test_pickling_common(self): + P = self.cls + for pathstr in ('a', 'a/', 'a/b', 'a/b/c', '/', '/a/b', '/a/b/c', 'a/b/c/'): + with self.subTest(pathstr=pathstr): + p = P(pathstr) + for proto in range(0, pickle.HIGHEST_PROTOCOL + 1): + dumped = pickle.dumps(p, proto) + pp = pickle.loads(dumped) + self.assertIs(pp.__class__, p.__class__) + self.assertEqual(pp, p) + self.assertEqual(hash(pp), hash(p)) + self.assertEqual(str(pp), str(p)) + + def test_repr_common(self): + for pathstr in ('a', 'a/b', 'a/b/c', '/', '/a/b', '/a/b/c'): + with self.subTest(pathstr=pathstr): + p = self.cls(pathstr) + clsname = p.__class__.__name__ + r = repr(p) + # The repr() is in the form ClassName("forward-slashes path"). + self.assertTrue(r.startswith(clsname + '('), r) + self.assertTrue(r.endswith(')'), r) + inner = r[len(clsname) + 1 : -1] + self.assertEqual(eval(inner), p.as_posix()) + + def test_fspath_common(self): + P = self.cls + p = P('a/b') + self._check_str(p.__fspath__(), ('a/b',)) + self._check_str(os.fspath(p), ('a/b',)) + + def test_bytes_exc_message(self): + P = self.cls + message = (r"argument should be a str or an os\.PathLike object " + r"where __fspath__ returns a str, not 'bytes'") + with self.assertRaisesRegex(TypeError, message): + P(b'a') + with self.assertRaisesRegex(TypeError, message): + P(b'a', 'b') + with self.assertRaisesRegex(TypeError, message): + P('a', b'b') + + def test_as_bytes_common(self): + sep = os.fsencode(self.sep) + P = self.cls + self.assertEqual(bytes(P('a/b')), b'a' + sep + b'b') + + def test_eq_common(self): + P = self.cls + self.assertEqual(P('a/b'), P('a/b')) + self.assertEqual(P('a/b'), P('a', 'b')) + self.assertNotEqual(P('a/b'), P('a')) + self.assertNotEqual(P('a/b'), P('/a/b')) + self.assertNotEqual(P('a/b'), P()) + self.assertNotEqual(P('/a/b'), P('/')) + self.assertNotEqual(P(), P('/')) + self.assertNotEqual(P(), "") + self.assertNotEqual(P(), {}) + self.assertNotEqual(P(), int) + + def test_equivalences(self, equivalences=None): + if equivalences is None: + equivalences = self.equivalences + for k, tuples in equivalences.items(): + canon = k.replace('/', self.sep) + posix = k.replace(self.sep, '/') + if canon != posix: + tuples = tuples + [ + tuple(part.replace('/', self.sep) for part in t) + for t in tuples + ] + tuples.append((posix, )) + pcanon = self.cls(canon) + for t in tuples: + p = self.cls(*t) + self.assertEqual(p, pcanon, "failed with args {}".format(t)) + self.assertEqual(hash(p), hash(pcanon)) + self.assertEqual(str(p), canon) + self.assertEqual(p.as_posix(), posix) + + def test_ordering_common(self): + # Ordering is tuple-alike. + def assertLess(a, b): + self.assertLess(a, b) + self.assertGreater(b, a) + P = self.cls + a = P('a') + b = P('a/b') + c = P('abc') + d = P('b') + assertLess(a, b) + assertLess(a, c) + assertLess(a, d) + assertLess(b, c) + assertLess(c, d) + P = self.cls + a = P('/a') + b = P('/a/b') + c = P('/abc') + d = P('/b') + assertLess(a, b) + assertLess(a, c) + assertLess(a, d) + assertLess(b, c) + assertLess(c, d) + with self.assertRaises(TypeError): + P() < {} + + def test_as_uri_common(self): + P = self.cls + with self.assertRaises(ValueError): + P('a').as_uri() + with self.assertRaises(ValueError): + P().as_uri() + + def test_repr_roundtrips(self): + for pathstr in ('a', 'a/b', 'a/b/c', '/', '/a/b', '/a/b/c'): + with self.subTest(pathstr=pathstr): + p = self.cls(pathstr) + r = repr(p) + # The repr() roundtrips. + q = eval(r, pathlib.__dict__) + self.assertIs(q.__class__, p.__class__) + self.assertEqual(q, p) + self.assertEqual(repr(q), r) + + def test_name_empty(self): + P = self.cls + self.assertEqual(P('').name, '') + self.assertEqual(P('.').name, '') + self.assertEqual(P('/a/b/.').name, 'b') + + def test_stem_empty(self): + P = self.cls + self.assertEqual(P('').stem, '') + self.assertEqual(P('.').stem, '') + + def test_with_name_empty(self): + P = self.cls + self.assertRaises(ValueError, P('').with_name, 'd.xml') + self.assertRaises(ValueError, P('.').with_name, 'd.xml') + self.assertRaises(ValueError, P('/').with_name, 'd.xml') + self.assertRaises(ValueError, P('a/b').with_name, '') + self.assertRaises(ValueError, P('a/b').with_name, '.') + + def test_with_stem_empty(self): + P = self.cls + self.assertRaises(ValueError, P('').with_stem, 'd') + self.assertRaises(ValueError, P('.').with_stem, 'd') + self.assertRaises(ValueError, P('/').with_stem, 'd') + self.assertRaises(ValueError, P('a/b').with_stem, '') + self.assertRaises(ValueError, P('a/b').with_stem, '.') + + def test_relative_to_several_args(self): + P = self.cls + p = P('a/b') + with self.assertWarns(DeprecationWarning): + p.relative_to('a', 'b') + p.relative_to('a', 'b', walk_up=True) + + def test_is_relative_to_several_args(self): + P = self.cls + p = P('a/b') + with self.assertWarns(DeprecationWarning): + p.is_relative_to('a', 'b') + + def test_is_reserved_deprecated(self): + P = self.cls + p = P('a/b') + with self.assertWarns(DeprecationWarning): + p.is_reserved() + + def test_match_empty(self): + P = self.cls + self.assertRaises(ValueError, P('a').match, '') + self.assertRaises(ValueError, P('a').match, '.') + + @needs_posix + def test_parse_path_posix(self): + check = self._check_parse_path + # Collapsing of excess leading slashes, except for the double-slash + # special case. + check('//a/b', '', '//', ['a', 'b']) + check('///a/b', '', '/', ['a', 'b']) + check('////a/b', '', '/', ['a', 'b']) + # Paths which look like NT paths aren't treated specially. + check('c:a', '', '', ['c:a',]) + check('c:\\a', '', '', ['c:\\a',]) + check('\\a', '', '', ['\\a',]) + + @needs_posix + def test_eq_posix(self): + P = self.cls + self.assertNotEqual(P('a/b'), P('A/b')) + self.assertEqual(P('/a'), P('///a')) + self.assertNotEqual(P('/a'), P('//a')) + + @needs_posix + def test_as_uri_posix(self): + P = self.cls + self.assertEqual(P('/').as_uri(), 'file:///') + self.assertEqual(P('/a/b.c').as_uri(), 'file:///a/b.c') + self.assertEqual(P('/a/b%#c').as_uri(), 'file:///a/b%25%23c') + + @needs_posix + def test_as_uri_non_ascii(self): + from urllib.parse import quote_from_bytes + P = self.cls + try: + os.fsencode('\xe9') + except UnicodeEncodeError: + self.skipTest("\\xe9 cannot be encoded to the filesystem encoding") + self.assertEqual(P('/a/b\xe9').as_uri(), + 'file:///a/b' + quote_from_bytes(os.fsencode('\xe9'))) + + @needs_posix + def test_parse_windows_path(self): + P = self.cls + p = P('c:', 'a', 'b') + pp = P(pathlib.PureWindowsPath('c:\\a\\b')) + self.assertEqual(p, pp) + + windows_equivalences = { + './a:b': [ ('./a:b',) ], + 'c:a': [ ('c:', 'a'), ('c:', 'a/'), ('.', 'c:', 'a') ], + 'c:/a': [ + ('c:/', 'a'), ('c:', '/', 'a'), ('c:', '/a'), + ('/z', 'c:/', 'a'), ('//x/y', 'c:/', 'a'), + ], + '//a/b/': [ ('//a/b',) ], + '//a/b/c': [ + ('//a/b', 'c'), ('//a/b/', 'c'), + ], + } + + @needs_windows + def test_equivalences_windows(self): + self.test_equivalences(self.windows_equivalences) + + @needs_windows + def test_parse_path_windows(self): + check = self._check_parse_path + # First part is anchored. + check('c:', 'c:', '', []) + check('c:/', 'c:', '\\', []) + check('/', '', '\\', []) + check('c:a', 'c:', '', ['a']) + check('c:/a', 'c:', '\\', ['a']) + check('/a', '', '\\', ['a']) + # UNC paths. + check('//', '\\\\', '', []) + check('//a', '\\\\a', '', []) + check('//a/', '\\\\a\\', '', []) + check('//a/b', '\\\\a\\b', '\\', []) + check('//a/b/', '\\\\a\\b', '\\', []) + check('//a/b/c', '\\\\a\\b', '\\', ['c']) + # Collapsing and stripping excess slashes. + check('Z://b//c/d/', 'Z:', '\\', ['b', 'c', 'd']) + # UNC paths. + check('//b/c//d', '\\\\b\\c', '\\', ['d']) + # Extended paths. + check('//./c:', '\\\\.\\c:', '', []) + check('//?/c:/', '\\\\?\\c:', '\\', []) + check('//?/c:/a', '\\\\?\\c:', '\\', ['a']) + # Extended UNC paths (format is "\\?\UNC\server\share"). + check('//?', '\\\\?', '', []) + check('//?/', '\\\\?\\', '', []) + check('//?/UNC', '\\\\?\\UNC', '', []) + check('//?/UNC/', '\\\\?\\UNC\\', '', []) + check('//?/UNC/b', '\\\\?\\UNC\\b', '', []) + check('//?/UNC/b/', '\\\\?\\UNC\\b\\', '', []) + check('//?/UNC/b/c', '\\\\?\\UNC\\b\\c', '\\', []) + check('//?/UNC/b/c/', '\\\\?\\UNC\\b\\c', '\\', []) + check('//?/UNC/b/c/d', '\\\\?\\UNC\\b\\c', '\\', ['d']) + # UNC device paths + check('//./BootPartition/', '\\\\.\\BootPartition', '\\', []) + check('//?/BootPartition/', '\\\\?\\BootPartition', '\\', []) + check('//./PhysicalDrive0', '\\\\.\\PhysicalDrive0', '', []) + check('//?/Volume{}/', '\\\\?\\Volume{}', '\\', []) + check('//./nul', '\\\\.\\nul', '', []) + # Paths to files with NTFS alternate data streams + check('./c:s', '', '', ['c:s']) + check('cc:s', '', '', ['cc:s']) + check('C:c:s', 'C:', '', ['c:s']) + check('C:/c:s', 'C:', '\\', ['c:s']) + check('D:a/c:b', 'D:', '', ['a', 'c:b']) + check('D:/a/c:b', 'D:', '\\', ['a', 'c:b']) + + @needs_windows + def test_eq_windows(self): + P = self.cls + self.assertEqual(P('c:a/b'), P('c:a/b')) + self.assertEqual(P('c:a/b'), P('c:', 'a', 'b')) + self.assertNotEqual(P('c:a/b'), P('d:a/b')) + self.assertNotEqual(P('c:a/b'), P('c:/a/b')) + self.assertNotEqual(P('/a/b'), P('c:/a/b')) + # Case-insensitivity. + self.assertEqual(P('a/B'), P('A/b')) + self.assertEqual(P('C:a/B'), P('c:A/b')) + self.assertEqual(P('//Some/SHARE/a/B'), P('//somE/share/A/b')) + self.assertEqual(P('\u0130'), P('i\u0307')) + + @needs_windows + def test_as_uri_windows(self): + P = self.cls + with self.assertRaises(ValueError): + P('/a/b').as_uri() + with self.assertRaises(ValueError): + P('c:a/b').as_uri() + self.assertEqual(P('c:/').as_uri(), 'file:///c:/') + self.assertEqual(P('c:/a/b.c').as_uri(), 'file:///c:/a/b.c') + self.assertEqual(P('c:/a/b%#c').as_uri(), 'file:///c:/a/b%25%23c') + self.assertEqual(P('//some/share/').as_uri(), 'file://some/share/') + self.assertEqual(P('//some/share/a/b.c').as_uri(), + 'file://some/share/a/b.c') + + from urllib.parse import quote_from_bytes + QUOTED_FS_NONASCII = quote_from_bytes(os.fsencode(FS_NONASCII)) + self.assertEqual(P('c:/a/b' + FS_NONASCII).as_uri(), + 'file:///c:/a/b' + QUOTED_FS_NONASCII) + self.assertEqual(P('//some/share/a/b%#c' + FS_NONASCII).as_uri(), + 'file://some/share/a/b%25%23c' + QUOTED_FS_NONASCII) + + @needs_windows + def test_ordering_windows(self): + # Case-insensitivity. + def assertOrderedEqual(a, b): + self.assertLessEqual(a, b) + self.assertGreaterEqual(b, a) + P = self.cls + p = P('c:A/b') + q = P('C:a/B') + assertOrderedEqual(p, q) + self.assertFalse(p < q) + self.assertFalse(p > q) + p = P('//some/Share/A/b') + q = P('//Some/SHARE/a/B') + assertOrderedEqual(p, q) + self.assertFalse(p < q) + self.assertFalse(p > q) + + +class PurePosixPathTest(PurePathTest): + cls = pathlib.PurePosixPath + + +class PureWindowsPathTest(PurePathTest): + cls = pathlib.PureWindowsPath + + +class PurePathSubclassTest(PurePathTest): + class cls(pathlib.PurePath): + pass + + # repr() roundtripping is not supported in custom subclass. + test_repr_roundtrips = None + + +# +# Tests for the concrete classes. +# + +class PathTest(test_pathlib_abc.DummyPathTest, PurePathTest): + """Tests for the FS-accessing functionalities of the Path classes.""" + cls = pathlib.Path + can_symlink = os_helper.can_symlink() + + def setUp(self): + super().setUp() + os.chmod(self.parser.join(self.base, 'dirE'), 0) + + def tearDown(self): + os.chmod(self.parser.join(self.base, 'dirE'), 0o777) + os_helper.rmtree(self.base) + + def tempdir(self): + d = os_helper._longpath(tempfile.mkdtemp(suffix='-dirD', + dir=os.getcwd())) + self.addCleanup(os_helper.rmtree, d) + return d + + def test_matches_pathbase_api(self): + our_names = {name for name in dir(self.cls) if name[0] != '_'} + our_names.remove('is_reserved') # only present in PurePath + path_names = {name for name in dir(pathlib._abc.PathBase) if name[0] != '_'} + self.assertEqual(our_names, path_names) + for attr_name in our_names: + if attr_name == 'parser': + # On Windows, Path.parser is ntpath, but PathBase.parser is + # posixpath, and so their docstrings differ. + continue + our_attr = getattr(self.cls, attr_name) + path_attr = getattr(pathlib._abc.PathBase, attr_name) + self.assertEqual(our_attr.__doc__, path_attr.__doc__) + + def test_concrete_class(self): + if self.cls is pathlib.Path: + expected = pathlib.WindowsPath if os.name == 'nt' else pathlib.PosixPath + else: + expected = self.cls + p = self.cls('a') + self.assertIs(type(p), expected) + + def test_unsupported_parser(self): + if self.cls.parser is os.path: + self.skipTest("path parser is supported") + else: + self.assertRaises(pathlib.UnsupportedOperation, self.cls) + + def _test_cwd(self, p): + q = self.cls(os.getcwd()) + self.assertEqual(p, q) + self.assertEqualNormCase(str(p), str(q)) + self.assertIs(type(p), type(q)) + self.assertTrue(p.is_absolute()) + + def test_cwd(self): + p = self.cls.cwd() + self._test_cwd(p) + + def test_absolute_common(self): + P = self.cls + + with mock.patch("os.getcwd") as getcwd: + getcwd.return_value = self.base + + # Simple relative paths. + self.assertEqual(str(P().absolute()), self.base) + self.assertEqual(str(P('.').absolute()), self.base) + self.assertEqual(str(P('a').absolute()), os.path.join(self.base, 'a')) + self.assertEqual(str(P('a', 'b', 'c').absolute()), os.path.join(self.base, 'a', 'b', 'c')) + + # Symlinks should not be resolved. + self.assertEqual(str(P('linkB', 'fileB').absolute()), os.path.join(self.base, 'linkB', 'fileB')) + self.assertEqual(str(P('brokenLink').absolute()), os.path.join(self.base, 'brokenLink')) + self.assertEqual(str(P('brokenLinkLoop').absolute()), os.path.join(self.base, 'brokenLinkLoop')) + + # '..' entries should be preserved and not normalised. + self.assertEqual(str(P('..').absolute()), os.path.join(self.base, '..')) + self.assertEqual(str(P('a', '..').absolute()), os.path.join(self.base, 'a', '..')) + self.assertEqual(str(P('..', 'b').absolute()), os.path.join(self.base, '..', 'b')) + + def _test_home(self, p): + q = self.cls(os.path.expanduser('~')) + self.assertEqual(p, q) + self.assertEqualNormCase(str(p), str(q)) + self.assertIs(type(p), type(q)) + self.assertTrue(p.is_absolute()) + + @unittest.skipIf( + pwd is None, reason="Test requires pwd module to get homedir." + ) + def test_home(self): + with os_helper.EnvironmentVarGuard() as env: + self._test_home(self.cls.home()) + + env.clear() + env['USERPROFILE'] = os.path.join(self.base, 'userprofile') + self._test_home(self.cls.home()) + + # bpo-38883: ignore `HOME` when set on windows + env['HOME'] = os.path.join(self.base, 'home') + self._test_home(self.cls.home()) + + @unittest.skipIf(is_wasi, "WASI has no user accounts.") + def test_expanduser_common(self): + P = self.cls + p = P('~') + self.assertEqual(p.expanduser(), P(os.path.expanduser('~'))) + p = P('foo') + self.assertEqual(p.expanduser(), p) + p = P('/~') + self.assertEqual(p.expanduser(), p) + p = P('../~') + self.assertEqual(p.expanduser(), p) + p = P(P('').absolute().anchor) / '~' + self.assertEqual(p.expanduser(), p) + p = P('~/a:b') + self.assertEqual(p.expanduser(), P(os.path.expanduser('~'), './a:b')) + + def test_with_segments(self): + class P(self.cls): + def __init__(self, *pathsegments, session_id): + super().__init__(*pathsegments) + self.session_id = session_id + + def with_segments(self, *pathsegments): + return type(self)(*pathsegments, session_id=self.session_id) + p = P(self.base, session_id=42) + self.assertEqual(42, p.absolute().session_id) + self.assertEqual(42, p.resolve().session_id) + if not is_wasi: # WASI has no user accounts. + self.assertEqual(42, p.with_segments('~').expanduser().session_id) + self.assertEqual(42, (p / 'fileA').rename(p / 'fileB').session_id) + self.assertEqual(42, (p / 'fileB').replace(p / 'fileA').session_id) + if self.can_symlink: + self.assertEqual(42, (p / 'linkA').readlink().session_id) + for path in p.iterdir(): + self.assertEqual(42, path.session_id) + for path in p.glob('*'): + self.assertEqual(42, path.session_id) + for path in p.rglob('*'): + self.assertEqual(42, path.session_id) + for dirpath, dirnames, filenames in p.walk(): + self.assertEqual(42, dirpath.session_id) + + def test_open_unbuffered(self): + p = self.cls(self.base) + with (p / 'fileA').open('rb', buffering=0) as f: + self.assertIsInstance(f, io.RawIOBase) + self.assertEqual(f.read().strip(), b"this is file A") + + def test_resolve_nonexist_relative_issue38671(self): + p = self.cls('non', 'exist') + + old_cwd = os.getcwd() + os.chdir(self.base) + try: + self.assertEqual(p.resolve(), self.cls(self.base, p)) + finally: + os.chdir(old_cwd) + + @os_helper.skip_unless_working_chmod + def test_chmod(self): + p = self.cls(self.base) / 'fileA' + mode = p.stat().st_mode + # Clear writable bit. + new_mode = mode & ~0o222 + p.chmod(new_mode) + self.assertEqual(p.stat().st_mode, new_mode) + # Set writable bit. + new_mode = mode | 0o222 + p.chmod(new_mode) + self.assertEqual(p.stat().st_mode, new_mode) + + # On Windows, os.chmod does not follow symlinks (issue #15411) + @needs_posix + @os_helper.skip_unless_working_chmod + def test_chmod_follow_symlinks_true(self): + p = self.cls(self.base) / 'linkA' + q = p.resolve() + mode = q.stat().st_mode + # Clear writable bit. + new_mode = mode & ~0o222 + p.chmod(new_mode, follow_symlinks=True) + self.assertEqual(q.stat().st_mode, new_mode) + # Set writable bit + new_mode = mode | 0o222 + p.chmod(new_mode, follow_symlinks=True) + self.assertEqual(q.stat().st_mode, new_mode) + + # XXX also need a test for lchmod. + + def _get_pw_name_or_skip_test(self, uid): + try: + return pwd.getpwuid(uid).pw_name + except KeyError: + self.skipTest( + "user %d doesn't have an entry in the system database" % uid) + + @unittest.skipUnless(pwd, "the pwd module is needed for this test") + def test_owner(self): + p = self.cls(self.base) / 'fileA' + expected_uid = p.stat().st_uid + expected_name = self._get_pw_name_or_skip_test(expected_uid) + + self.assertEqual(expected_name, p.owner()) + + @unittest.skipUnless(pwd, "the pwd module is needed for this test") + @unittest.skipUnless(root_in_posix, "test needs root privilege") + def test_owner_no_follow_symlinks(self): + all_users = [u.pw_uid for u in pwd.getpwall()] + if len(all_users) < 2: + self.skipTest("test needs more than one user") + + target = self.cls(self.base) / 'fileA' + link = self.cls(self.base) / 'linkA' + + uid_1, uid_2 = all_users[:2] + os.chown(target, uid_1, -1) + os.chown(link, uid_2, -1, follow_symlinks=False) + + expected_uid = link.stat(follow_symlinks=False).st_uid + expected_name = self._get_pw_name_or_skip_test(expected_uid) + + self.assertEqual(expected_uid, uid_2) + self.assertEqual(expected_name, link.owner(follow_symlinks=False)) + + def _get_gr_name_or_skip_test(self, gid): + try: + return grp.getgrgid(gid).gr_name + except KeyError: + self.skipTest( + "group %d doesn't have an entry in the system database" % gid) + + @unittest.skipUnless(grp, "the grp module is needed for this test") + def test_group(self): + p = self.cls(self.base) / 'fileA' + expected_gid = p.stat().st_gid + expected_name = self._get_gr_name_or_skip_test(expected_gid) + + self.assertEqual(expected_name, p.group()) + + @unittest.skipUnless(grp, "the grp module is needed for this test") + @unittest.skipUnless(root_in_posix, "test needs root privilege") + def test_group_no_follow_symlinks(self): + all_groups = [g.gr_gid for g in grp.getgrall()] + if len(all_groups) < 2: + self.skipTest("test needs more than one group") + + target = self.cls(self.base) / 'fileA' + link = self.cls(self.base) / 'linkA' + + gid_1, gid_2 = all_groups[:2] + os.chown(target, -1, gid_1) + os.chown(link, -1, gid_2, follow_symlinks=False) + + expected_gid = link.stat(follow_symlinks=False).st_gid + expected_name = self._get_gr_name_or_skip_test(expected_gid) + + self.assertEqual(expected_gid, gid_2) + self.assertEqual(expected_name, link.group(follow_symlinks=False)) + + def test_unlink(self): + p = self.cls(self.base) / 'fileA' + p.unlink() + self.assertFileNotFound(p.stat) + self.assertFileNotFound(p.unlink) + + def test_unlink_missing_ok(self): + p = self.cls(self.base) / 'fileAAA' + self.assertFileNotFound(p.unlink) + p.unlink(missing_ok=True) + + def test_rmdir(self): + p = self.cls(self.base) / 'dirA' + for q in p.iterdir(): + q.unlink() + p.rmdir() + self.assertFileNotFound(p.stat) + self.assertFileNotFound(p.unlink) + + @os_helper.skip_unless_hardlink + def test_hardlink_to(self): + P = self.cls(self.base) + target = P / 'fileA' + size = target.stat().st_size + # linking to another path. + link = P / 'dirA' / 'fileAA' + link.hardlink_to(target) + self.assertEqual(link.stat().st_size, size) + self.assertTrue(os.path.samefile(target, link)) + self.assertTrue(target.exists()) + # Linking to a str of a relative path. + link2 = P / 'dirA' / 'fileAAA' + target2 = self.parser.join(TESTFN, 'fileA') + link2.hardlink_to(target2) + self.assertEqual(os.stat(target2).st_size, size) + self.assertTrue(link2.exists()) + + @unittest.skipIf(hasattr(os, "link"), "os.link() is present") + def test_hardlink_to_unsupported(self): + P = self.cls(self.base) + p = P / 'fileA' + # linking to another path. + q = P / 'dirA' / 'fileAA' + with self.assertRaises(pathlib.UnsupportedOperation): + q.hardlink_to(p) + + def test_rename(self): + P = self.cls(self.base) + p = P / 'fileA' + size = p.stat().st_size + # Renaming to another path. + q = P / 'dirA' / 'fileAA' + renamed_p = p.rename(q) + self.assertEqual(renamed_p, q) + self.assertEqual(q.stat().st_size, size) + self.assertFileNotFound(p.stat) + # Renaming to a str of a relative path. + r = self.parser.join(TESTFN, 'fileAAA') + renamed_q = q.rename(r) + self.assertEqual(renamed_q, self.cls(r)) + self.assertEqual(os.stat(r).st_size, size) + self.assertFileNotFound(q.stat) + + def test_replace(self): + P = self.cls(self.base) + p = P / 'fileA' + size = p.stat().st_size + # Replacing a non-existing path. + q = P / 'dirA' / 'fileAA' + replaced_p = p.replace(q) + self.assertEqual(replaced_p, q) + self.assertEqual(q.stat().st_size, size) + self.assertFileNotFound(p.stat) + # Replacing another (existing) path. + r = self.parser.join(TESTFN, 'dirB', 'fileB') + replaced_q = q.replace(r) + self.assertEqual(replaced_q, self.cls(r)) + self.assertEqual(os.stat(r).st_size, size) + self.assertFileNotFound(q.stat) + + def test_touch_common(self): + P = self.cls(self.base) + p = P / 'newfileA' + self.assertFalse(p.exists()) + p.touch() + self.assertTrue(p.exists()) + st = p.stat() + old_mtime = st.st_mtime + old_mtime_ns = st.st_mtime_ns + # Rewind the mtime sufficiently far in the past to work around + # filesystem-specific timestamp granularity. + os.utime(str(p), (old_mtime - 10, old_mtime - 10)) + # The file mtime should be refreshed by calling touch() again. + p.touch() + st = p.stat() + self.assertGreaterEqual(st.st_mtime_ns, old_mtime_ns) + self.assertGreaterEqual(st.st_mtime, old_mtime) + # Now with exist_ok=False. + p = P / 'newfileB' + self.assertFalse(p.exists()) + p.touch(mode=0o700, exist_ok=False) + self.assertTrue(p.exists()) + self.assertRaises(OSError, p.touch, exist_ok=False) + + def test_touch_nochange(self): + P = self.cls(self.base) + p = P / 'fileA' + p.touch() + with p.open('rb') as f: + self.assertEqual(f.read().strip(), b"this is file A") + + def test_mkdir(self): + P = self.cls(self.base) + p = P / 'newdirA' + self.assertFalse(p.exists()) + p.mkdir() + self.assertTrue(p.exists()) + self.assertTrue(p.is_dir()) + with self.assertRaises(OSError) as cm: + p.mkdir() + self.assertEqual(cm.exception.errno, errno.EEXIST) + + def test_mkdir_parents(self): + # Creating a chain of directories. + p = self.cls(self.base, 'newdirB', 'newdirC') + self.assertFalse(p.exists()) + with self.assertRaises(OSError) as cm: + p.mkdir() + self.assertEqual(cm.exception.errno, errno.ENOENT) + p.mkdir(parents=True) + self.assertTrue(p.exists()) + self.assertTrue(p.is_dir()) + with self.assertRaises(OSError) as cm: + p.mkdir(parents=True) + self.assertEqual(cm.exception.errno, errno.EEXIST) + # Test `mode` arg. + mode = stat.S_IMODE(p.stat().st_mode) # Default mode. + p = self.cls(self.base, 'newdirD', 'newdirE') + p.mkdir(0o555, parents=True) + self.assertTrue(p.exists()) + self.assertTrue(p.is_dir()) + if os.name != 'nt': + # The directory's permissions follow the mode argument. + self.assertEqual(stat.S_IMODE(p.stat().st_mode), 0o7555 & mode) + # The parent's permissions follow the default process settings. + self.assertEqual(stat.S_IMODE(p.parent.stat().st_mode), mode) + + def test_mkdir_exist_ok(self): + p = self.cls(self.base, 'dirB') + st_ctime_first = p.stat().st_ctime + self.assertTrue(p.exists()) + self.assertTrue(p.is_dir()) + with self.assertRaises(FileExistsError) as cm: + p.mkdir() + self.assertEqual(cm.exception.errno, errno.EEXIST) + p.mkdir(exist_ok=True) + self.assertTrue(p.exists()) + self.assertEqual(p.stat().st_ctime, st_ctime_first) + + def test_mkdir_exist_ok_with_parent(self): + p = self.cls(self.base, 'dirC') + self.assertTrue(p.exists()) + with self.assertRaises(FileExistsError) as cm: + p.mkdir() + self.assertEqual(cm.exception.errno, errno.EEXIST) + p = p / 'newdirC' + p.mkdir(parents=True) + st_ctime_first = p.stat().st_ctime + self.assertTrue(p.exists()) + with self.assertRaises(FileExistsError) as cm: + p.mkdir(parents=True) + self.assertEqual(cm.exception.errno, errno.EEXIST) + p.mkdir(parents=True, exist_ok=True) + self.assertTrue(p.exists()) + self.assertEqual(p.stat().st_ctime, st_ctime_first) + + @unittest.skipIf(is_emscripten, "FS root cannot be modified on Emscripten.") + def test_mkdir_exist_ok_root(self): + # Issue #25803: A drive root could raise PermissionError on Windows. + self.cls('/').resolve().mkdir(exist_ok=True) + self.cls('/').resolve().mkdir(parents=True, exist_ok=True) + + @needs_windows # XXX: not sure how to test this on POSIX. + def test_mkdir_with_unknown_drive(self): + for d in 'ZYXWVUTSRQPONMLKJIHGFEDCBA': + p = self.cls(d + ':\\') + if not p.is_dir(): + break + else: + self.skipTest("cannot find a drive that doesn't exist") + with self.assertRaises(OSError): + (p / 'child' / 'path').mkdir(parents=True) + + def test_mkdir_with_child_file(self): + p = self.cls(self.base, 'dirB', 'fileB') + self.assertTrue(p.exists()) + # An exception is raised when the last path component is an existing + # regular file, regardless of whether exist_ok is true or not. + with self.assertRaises(FileExistsError) as cm: + p.mkdir(parents=True) + self.assertEqual(cm.exception.errno, errno.EEXIST) + with self.assertRaises(FileExistsError) as cm: + p.mkdir(parents=True, exist_ok=True) + self.assertEqual(cm.exception.errno, errno.EEXIST) + + def test_mkdir_no_parents_file(self): + p = self.cls(self.base, 'fileA') + self.assertTrue(p.exists()) + # An exception is raised when the last path component is an existing + # regular file, regardless of whether exist_ok is true or not. + with self.assertRaises(FileExistsError) as cm: + p.mkdir() + self.assertEqual(cm.exception.errno, errno.EEXIST) + with self.assertRaises(FileExistsError) as cm: + p.mkdir(exist_ok=True) + self.assertEqual(cm.exception.errno, errno.EEXIST) + + def test_mkdir_concurrent_parent_creation(self): + for pattern_num in range(32): + p = self.cls(self.base, 'dirCPC%d' % pattern_num) + self.assertFalse(p.exists()) + + real_mkdir = os.mkdir + def my_mkdir(path, mode=0o777): + path = str(path) + # Emulate another process that would create the directory + # just before we try to create it ourselves. We do it + # in all possible pattern combinations, assuming that this + # function is called at most 5 times (dirCPC/dir1/dir2, + # dirCPC/dir1, dirCPC, dirCPC/dir1, dirCPC/dir1/dir2). + if pattern.pop(): + real_mkdir(path, mode) # From another process. + concurrently_created.add(path) + real_mkdir(path, mode) # Our real call. + + pattern = [bool(pattern_num & (1 << n)) for n in range(5)] + concurrently_created = set() + p12 = p / 'dir1' / 'dir2' + try: + with mock.patch("os.mkdir", my_mkdir): + p12.mkdir(parents=True, exist_ok=False) + except FileExistsError: + self.assertIn(str(p12), concurrently_created) + else: + self.assertNotIn(str(p12), concurrently_created) + self.assertTrue(p.exists()) + + @needs_symlinks + def test_symlink_to(self): + P = self.cls(self.base) + target = P / 'fileA' + # Symlinking a path target. + link = P / 'dirA' / 'linkAA' + link.symlink_to(target) + self.assertEqual(link.stat(), target.stat()) + self.assertNotEqual(link.lstat(), target.stat()) + # Symlinking a str target. + link = P / 'dirA' / 'linkAAA' + link.symlink_to(str(target)) + self.assertEqual(link.stat(), target.stat()) + self.assertNotEqual(link.lstat(), target.stat()) + self.assertFalse(link.is_dir()) + # Symlinking to a directory. + target = P / 'dirB' + link = P / 'dirA' / 'linkAAAA' + link.symlink_to(target, target_is_directory=True) + self.assertEqual(link.stat(), target.stat()) + self.assertNotEqual(link.lstat(), target.stat()) + self.assertTrue(link.is_dir()) + self.assertTrue(list(link.iterdir())) + + @unittest.skipIf(hasattr(os, "symlink"), "os.symlink() is present") + def test_symlink_to_unsupported(self): + P = self.cls(self.base) + p = P / 'fileA' + # linking to another path. + q = P / 'dirA' / 'fileAA' + with self.assertRaises(pathlib.UnsupportedOperation): + q.symlink_to(p) + + def test_is_junction(self): + P = self.cls(self.base) + + with mock.patch.object(P.parser, 'isjunction'): + self.assertEqual(P.is_junction(), P.parser.isjunction.return_value) + P.parser.isjunction.assert_called_once_with(P) + + @unittest.skipUnless(hasattr(os, "mkfifo"), "os.mkfifo() required") + @unittest.skipIf(sys.platform == "vxworks", + "fifo requires special path on VxWorks") + def test_is_fifo_true(self): + P = self.cls(self.base, 'myfifo') + try: + os.mkfifo(str(P)) + except PermissionError as e: + self.skipTest('os.mkfifo(): %s' % e) + self.assertTrue(P.is_fifo()) + self.assertFalse(P.is_socket()) + self.assertFalse(P.is_file()) + self.assertIs(self.cls(self.base, 'myfifo\udfff').is_fifo(), False) + self.assertIs(self.cls(self.base, 'myfifo\x00').is_fifo(), False) + + @unittest.skipUnless(hasattr(socket, "AF_UNIX"), "Unix sockets required") + @unittest.skipIf( + is_emscripten, "Unix sockets are not implemented on Emscripten." + ) + @unittest.skipIf( + is_wasi, "Cannot create socket on WASI." + ) + def test_is_socket_true(self): + P = self.cls(self.base, 'mysock') + sock = socket.socket(socket.AF_UNIX, socket.SOCK_STREAM) + self.addCleanup(sock.close) + try: + sock.bind(str(P)) + except OSError as e: + if (isinstance(e, PermissionError) or + "AF_UNIX path too long" in str(e)): + self.skipTest("cannot bind Unix socket: " + str(e)) + self.assertTrue(P.is_socket()) + self.assertFalse(P.is_fifo()) + self.assertFalse(P.is_file()) + self.assertIs(self.cls(self.base, 'mysock\udfff').is_socket(), False) + self.assertIs(self.cls(self.base, 'mysock\x00').is_socket(), False) + + def test_is_char_device_true(self): + # os.devnull should generally be a char device. + P = self.cls(os.devnull) + if not P.exists(): + self.skipTest("null device required") + self.assertTrue(P.is_char_device()) + self.assertFalse(P.is_block_device()) + self.assertFalse(P.is_file()) + self.assertIs(self.cls(f'{os.devnull}\udfff').is_char_device(), False) + self.assertIs(self.cls(f'{os.devnull}\x00').is_char_device(), False) + + def test_is_mount_root(self): + if os.name == 'nt': + R = self.cls('c:\\') + else: + R = self.cls('/') + self.assertTrue(R.is_mount()) + self.assertFalse((R / '\udfff').is_mount()) + + def test_passing_kwargs_deprecated(self): + with self.assertWarns(DeprecationWarning): + self.cls(foo="bar") + + def setUpWalk(self): + super().setUpWalk() + sub21_path= self.sub2_path / "SUB21" + tmp5_path = sub21_path / "tmp3" + broken_link3_path = self.sub2_path / "broken_link3" + + os.makedirs(sub21_path) + tmp5_path.write_text("I am tmp5, blame test_pathlib.") + if self.can_symlink: + os.symlink(tmp5_path, broken_link3_path) + self.sub2_tree[2].append('broken_link3') + self.sub2_tree[2].sort() + if not is_emscripten: + # Emscripten fails with inaccessible directories. + os.chmod(sub21_path, 0) + try: + os.listdir(sub21_path) + except PermissionError: + self.sub2_tree[1].append('SUB21') + else: + os.chmod(sub21_path, stat.S_IRWXU) + os.unlink(tmp5_path) + os.rmdir(sub21_path) + + def test_walk_bad_dir(self): + self.setUpWalk() + errors = [] + walk_it = self.walk_path.walk(on_error=errors.append) + root, dirs, files = next(walk_it) + self.assertEqual(errors, []) + dir1 = 'SUB1' + path1 = root / dir1 + path1new = (root / dir1).with_suffix(".new") + path1.rename(path1new) + try: + roots = [r for r, _, _ in walk_it] + self.assertTrue(errors) + self.assertNotIn(path1, roots) + self.assertNotIn(path1new, roots) + for dir2 in dirs: + if dir2 != dir1: + self.assertIn(root / dir2, roots) + finally: + path1new.rename(path1) + + def test_walk_many_open_files(self): + depth = 30 + base = self.cls(self.base, 'deep') + path = self.cls(base, *(['d']*depth)) + path.mkdir(parents=True) + + iters = [base.walk(top_down=False) for _ in range(100)] + for i in range(depth + 1): + expected = (path, ['d'] if i else [], []) + for it in iters: + self.assertEqual(next(it), expected) + path = path.parent + + iters = [base.walk(top_down=True) for _ in range(100)] + path = base + for i in range(depth + 1): + expected = (path, ['d'] if i < depth else [], []) + for it in iters: + self.assertEqual(next(it), expected) + path = path / 'd' + + def test_walk_above_recursion_limit(self): + recursion_limit = 40 + # directory_depth > recursion_limit + directory_depth = recursion_limit + 10 + base = self.cls(self.base, 'deep') + path = base.joinpath(*(['d'] * directory_depth)) + path.mkdir(parents=True) + + with infinite_recursion(recursion_limit): + list(base.walk()) + list(base.walk(top_down=False)) + + def test_glob_empty_pattern(self): + p = self.cls('') + with self.assertRaisesRegex(ValueError, 'Unacceptable pattern'): + list(p.glob('')) + with self.assertRaisesRegex(ValueError, 'Unacceptable pattern'): + list(p.glob('.')) + with self.assertRaisesRegex(ValueError, 'Unacceptable pattern'): + list(p.glob('./')) + + def test_glob_many_open_files(self): + depth = 30 + P = self.cls + p = base = P(self.base) / 'deep' + p.mkdir() + for _ in range(depth): + p /= 'd' + p.mkdir() + pattern = '/'.join(['*'] * depth) + iters = [base.glob(pattern) for j in range(100)] + for it in iters: + self.assertEqual(next(it), p) + iters = [base.rglob('d') for j in range(100)] + p = base + for i in range(depth): + p = p / 'd' + for it in iters: + self.assertEqual(next(it), p) + + def test_glob_above_recursion_limit(self): + recursion_limit = 50 + # directory_depth > recursion_limit + directory_depth = recursion_limit + 10 + base = self.cls(self.base, 'deep') + path = base.joinpath(*(['d'] * directory_depth)) + path.mkdir(parents=True) + + with infinite_recursion(recursion_limit): + list(base.glob('**/')) + + def test_glob_pathlike(self): + P = self.cls + p = P(self.base) + pattern = "dir*/file*" + expect = {p / "dirB/fileB", p / "dirC/fileC"} + self.assertEqual(expect, set(p.glob(P(pattern)))) + self.assertEqual(expect, set(p.glob(FakePath(pattern)))) + + @needs_symlinks + def test_glob_dot(self): + P = self.cls + with os_helper.change_cwd(P(self.base, "dirC")): + self.assertEqual( + set(P('.').glob('*')), {P("fileC"), P("novel.txt"), P("dirD")}) + self.assertEqual( + set(P('.').glob('**')), {P("fileC"), P("novel.txt"), P("dirD"), P("dirD/fileD"), P(".")}) + self.assertEqual( + set(P('.').glob('**/*')), {P("fileC"), P("novel.txt"), P("dirD"), P("dirD/fileD")}) + self.assertEqual( + set(P('.').glob('**/*/*')), {P("dirD/fileD")}) + + def test_glob_inaccessible(self): + P = self.cls + p = P(self.base, "mydir1", "mydir2") + p.mkdir(parents=True) + p.parent.chmod(0) + self.assertEqual(set(p.glob('*')), set()) + + def test_rglob_pathlike(self): + P = self.cls + p = P(self.base, "dirC") + pattern = "**/file*" + expect = {p / "fileC", p / "dirD/fileD"} + self.assertEqual(expect, set(p.rglob(P(pattern)))) + self.assertEqual(expect, set(p.rglob(FakePath(pattern)))) + + @needs_posix + def test_absolute_posix(self): + P = self.cls + self.assertEqual(str(P('/').absolute()), '/') + self.assertEqual(str(P('/a').absolute()), '/a') + self.assertEqual(str(P('/a/b').absolute()), '/a/b') + + # '//'-prefixed absolute path (supported by POSIX). + self.assertEqual(str(P('//').absolute()), '//') + self.assertEqual(str(P('//a').absolute()), '//a') + self.assertEqual(str(P('//a/b').absolute()), '//a/b') + + @unittest.skipIf( + is_emscripten or is_wasi, + "umask is not implemented on Emscripten/WASI." + ) + @needs_posix + def test_open_mode(self): + # Unmask all permissions except world-write, which may + # not be supported on some filesystems (see GH-85633.) + old_mask = os.umask(0o002) + self.addCleanup(os.umask, old_mask) + p = self.cls(self.base) + with (p / 'new_file').open('wb'): + pass + st = os.stat(self.parser.join(self.base, 'new_file')) + self.assertEqual(stat.S_IMODE(st.st_mode), 0o664) + os.umask(0o026) + with (p / 'other_new_file').open('wb'): + pass + st = os.stat(self.parser.join(self.base, 'other_new_file')) + self.assertEqual(stat.S_IMODE(st.st_mode), 0o640) + + @needs_posix + def test_resolve_root(self): + current_directory = os.getcwd() + try: + os.chdir('/') + p = self.cls('spam') + self.assertEqual(str(p.resolve()), '/spam') + finally: + os.chdir(current_directory) + + @unittest.skipIf( + is_emscripten or is_wasi, + "umask is not implemented on Emscripten/WASI." + ) + @needs_posix + def test_touch_mode(self): + # Unmask all permissions except world-write, which may + # not be supported on some filesystems (see GH-85633.) + old_mask = os.umask(0o002) + self.addCleanup(os.umask, old_mask) + p = self.cls(self.base) + (p / 'new_file').touch() + st = os.stat(self.parser.join(self.base, 'new_file')) + self.assertEqual(stat.S_IMODE(st.st_mode), 0o664) + os.umask(0o026) + (p / 'other_new_file').touch() + st = os.stat(self.parser.join(self.base, 'other_new_file')) + self.assertEqual(stat.S_IMODE(st.st_mode), 0o640) + (p / 'masked_new_file').touch(mode=0o750) + st = os.stat(self.parser.join(self.base, 'masked_new_file')) + self.assertEqual(stat.S_IMODE(st.st_mode), 0o750) + + @unittest.skipUnless(hasattr(pwd, 'getpwall'), + 'pwd module does not expose getpwall()') + @unittest.skipIf(sys.platform == "vxworks", + "no home directory on VxWorks") + @needs_posix + def test_expanduser_posix(self): + P = self.cls + import_helper.import_module('pwd') + import pwd + pwdent = pwd.getpwuid(os.getuid()) + username = pwdent.pw_name + userhome = pwdent.pw_dir.rstrip('/') or '/' + # Find arbitrary different user (if exists). + for pwdent in pwd.getpwall(): + othername = pwdent.pw_name + otherhome = pwdent.pw_dir.rstrip('/') + if othername != username and otherhome: + break + else: + othername = username + otherhome = userhome + + fakename = 'fakeuser' + # This user can theoretically exist on a test runner. Create unique name: + try: + while pwd.getpwnam(fakename): + fakename += '1' + except KeyError: + pass # Non-existent name found + + p1 = P('~/Documents') + p2 = P(f'~{username}/Documents') + p3 = P(f'~{othername}/Documents') + p4 = P(f'../~{username}/Documents') + p5 = P(f'/~{username}/Documents') + p6 = P('') + p7 = P(f'~{fakename}/Documents') + + with os_helper.EnvironmentVarGuard() as env: + env.unset('HOME') + + self.assertEqual(p1.expanduser(), P(userhome) / 'Documents') + self.assertEqual(p2.expanduser(), P(userhome) / 'Documents') + self.assertEqual(p3.expanduser(), P(otherhome) / 'Documents') + self.assertEqual(p4.expanduser(), p4) + self.assertEqual(p5.expanduser(), p5) + self.assertEqual(p6.expanduser(), p6) + self.assertRaises(RuntimeError, p7.expanduser) + + env['HOME'] = '/tmp' + self.assertEqual(p1.expanduser(), P('/tmp/Documents')) + self.assertEqual(p2.expanduser(), P(userhome) / 'Documents') + self.assertEqual(p3.expanduser(), P(otherhome) / 'Documents') + self.assertEqual(p4.expanduser(), p4) + self.assertEqual(p5.expanduser(), p5) + self.assertEqual(p6.expanduser(), p6) + self.assertRaises(RuntimeError, p7.expanduser) + + @unittest.skipIf(sys.platform != "darwin", + "Bad file descriptor in /dev/fd affects only macOS") + @needs_posix + def test_handling_bad_descriptor(self): + try: + file_descriptors = list(pathlib.Path('/dev/fd').rglob("*"))[3:] + if not file_descriptors: + self.skipTest("no file descriptors - issue was not reproduced") + # Checking all file descriptors because there is no guarantee + # which one will fail. + for f in file_descriptors: + f.exists() + f.is_dir() + f.is_file() + f.is_symlink() + f.is_block_device() + f.is_char_device() + f.is_fifo() + f.is_socket() + except OSError as e: + if e.errno == errno.EBADF: + self.fail("Bad file descriptor not handled.") + raise + + @needs_posix + def test_from_uri_posix(self): + P = self.cls + self.assertEqual(P.from_uri('file:/foo/bar'), P('/foo/bar')) + self.assertEqual(P.from_uri('file://foo/bar'), P('//foo/bar')) + self.assertEqual(P.from_uri('file:///foo/bar'), P('/foo/bar')) + self.assertEqual(P.from_uri('file:////foo/bar'), P('//foo/bar')) + self.assertEqual(P.from_uri('file://localhost/foo/bar'), P('/foo/bar')) + self.assertRaises(ValueError, P.from_uri, 'foo/bar') + self.assertRaises(ValueError, P.from_uri, '/foo/bar') + self.assertRaises(ValueError, P.from_uri, '//foo/bar') + self.assertRaises(ValueError, P.from_uri, 'file:foo/bar') + self.assertRaises(ValueError, P.from_uri, 'http://foo/bar') + + @needs_posix + def test_from_uri_pathname2url_posix(self): + P = self.cls + self.assertEqual(P.from_uri('file:' + pathname2url('/foo/bar')), P('/foo/bar')) + self.assertEqual(P.from_uri('file:' + pathname2url('//foo/bar')), P('//foo/bar')) + + @needs_windows + def test_absolute_windows(self): + P = self.cls + + # Simple absolute paths. + self.assertEqual(str(P('c:\\').absolute()), 'c:\\') + self.assertEqual(str(P('c:\\a').absolute()), 'c:\\a') + self.assertEqual(str(P('c:\\a\\b').absolute()), 'c:\\a\\b') + + # UNC absolute paths. + share = '\\\\server\\share\\' + self.assertEqual(str(P(share).absolute()), share) + self.assertEqual(str(P(share + 'a').absolute()), share + 'a') + self.assertEqual(str(P(share + 'a\\b').absolute()), share + 'a\\b') + + # UNC relative paths. + with mock.patch("os.getcwd") as getcwd: + getcwd.return_value = share + + self.assertEqual(str(P().absolute()), share) + self.assertEqual(str(P('.').absolute()), share) + self.assertEqual(str(P('a').absolute()), os.path.join(share, 'a')) + self.assertEqual(str(P('a', 'b', 'c').absolute()), + os.path.join(share, 'a', 'b', 'c')) + + drive = os.path.splitdrive(self.base)[0] + with os_helper.change_cwd(self.base): + # Relative path with root + self.assertEqual(str(P('\\').absolute()), drive + '\\') + self.assertEqual(str(P('\\foo').absolute()), drive + '\\foo') + + # Relative path on current drive + self.assertEqual(str(P(drive).absolute()), self.base) + self.assertEqual(str(P(drive + 'foo').absolute()), os.path.join(self.base, 'foo')) + + with os_helper.subst_drive(self.base) as other_drive: + # Set the working directory on the substitute drive + saved_cwd = os.getcwd() + other_cwd = f'{other_drive}\\dirA' + os.chdir(other_cwd) + os.chdir(saved_cwd) + + # Relative path on another drive + self.assertEqual(str(P(other_drive).absolute()), other_cwd) + self.assertEqual(str(P(other_drive + 'foo').absolute()), other_cwd + '\\foo') + + @needs_windows + def test_expanduser_windows(self): + P = self.cls + with os_helper.EnvironmentVarGuard() as env: + env.unset('HOME', 'USERPROFILE', 'HOMEPATH', 'HOMEDRIVE') + env['USERNAME'] = 'alice' + + # test that the path returns unchanged + p1 = P('~/My Documents') + p2 = P('~alice/My Documents') + p3 = P('~bob/My Documents') + p4 = P('/~/My Documents') + p5 = P('d:~/My Documents') + p6 = P('') + self.assertRaises(RuntimeError, p1.expanduser) + self.assertRaises(RuntimeError, p2.expanduser) + self.assertRaises(RuntimeError, p3.expanduser) + self.assertEqual(p4.expanduser(), p4) + self.assertEqual(p5.expanduser(), p5) + self.assertEqual(p6.expanduser(), p6) + + def check(): + env.pop('USERNAME', None) + self.assertEqual(p1.expanduser(), + P('C:/Users/alice/My Documents')) + self.assertRaises(RuntimeError, p2.expanduser) + env['USERNAME'] = 'alice' + self.assertEqual(p2.expanduser(), + P('C:/Users/alice/My Documents')) + self.assertEqual(p3.expanduser(), + P('C:/Users/bob/My Documents')) + self.assertEqual(p4.expanduser(), p4) + self.assertEqual(p5.expanduser(), p5) + self.assertEqual(p6.expanduser(), p6) + + env['HOMEPATH'] = 'C:\\Users\\alice' + check() + + env['HOMEDRIVE'] = 'C:\\' + env['HOMEPATH'] = 'Users\\alice' + check() + + env.unset('HOMEDRIVE', 'HOMEPATH') + env['USERPROFILE'] = 'C:\\Users\\alice' + check() + + # bpo-38883: ignore `HOME` when set on windows + env['HOME'] = 'C:\\Users\\eve' + check() + + @needs_windows + def test_from_uri_windows(self): + P = self.cls + # DOS drive paths + self.assertEqual(P.from_uri('file:c:/path/to/file'), P('c:/path/to/file')) + self.assertEqual(P.from_uri('file:c|/path/to/file'), P('c:/path/to/file')) + self.assertEqual(P.from_uri('file:/c|/path/to/file'), P('c:/path/to/file')) + self.assertEqual(P.from_uri('file:///c|/path/to/file'), P('c:/path/to/file')) + # UNC paths + self.assertEqual(P.from_uri('file://server/path/to/file'), P('//server/path/to/file')) + self.assertEqual(P.from_uri('file:////server/path/to/file'), P('//server/path/to/file')) + self.assertEqual(P.from_uri('file://///server/path/to/file'), P('//server/path/to/file')) + # Localhost paths + self.assertEqual(P.from_uri('file://localhost/c:/path/to/file'), P('c:/path/to/file')) + self.assertEqual(P.from_uri('file://localhost/c|/path/to/file'), P('c:/path/to/file')) + # Invalid paths + self.assertRaises(ValueError, P.from_uri, 'foo/bar') + self.assertRaises(ValueError, P.from_uri, 'c:/foo/bar') + self.assertRaises(ValueError, P.from_uri, '//foo/bar') + self.assertRaises(ValueError, P.from_uri, 'file:foo/bar') + self.assertRaises(ValueError, P.from_uri, 'http://foo/bar') + + @needs_windows + def test_from_uri_pathname2url_windows(self): + P = self.cls + self.assertEqual(P.from_uri('file:' + pathname2url(r'c:\path\to\file')), P('c:/path/to/file')) + self.assertEqual(P.from_uri('file:' + pathname2url(r'\\server\path\to\file')), P('//server/path/to/file')) + + @needs_windows + def test_owner_windows(self): + P = self.cls + with self.assertRaises(pathlib.UnsupportedOperation): + P('c:/').owner() + + @needs_windows + def test_group_windows(self): + P = self.cls + with self.assertRaises(pathlib.UnsupportedOperation): + P('c:/').group() + + +@unittest.skipIf(os.name == 'nt', 'test requires a POSIX-compatible system') +class PosixPathTest(PathTest, PurePosixPathTest): + cls = pathlib.PosixPath + + +@unittest.skipIf(os.name != 'nt', 'test requires a Windows-compatible system') +class WindowsPathTest(PathTest, PureWindowsPathTest): + cls = pathlib.WindowsPath + + +class PathSubclassTest(PathTest): + class cls(pathlib.Path): + pass + + # repr() roundtripping is not supported in custom subclass. + test_repr_roundtrips = None + + +class CompatiblePathTest(unittest.TestCase): + """ + Test that a type can be made compatible with PurePath + derivatives by implementing division operator overloads. + """ + + class CompatPath: + """ + Minimum viable class to test PurePath compatibility. + Simply uses the division operator to join a given + string and the string value of another object with + a forward slash. + """ + def __init__(self, string): + self.string = string + + def __truediv__(self, other): + return type(self)(f"{self.string}/{other}") + + def __rtruediv__(self, other): + return type(self)(f"{other}/{self.string}") + + def test_truediv(self): + result = pathlib.PurePath("test") / self.CompatPath("right") + self.assertIsInstance(result, self.CompatPath) + self.assertEqual(result.string, "test/right") + + with self.assertRaises(TypeError): + # Verify improper operations still raise a TypeError + pathlib.PurePath("test") / 10 + + def test_rtruediv(self): + result = self.CompatPath("left") / pathlib.PurePath("test") + self.assertIsInstance(result, self.CompatPath) + self.assertEqual(result.string, "left/test") + + with self.assertRaises(TypeError): + # Verify improper operations still raise a TypeError + 10 / pathlib.PurePath("test") + + +if __name__ == "__main__": + unittest.main() diff --git a/Lib/test/test_pathlib/test_pathlib_abc.py b/Lib/test/test_pathlib/test_pathlib_abc.py new file mode 100644 index 00000000000..2e050362158 --- /dev/null +++ b/Lib/test/test_pathlib/test_pathlib_abc.py @@ -0,0 +1,2499 @@ +# TODO: RUSTPYTHON +# Has not been tested with Windows style paths +import collections +import io +import os +import errno +import stat +import unittest + +from pathlib._abc import UnsupportedOperation, ParserBase, PurePathBase, PathBase +import posixpath + +from test.support import is_wasi +from test.support.os_helper import TESTFN + + +_tests_needing_posix = set() +_tests_needing_windows = set() +_tests_needing_symlinks = set() + + +def needs_posix(fn): + """Decorator that marks a test as requiring a POSIX-flavoured path class.""" + _tests_needing_posix.add(fn.__name__) + return fn + +def needs_windows(fn): + """Decorator that marks a test as requiring a Windows-flavoured path class.""" + _tests_needing_windows.add(fn.__name__) + return fn + +def needs_symlinks(fn): + """Decorator that marks a test as requiring a path class that supports symlinks.""" + _tests_needing_symlinks.add(fn.__name__) + return fn + + +class UnsupportedOperationTest(unittest.TestCase): + def test_is_notimplemented(self): + self.assertTrue(issubclass(UnsupportedOperation, NotImplementedError)) + self.assertTrue(isinstance(UnsupportedOperation(), NotImplementedError)) + + +class ParserBaseTest(unittest.TestCase): + cls = ParserBase + + def test_unsupported_operation(self): + m = self.cls() + e = UnsupportedOperation + with self.assertRaises(e): + m.sep + self.assertRaises(e, m.join, 'foo') + self.assertRaises(e, m.split, 'foo') + self.assertRaises(e, m.splitdrive, 'foo') + self.assertRaises(e, m.normcase, 'foo') + self.assertRaises(e, m.isabs, 'foo') + +# +# Tests for the pure classes. +# + + +class PurePathBaseTest(unittest.TestCase): + cls = PurePathBase + + def test_unsupported_operation_pure(self): + p = self.cls('foo') + e = UnsupportedOperation + with self.assertRaises(e): + p.drive + with self.assertRaises(e): + p.root + with self.assertRaises(e): + p.anchor + with self.assertRaises(e): + p.parts + with self.assertRaises(e): + p.parent + with self.assertRaises(e): + p.parents + with self.assertRaises(e): + p.name + with self.assertRaises(e): + p.stem + with self.assertRaises(e): + p.suffix + with self.assertRaises(e): + p.suffixes + with self.assertRaises(e): + p / 'bar' + with self.assertRaises(e): + 'bar' / p + self.assertRaises(e, p.joinpath, 'bar') + self.assertRaises(e, p.with_name, 'bar') + self.assertRaises(e, p.with_stem, 'bar') + self.assertRaises(e, p.with_suffix, '.txt') + self.assertRaises(e, p.relative_to, '') + self.assertRaises(e, p.is_relative_to, '') + self.assertRaises(e, p.is_absolute) + self.assertRaises(e, p.match, '*') + + def test_magic_methods(self): + P = self.cls + self.assertFalse(hasattr(P, '__fspath__')) + self.assertFalse(hasattr(P, '__bytes__')) + self.assertIs(P.__reduce__, object.__reduce__) + self.assertIs(P.__repr__, object.__repr__) + self.assertIs(P.__hash__, object.__hash__) + self.assertIs(P.__eq__, object.__eq__) + self.assertIs(P.__lt__, object.__lt__) + self.assertIs(P.__le__, object.__le__) + self.assertIs(P.__gt__, object.__gt__) + self.assertIs(P.__ge__, object.__ge__) + + def test_parser(self): + self.assertIsInstance(self.cls.parser, ParserBase) + + +class DummyPurePath(PurePathBase): + __slots__ = () + parser = posixpath + + def __eq__(self, other): + if not isinstance(other, DummyPurePath): + return NotImplemented + return str(self) == str(other) + + def __hash__(self): + return hash(str(self)) + + def __repr__(self): + return "{}({!r})".format(self.__class__.__name__, self.as_posix()) + + +class DummyPurePathTest(unittest.TestCase): + cls = DummyPurePath + + # Use a base path that's unrelated to any real filesystem path. + base = f'/this/path/kills/fascists/{TESTFN}' + + def setUp(self): + name = self.id().split('.')[-1] + if name in _tests_needing_posix and self.cls.parser is not posixpath: + self.skipTest('requires POSIX-flavoured path class') + if name in _tests_needing_windows and self.cls.parser is posixpath: + self.skipTest('requires Windows-flavoured path class') + p = self.cls('a') + self.parser = p.parser + self.sep = self.parser.sep + self.altsep = self.parser.altsep + + def test_constructor_common(self): + P = self.cls + p = P('a') + self.assertIsInstance(p, P) + P('a', 'b', 'c') + P('/a', 'b', 'c') + P('a/b/c') + P('/a/b/c') + + def test_bytes(self): + P = self.cls + with self.assertRaises(TypeError): + P(b'a') + with self.assertRaises(TypeError): + P(b'a', 'b') + with self.assertRaises(TypeError): + P('a', b'b') + with self.assertRaises(TypeError): + P('a').joinpath(b'b') + with self.assertRaises(TypeError): + P('a') / b'b' + with self.assertRaises(TypeError): + b'a' / P('b') + with self.assertRaises(TypeError): + P('a').match(b'b') + with self.assertRaises(TypeError): + P('a').relative_to(b'b') + with self.assertRaises(TypeError): + P('a').with_name(b'b') + with self.assertRaises(TypeError): + P('a').with_stem(b'b') + with self.assertRaises(TypeError): + P('a').with_suffix(b'b') + + def _check_str_subclass(self, *args): + # Issue #21127: it should be possible to construct a PurePath object + # from a str subclass instance, and it then gets converted to + # a pure str object. + class StrSubclass(str): + pass + P = self.cls + p = P(*(StrSubclass(x) for x in args)) + self.assertEqual(p, P(*args)) + for part in p.parts: + self.assertIs(type(part), str) + + def test_str_subclass_common(self): + self._check_str_subclass('') + self._check_str_subclass('.') + self._check_str_subclass('a') + self._check_str_subclass('a/b.txt') + self._check_str_subclass('/a/b.txt') + + @needs_windows + def test_str_subclass_windows(self): + self._check_str_subclass('.\\a:b') + self._check_str_subclass('c:') + self._check_str_subclass('c:a') + self._check_str_subclass('c:a\\b.txt') + self._check_str_subclass('c:\\') + self._check_str_subclass('c:\\a') + self._check_str_subclass('c:\\a\\b.txt') + self._check_str_subclass('\\\\some\\share') + self._check_str_subclass('\\\\some\\share\\a') + self._check_str_subclass('\\\\some\\share\\a\\b.txt') + + def test_with_segments_common(self): + class P(self.cls): + def __init__(self, *pathsegments, session_id): + super().__init__(*pathsegments) + self.session_id = session_id + + def with_segments(self, *pathsegments): + return type(self)(*pathsegments, session_id=self.session_id) + p = P('foo', 'bar', session_id=42) + self.assertEqual(42, (p / 'foo').session_id) + self.assertEqual(42, ('foo' / p).session_id) + self.assertEqual(42, p.joinpath('foo').session_id) + self.assertEqual(42, p.with_name('foo').session_id) + self.assertEqual(42, p.with_stem('foo').session_id) + self.assertEqual(42, p.with_suffix('.foo').session_id) + self.assertEqual(42, p.with_segments('foo').session_id) + self.assertEqual(42, p.relative_to('foo').session_id) + self.assertEqual(42, p.parent.session_id) + for parent in p.parents: + self.assertEqual(42, parent.session_id) + + def test_join_common(self): + P = self.cls + p = P('a/b') + pp = p.joinpath('c') + self.assertEqual(pp, P('a/b/c')) + self.assertIs(type(pp), type(p)) + pp = p.joinpath('c', 'd') + self.assertEqual(pp, P('a/b/c/d')) + pp = p.joinpath('/c') + self.assertEqual(pp, P('/c')) + + @needs_posix + def test_join_posix(self): + P = self.cls + p = P('//a') + pp = p.joinpath('b') + self.assertEqual(pp, P('//a/b')) + pp = P('/a').joinpath('//c') + self.assertEqual(pp, P('//c')) + pp = P('//a').joinpath('/c') + self.assertEqual(pp, P('/c')) + + @needs_windows + def test_join_windows(self): + P = self.cls + p = P('C:/a/b') + pp = p.joinpath('x/y') + self.assertEqual(pp, P('C:/a/b/x/y')) + pp = p.joinpath('/x/y') + self.assertEqual(pp, P('C:/x/y')) + # Joining with a different drive => the first path is ignored, even + # if the second path is relative. + pp = p.joinpath('D:x/y') + self.assertEqual(pp, P('D:x/y')) + pp = p.joinpath('D:/x/y') + self.assertEqual(pp, P('D:/x/y')) + pp = p.joinpath('//host/share/x/y') + self.assertEqual(pp, P('//host/share/x/y')) + # Joining with the same drive => the first path is appended to if + # the second path is relative. + pp = p.joinpath('c:x/y') + self.assertEqual(pp, P('C:/a/b/x/y')) + pp = p.joinpath('c:/x/y') + self.assertEqual(pp, P('C:/x/y')) + # Joining with files with NTFS data streams => the filename should + # not be parsed as a drive letter + pp = p.joinpath(P('./d:s')) + self.assertEqual(pp, P('C:/a/b/d:s')) + pp = p.joinpath(P('./dd:s')) + self.assertEqual(pp, P('C:/a/b/dd:s')) + pp = p.joinpath(P('E:d:s')) + self.assertEqual(pp, P('E:d:s')) + # Joining onto a UNC path with no root + pp = P('//').joinpath('server') + self.assertEqual(pp, P('//server')) + pp = P('//server').joinpath('share') + self.assertEqual(pp, P('//server/share')) + pp = P('//./BootPartition').joinpath('Windows') + self.assertEqual(pp, P('//./BootPartition/Windows')) + + def test_div_common(self): + # Basically the same as joinpath(). + P = self.cls + p = P('a/b') + pp = p / 'c' + self.assertEqual(pp, P('a/b/c')) + self.assertIs(type(pp), type(p)) + pp = p / 'c/d' + self.assertEqual(pp, P('a/b/c/d')) + pp = p / 'c' / 'd' + self.assertEqual(pp, P('a/b/c/d')) + pp = 'c' / p / 'd' + self.assertEqual(pp, P('c/a/b/d')) + pp = p/ '/c' + self.assertEqual(pp, P('/c')) + + @needs_posix + def test_div_posix(self): + # Basically the same as joinpath(). + P = self.cls + p = P('//a') + pp = p / 'b' + self.assertEqual(pp, P('//a/b')) + pp = P('/a') / '//c' + self.assertEqual(pp, P('//c')) + pp = P('//a') / '/c' + self.assertEqual(pp, P('/c')) + + @needs_windows + def test_div_windows(self): + # Basically the same as joinpath(). + P = self.cls + p = P('C:/a/b') + self.assertEqual(p / 'x/y', P('C:/a/b/x/y')) + self.assertEqual(p / 'x' / 'y', P('C:/a/b/x/y')) + self.assertEqual(p / '/x/y', P('C:/x/y')) + self.assertEqual(p / '/x' / 'y', P('C:/x/y')) + # Joining with a different drive => the first path is ignored, even + # if the second path is relative. + self.assertEqual(p / 'D:x/y', P('D:x/y')) + self.assertEqual(p / 'D:' / 'x/y', P('D:x/y')) + self.assertEqual(p / 'D:/x/y', P('D:/x/y')) + self.assertEqual(p / 'D:' / '/x/y', P('D:/x/y')) + self.assertEqual(p / '//host/share/x/y', P('//host/share/x/y')) + # Joining with the same drive => the first path is appended to if + # the second path is relative. + self.assertEqual(p / 'c:x/y', P('C:/a/b/x/y')) + self.assertEqual(p / 'c:/x/y', P('C:/x/y')) + # Joining with files with NTFS data streams => the filename should + # not be parsed as a drive letter + self.assertEqual(p / P('./d:s'), P('C:/a/b/d:s')) + self.assertEqual(p / P('./dd:s'), P('C:/a/b/dd:s')) + self.assertEqual(p / P('E:d:s'), P('E:d:s')) + + def _check_str(self, expected, args): + p = self.cls(*args) + self.assertEqual(str(p), expected.replace('/', self.sep)) + + def test_str_common(self): + # Canonicalized paths roundtrip. + for pathstr in ('a', 'a/b', 'a/b/c', '/', '/a/b', '/a/b/c'): + self._check_str(pathstr, (pathstr,)) + # Other tests for str() are in test_equivalences(). + + @needs_windows + def test_str_windows(self): + p = self.cls('a/b/c') + self.assertEqual(str(p), 'a\\b\\c') + p = self.cls('c:/a/b/c') + self.assertEqual(str(p), 'c:\\a\\b\\c') + p = self.cls('//a/b') + self.assertEqual(str(p), '\\\\a\\b\\') + p = self.cls('//a/b/c') + self.assertEqual(str(p), '\\\\a\\b\\c') + p = self.cls('//a/b/c/d') + self.assertEqual(str(p), '\\\\a\\b\\c\\d') + + def test_as_posix_common(self): + P = self.cls + for pathstr in ('a', 'a/b', 'a/b/c', '/', '/a/b', '/a/b/c'): + self.assertEqual(P(pathstr).as_posix(), pathstr) + # Other tests for as_posix() are in test_equivalences(). + + def test_match_empty(self): + P = self.cls + self.assertRaises(ValueError, P('a').match, '') + + def test_match_common(self): + P = self.cls + # Simple relative pattern. + self.assertTrue(P('b.py').match('b.py')) + self.assertTrue(P('a/b.py').match('b.py')) + self.assertTrue(P('/a/b.py').match('b.py')) + self.assertFalse(P('a.py').match('b.py')) + self.assertFalse(P('b/py').match('b.py')) + self.assertFalse(P('/a.py').match('b.py')) + self.assertFalse(P('b.py/c').match('b.py')) + # Wildcard relative pattern. + self.assertTrue(P('b.py').match('*.py')) + self.assertTrue(P('a/b.py').match('*.py')) + self.assertTrue(P('/a/b.py').match('*.py')) + self.assertFalse(P('b.pyc').match('*.py')) + self.assertFalse(P('b./py').match('*.py')) + self.assertFalse(P('b.py/c').match('*.py')) + # Multi-part relative pattern. + self.assertTrue(P('ab/c.py').match('a*/*.py')) + self.assertTrue(P('/d/ab/c.py').match('a*/*.py')) + self.assertFalse(P('a.py').match('a*/*.py')) + self.assertFalse(P('/dab/c.py').match('a*/*.py')) + self.assertFalse(P('ab/c.py/d').match('a*/*.py')) + # Absolute pattern. + self.assertTrue(P('/b.py').match('/*.py')) + self.assertFalse(P('b.py').match('/*.py')) + self.assertFalse(P('a/b.py').match('/*.py')) + self.assertFalse(P('/a/b.py').match('/*.py')) + # Multi-part absolute pattern. + self.assertTrue(P('/a/b.py').match('/a/*.py')) + self.assertFalse(P('/ab.py').match('/a/*.py')) + self.assertFalse(P('/a/b/c.py').match('/a/*.py')) + # Multi-part glob-style pattern. + self.assertFalse(P('/a/b/c.py').match('/**/*.py')) + self.assertTrue(P('/a/b/c.py').match('/a/**/*.py')) + # Case-sensitive flag + self.assertFalse(P('A.py').match('a.PY', case_sensitive=True)) + self.assertTrue(P('A.py').match('a.PY', case_sensitive=False)) + self.assertFalse(P('c:/a/B.Py').match('C:/A/*.pY', case_sensitive=True)) + self.assertTrue(P('/a/b/c.py').match('/A/*/*.Py', case_sensitive=False)) + # Matching against empty path + self.assertFalse(P('').match('*')) + self.assertFalse(P('').match('**')) + self.assertFalse(P('').match('**/*')) + + @needs_posix + def test_match_posix(self): + P = self.cls + self.assertFalse(P('A.py').match('a.PY')) + + @needs_windows + def test_match_windows(self): + P = self.cls + # Absolute patterns. + self.assertTrue(P('c:/b.py').match('*:/*.py')) + self.assertTrue(P('c:/b.py').match('c:/*.py')) + self.assertFalse(P('d:/b.py').match('c:/*.py')) # wrong drive + self.assertFalse(P('b.py').match('/*.py')) + self.assertFalse(P('b.py').match('c:*.py')) + self.assertFalse(P('b.py').match('c:/*.py')) + self.assertFalse(P('c:b.py').match('/*.py')) + self.assertFalse(P('c:b.py').match('c:/*.py')) + self.assertFalse(P('/b.py').match('c:*.py')) + self.assertFalse(P('/b.py').match('c:/*.py')) + # UNC patterns. + self.assertTrue(P('//some/share/a.py').match('//*/*/*.py')) + self.assertTrue(P('//some/share/a.py').match('//some/share/*.py')) + self.assertFalse(P('//other/share/a.py').match('//some/share/*.py')) + self.assertFalse(P('//some/share/a/b.py').match('//some/share/*.py')) + # Case-insensitivity. + self.assertTrue(P('B.py').match('b.PY')) + self.assertTrue(P('c:/a/B.Py').match('C:/A/*.pY')) + self.assertTrue(P('//Some/Share/B.Py').match('//somE/sharE/*.pY')) + # Path anchor doesn't match pattern anchor + self.assertFalse(P('c:/b.py').match('/*.py')) # 'c:/' vs '/' + self.assertFalse(P('c:/b.py').match('c:*.py')) # 'c:/' vs 'c:' + self.assertFalse(P('//some/share/a.py').match('/*.py')) # '//some/share/' vs '/' + + def test_full_match_common(self): + P = self.cls + # Simple relative pattern. + self.assertTrue(P('b.py').full_match('b.py')) + self.assertFalse(P('a/b.py').full_match('b.py')) + self.assertFalse(P('/a/b.py').full_match('b.py')) + self.assertFalse(P('a.py').full_match('b.py')) + self.assertFalse(P('b/py').full_match('b.py')) + self.assertFalse(P('/a.py').full_match('b.py')) + self.assertFalse(P('b.py/c').full_match('b.py')) + # Wildcard relative pattern. + self.assertTrue(P('b.py').full_match('*.py')) + self.assertFalse(P('a/b.py').full_match('*.py')) + self.assertFalse(P('/a/b.py').full_match('*.py')) + self.assertFalse(P('b.pyc').full_match('*.py')) + self.assertFalse(P('b./py').full_match('*.py')) + self.assertFalse(P('b.py/c').full_match('*.py')) + # Multi-part relative pattern. + self.assertTrue(P('ab/c.py').full_match('a*/*.py')) + self.assertFalse(P('/d/ab/c.py').full_match('a*/*.py')) + self.assertFalse(P('a.py').full_match('a*/*.py')) + self.assertFalse(P('/dab/c.py').full_match('a*/*.py')) + self.assertFalse(P('ab/c.py/d').full_match('a*/*.py')) + # Absolute pattern. + self.assertTrue(P('/b.py').full_match('/*.py')) + self.assertFalse(P('b.py').full_match('/*.py')) + self.assertFalse(P('a/b.py').full_match('/*.py')) + self.assertFalse(P('/a/b.py').full_match('/*.py')) + # Multi-part absolute pattern. + self.assertTrue(P('/a/b.py').full_match('/a/*.py')) + self.assertFalse(P('/ab.py').full_match('/a/*.py')) + self.assertFalse(P('/a/b/c.py').full_match('/a/*.py')) + # Multi-part glob-style pattern. + self.assertTrue(P('a').full_match('**')) + self.assertTrue(P('c.py').full_match('**')) + self.assertTrue(P('a/b/c.py').full_match('**')) + self.assertTrue(P('/a/b/c.py').full_match('**')) + self.assertTrue(P('/a/b/c.py').full_match('/**')) + self.assertTrue(P('/a/b/c.py').full_match('/a/**')) + self.assertTrue(P('/a/b/c.py').full_match('**/*.py')) + self.assertTrue(P('/a/b/c.py').full_match('/**/*.py')) + self.assertTrue(P('/a/b/c.py').full_match('/a/**/*.py')) + self.assertTrue(P('/a/b/c.py').full_match('/a/b/**/*.py')) + self.assertTrue(P('/a/b/c.py').full_match('/**/**/**/**/*.py')) + self.assertFalse(P('c.py').full_match('**/a.py')) + self.assertFalse(P('c.py').full_match('c/**')) + self.assertFalse(P('a/b/c.py').full_match('**/a')) + self.assertFalse(P('a/b/c.py').full_match('**/a/b')) + self.assertFalse(P('a/b/c.py').full_match('**/a/b/c')) + self.assertFalse(P('a/b/c.py').full_match('**/a/b/c.')) + self.assertFalse(P('a/b/c.py').full_match('**/a/b/c./**')) + self.assertFalse(P('a/b/c.py').full_match('**/a/b/c./**')) + self.assertFalse(P('a/b/c.py').full_match('/a/b/c.py/**')) + self.assertFalse(P('a/b/c.py').full_match('/**/a/b/c.py')) + # Case-sensitive flag + self.assertFalse(P('A.py').full_match('a.PY', case_sensitive=True)) + self.assertTrue(P('A.py').full_match('a.PY', case_sensitive=False)) + self.assertFalse(P('c:/a/B.Py').full_match('C:/A/*.pY', case_sensitive=True)) + self.assertTrue(P('/a/b/c.py').full_match('/A/*/*.Py', case_sensitive=False)) + # Matching against empty path + self.assertFalse(P('').full_match('*')) + self.assertTrue(P('').full_match('**')) + self.assertFalse(P('').full_match('**/*')) + # Matching with empty pattern + self.assertTrue(P('').full_match('')) + self.assertTrue(P('.').full_match('.')) + self.assertFalse(P('/').full_match('')) + self.assertFalse(P('/').full_match('.')) + self.assertFalse(P('foo').full_match('')) + self.assertFalse(P('foo').full_match('.')) + + def test_parts_common(self): + # `parts` returns a tuple. + sep = self.sep + P = self.cls + p = P('a/b') + parts = p.parts + self.assertEqual(parts, ('a', 'b')) + # When the path is absolute, the anchor is a separate part. + p = P('/a/b') + parts = p.parts + self.assertEqual(parts, (sep, 'a', 'b')) + + @needs_windows + def test_parts_windows(self): + P = self.cls + p = P('c:a/b') + parts = p.parts + self.assertEqual(parts, ('c:', 'a', 'b')) + p = P('c:/a/b') + parts = p.parts + self.assertEqual(parts, ('c:\\', 'a', 'b')) + p = P('//a/b/c/d') + parts = p.parts + self.assertEqual(parts, ('\\\\a\\b\\', 'c', 'd')) + + def test_parent_common(self): + # Relative + P = self.cls + p = P('a/b/c') + self.assertEqual(p.parent, P('a/b')) + self.assertEqual(p.parent.parent, P('a')) + self.assertEqual(p.parent.parent.parent, P('')) + self.assertEqual(p.parent.parent.parent.parent, P('')) + # Anchored + p = P('/a/b/c') + self.assertEqual(p.parent, P('/a/b')) + self.assertEqual(p.parent.parent, P('/a')) + self.assertEqual(p.parent.parent.parent, P('/')) + self.assertEqual(p.parent.parent.parent.parent, P('/')) + + @needs_windows + def test_parent_windows(self): + # Anchored + P = self.cls + p = P('z:a/b/c') + self.assertEqual(p.parent, P('z:a/b')) + self.assertEqual(p.parent.parent, P('z:a')) + self.assertEqual(p.parent.parent.parent, P('z:')) + self.assertEqual(p.parent.parent.parent.parent, P('z:')) + p = P('z:/a/b/c') + self.assertEqual(p.parent, P('z:/a/b')) + self.assertEqual(p.parent.parent, P('z:/a')) + self.assertEqual(p.parent.parent.parent, P('z:/')) + self.assertEqual(p.parent.parent.parent.parent, P('z:/')) + p = P('//a/b/c/d') + self.assertEqual(p.parent, P('//a/b/c')) + self.assertEqual(p.parent.parent, P('//a/b')) + self.assertEqual(p.parent.parent.parent, P('//a/b')) + + def test_parents_common(self): + # Relative + P = self.cls + p = P('a/b/c') + par = p.parents + self.assertEqual(len(par), 3) + self.assertEqual(par[0], P('a/b')) + self.assertEqual(par[1], P('a')) + self.assertEqual(par[2], P('')) + self.assertEqual(par[-1], P('')) + self.assertEqual(par[-2], P('a')) + self.assertEqual(par[-3], P('a/b')) + self.assertEqual(par[0:1], (P('a/b'),)) + self.assertEqual(par[:2], (P('a/b'), P('a'))) + self.assertEqual(par[:-1], (P('a/b'), P('a'))) + self.assertEqual(par[1:], (P('a'), P(''))) + self.assertEqual(par[::2], (P('a/b'), P(''))) + self.assertEqual(par[::-1], (P(''), P('a'), P('a/b'))) + self.assertEqual(list(par), [P('a/b'), P('a'), P('')]) + with self.assertRaises(IndexError): + par[-4] + with self.assertRaises(IndexError): + par[3] + with self.assertRaises(TypeError): + par[0] = p + # Anchored + p = P('/a/b/c') + par = p.parents + self.assertEqual(len(par), 3) + self.assertEqual(par[0], P('/a/b')) + self.assertEqual(par[1], P('/a')) + self.assertEqual(par[2], P('/')) + self.assertEqual(par[-1], P('/')) + self.assertEqual(par[-2], P('/a')) + self.assertEqual(par[-3], P('/a/b')) + self.assertEqual(par[0:1], (P('/a/b'),)) + self.assertEqual(par[:2], (P('/a/b'), P('/a'))) + self.assertEqual(par[:-1], (P('/a/b'), P('/a'))) + self.assertEqual(par[1:], (P('/a'), P('/'))) + self.assertEqual(par[::2], (P('/a/b'), P('/'))) + self.assertEqual(par[::-1], (P('/'), P('/a'), P('/a/b'))) + self.assertEqual(list(par), [P('/a/b'), P('/a'), P('/')]) + with self.assertRaises(IndexError): + par[-4] + with self.assertRaises(IndexError): + par[3] + + @needs_windows + def test_parents_windows(self): + # Anchored + P = self.cls + p = P('z:a/b/') + par = p.parents + self.assertEqual(len(par), 2) + self.assertEqual(par[0], P('z:a')) + self.assertEqual(par[1], P('z:')) + self.assertEqual(par[0:1], (P('z:a'),)) + self.assertEqual(par[:-1], (P('z:a'),)) + self.assertEqual(par[:2], (P('z:a'), P('z:'))) + self.assertEqual(par[1:], (P('z:'),)) + self.assertEqual(par[::2], (P('z:a'),)) + self.assertEqual(par[::-1], (P('z:'), P('z:a'))) + self.assertEqual(list(par), [P('z:a'), P('z:')]) + with self.assertRaises(IndexError): + par[2] + p = P('z:/a/b/') + par = p.parents + self.assertEqual(len(par), 2) + self.assertEqual(par[0], P('z:/a')) + self.assertEqual(par[1], P('z:/')) + self.assertEqual(par[0:1], (P('z:/a'),)) + self.assertEqual(par[0:-1], (P('z:/a'),)) + self.assertEqual(par[:2], (P('z:/a'), P('z:/'))) + self.assertEqual(par[1:], (P('z:/'),)) + self.assertEqual(par[::2], (P('z:/a'),)) + self.assertEqual(par[::-1], (P('z:/'), P('z:/a'),)) + self.assertEqual(list(par), [P('z:/a'), P('z:/')]) + with self.assertRaises(IndexError): + par[2] + p = P('//a/b/c/d') + par = p.parents + self.assertEqual(len(par), 2) + self.assertEqual(par[0], P('//a/b/c')) + self.assertEqual(par[1], P('//a/b')) + self.assertEqual(par[0:1], (P('//a/b/c'),)) + self.assertEqual(par[0:-1], (P('//a/b/c'),)) + self.assertEqual(par[:2], (P('//a/b/c'), P('//a/b'))) + self.assertEqual(par[1:], (P('//a/b'),)) + self.assertEqual(par[::2], (P('//a/b/c'),)) + self.assertEqual(par[::-1], (P('//a/b'), P('//a/b/c'))) + self.assertEqual(list(par), [P('//a/b/c'), P('//a/b')]) + with self.assertRaises(IndexError): + par[2] + + def test_drive_common(self): + P = self.cls + self.assertEqual(P('a/b').drive, '') + self.assertEqual(P('/a/b').drive, '') + self.assertEqual(P('').drive, '') + + @needs_windows + def test_drive_windows(self): + P = self.cls + self.assertEqual(P('c:').drive, 'c:') + self.assertEqual(P('c:a/b').drive, 'c:') + self.assertEqual(P('c:/').drive, 'c:') + self.assertEqual(P('c:/a/b/').drive, 'c:') + self.assertEqual(P('//a/b').drive, '\\\\a\\b') + self.assertEqual(P('//a/b/').drive, '\\\\a\\b') + self.assertEqual(P('//a/b/c/d').drive, '\\\\a\\b') + self.assertEqual(P('./c:a').drive, '') + + def test_root_common(self): + P = self.cls + sep = self.sep + self.assertEqual(P('').root, '') + self.assertEqual(P('a/b').root, '') + self.assertEqual(P('/').root, sep) + self.assertEqual(P('/a/b').root, sep) + + @needs_posix + def test_root_posix(self): + P = self.cls + self.assertEqual(P('/a/b').root, '/') + # POSIX special case for two leading slashes. + self.assertEqual(P('//a/b').root, '//') + + @needs_windows + def test_root_windows(self): + P = self.cls + self.assertEqual(P('c:').root, '') + self.assertEqual(P('c:a/b').root, '') + self.assertEqual(P('c:/').root, '\\') + self.assertEqual(P('c:/a/b/').root, '\\') + self.assertEqual(P('//a/b').root, '\\') + self.assertEqual(P('//a/b/').root, '\\') + self.assertEqual(P('//a/b/c/d').root, '\\') + + def test_anchor_common(self): + P = self.cls + sep = self.sep + self.assertEqual(P('').anchor, '') + self.assertEqual(P('a/b').anchor, '') + self.assertEqual(P('/').anchor, sep) + self.assertEqual(P('/a/b').anchor, sep) + + @needs_windows + def test_anchor_windows(self): + P = self.cls + self.assertEqual(P('c:').anchor, 'c:') + self.assertEqual(P('c:a/b').anchor, 'c:') + self.assertEqual(P('c:/').anchor, 'c:\\') + self.assertEqual(P('c:/a/b/').anchor, 'c:\\') + self.assertEqual(P('//a/b').anchor, '\\\\a\\b\\') + self.assertEqual(P('//a/b/').anchor, '\\\\a\\b\\') + self.assertEqual(P('//a/b/c/d').anchor, '\\\\a\\b\\') + + def test_name_empty(self): + P = self.cls + self.assertEqual(P('').name, '') + self.assertEqual(P('.').name, '.') + self.assertEqual(P('/a/b/.').name, '.') + + def test_name_common(self): + P = self.cls + self.assertEqual(P('/').name, '') + self.assertEqual(P('a/b').name, 'b') + self.assertEqual(P('/a/b').name, 'b') + self.assertEqual(P('a/b.py').name, 'b.py') + self.assertEqual(P('/a/b.py').name, 'b.py') + + @needs_windows + def test_name_windows(self): + P = self.cls + self.assertEqual(P('c:').name, '') + self.assertEqual(P('c:/').name, '') + self.assertEqual(P('c:a/b').name, 'b') + self.assertEqual(P('c:/a/b').name, 'b') + self.assertEqual(P('c:a/b.py').name, 'b.py') + self.assertEqual(P('c:/a/b.py').name, 'b.py') + self.assertEqual(P('//My.py/Share.php').name, '') + self.assertEqual(P('//My.py/Share.php/a/b').name, 'b') + + def test_suffix_common(self): + P = self.cls + self.assertEqual(P('').suffix, '') + self.assertEqual(P('.').suffix, '') + self.assertEqual(P('..').suffix, '') + self.assertEqual(P('/').suffix, '') + self.assertEqual(P('a/b').suffix, '') + self.assertEqual(P('/a/b').suffix, '') + self.assertEqual(P('/a/b/.').suffix, '') + self.assertEqual(P('a/b.py').suffix, '.py') + self.assertEqual(P('/a/b.py').suffix, '.py') + self.assertEqual(P('a/.hgrc').suffix, '') + self.assertEqual(P('/a/.hgrc').suffix, '') + self.assertEqual(P('a/.hg.rc').suffix, '.rc') + self.assertEqual(P('/a/.hg.rc').suffix, '.rc') + self.assertEqual(P('a/b.tar.gz').suffix, '.gz') + self.assertEqual(P('/a/b.tar.gz').suffix, '.gz') + self.assertEqual(P('a/Some name. Ending with a dot.').suffix, '') + self.assertEqual(P('/a/Some name. Ending with a dot.').suffix, '') + + @needs_windows + def test_suffix_windows(self): + P = self.cls + self.assertEqual(P('c:').suffix, '') + self.assertEqual(P('c:/').suffix, '') + self.assertEqual(P('c:a/b').suffix, '') + self.assertEqual(P('c:/a/b').suffix, '') + self.assertEqual(P('c:a/b.py').suffix, '.py') + self.assertEqual(P('c:/a/b.py').suffix, '.py') + self.assertEqual(P('c:a/.hgrc').suffix, '') + self.assertEqual(P('c:/a/.hgrc').suffix, '') + self.assertEqual(P('c:a/.hg.rc').suffix, '.rc') + self.assertEqual(P('c:/a/.hg.rc').suffix, '.rc') + self.assertEqual(P('c:a/b.tar.gz').suffix, '.gz') + self.assertEqual(P('c:/a/b.tar.gz').suffix, '.gz') + self.assertEqual(P('c:a/Some name. Ending with a dot.').suffix, '') + self.assertEqual(P('c:/a/Some name. Ending with a dot.').suffix, '') + self.assertEqual(P('//My.py/Share.php').suffix, '') + self.assertEqual(P('//My.py/Share.php/a/b').suffix, '') + + def test_suffixes_common(self): + P = self.cls + self.assertEqual(P('').suffixes, []) + self.assertEqual(P('.').suffixes, []) + self.assertEqual(P('/').suffixes, []) + self.assertEqual(P('a/b').suffixes, []) + self.assertEqual(P('/a/b').suffixes, []) + self.assertEqual(P('/a/b/.').suffixes, []) + self.assertEqual(P('a/b.py').suffixes, ['.py']) + self.assertEqual(P('/a/b.py').suffixes, ['.py']) + self.assertEqual(P('a/.hgrc').suffixes, []) + self.assertEqual(P('/a/.hgrc').suffixes, []) + self.assertEqual(P('a/.hg.rc').suffixes, ['.rc']) + self.assertEqual(P('/a/.hg.rc').suffixes, ['.rc']) + self.assertEqual(P('a/b.tar.gz').suffixes, ['.tar', '.gz']) + self.assertEqual(P('/a/b.tar.gz').suffixes, ['.tar', '.gz']) + self.assertEqual(P('a/Some name. Ending with a dot.').suffixes, []) + self.assertEqual(P('/a/Some name. Ending with a dot.').suffixes, []) + + @needs_windows + def test_suffixes_windows(self): + P = self.cls + self.assertEqual(P('c:').suffixes, []) + self.assertEqual(P('c:/').suffixes, []) + self.assertEqual(P('c:a/b').suffixes, []) + self.assertEqual(P('c:/a/b').suffixes, []) + self.assertEqual(P('c:a/b.py').suffixes, ['.py']) + self.assertEqual(P('c:/a/b.py').suffixes, ['.py']) + self.assertEqual(P('c:a/.hgrc').suffixes, []) + self.assertEqual(P('c:/a/.hgrc').suffixes, []) + self.assertEqual(P('c:a/.hg.rc').suffixes, ['.rc']) + self.assertEqual(P('c:/a/.hg.rc').suffixes, ['.rc']) + self.assertEqual(P('c:a/b.tar.gz').suffixes, ['.tar', '.gz']) + self.assertEqual(P('c:/a/b.tar.gz').suffixes, ['.tar', '.gz']) + self.assertEqual(P('//My.py/Share.php').suffixes, []) + self.assertEqual(P('//My.py/Share.php/a/b').suffixes, []) + self.assertEqual(P('c:a/Some name. Ending with a dot.').suffixes, []) + self.assertEqual(P('c:/a/Some name. Ending with a dot.').suffixes, []) + + def test_stem_empty(self): + P = self.cls + self.assertEqual(P('').stem, '') + self.assertEqual(P('.').stem, '.') + + def test_stem_common(self): + P = self.cls + self.assertEqual(P('..').stem, '..') + self.assertEqual(P('/').stem, '') + self.assertEqual(P('a/b').stem, 'b') + self.assertEqual(P('a/b.py').stem, 'b') + self.assertEqual(P('a/.hgrc').stem, '.hgrc') + self.assertEqual(P('a/.hg.rc').stem, '.hg') + self.assertEqual(P('a/b.tar.gz').stem, 'b.tar') + self.assertEqual(P('a/Some name. Ending with a dot.').stem, + 'Some name. Ending with a dot.') + + @needs_windows + def test_stem_windows(self): + P = self.cls + self.assertEqual(P('c:').stem, '') + self.assertEqual(P('c:.').stem, '') + self.assertEqual(P('c:..').stem, '..') + self.assertEqual(P('c:/').stem, '') + self.assertEqual(P('c:a/b').stem, 'b') + self.assertEqual(P('c:a/b.py').stem, 'b') + self.assertEqual(P('c:a/.hgrc').stem, '.hgrc') + self.assertEqual(P('c:a/.hg.rc').stem, '.hg') + self.assertEqual(P('c:a/b.tar.gz').stem, 'b.tar') + self.assertEqual(P('c:a/Some name. Ending with a dot.').stem, + 'Some name. Ending with a dot.') + def test_with_name_common(self): + P = self.cls + self.assertEqual(P('a/b').with_name('d.xml'), P('a/d.xml')) + self.assertEqual(P('/a/b').with_name('d.xml'), P('/a/d.xml')) + self.assertEqual(P('a/b.py').with_name('d.xml'), P('a/d.xml')) + self.assertEqual(P('/a/b.py').with_name('d.xml'), P('/a/d.xml')) + self.assertEqual(P('a/Dot ending.').with_name('d.xml'), P('a/d.xml')) + self.assertEqual(P('/a/Dot ending.').with_name('d.xml'), P('/a/d.xml')) + + @needs_windows + def test_with_name_windows(self): + P = self.cls + self.assertEqual(P('c:a/b').with_name('d.xml'), P('c:a/d.xml')) + self.assertEqual(P('c:/a/b').with_name('d.xml'), P('c:/a/d.xml')) + self.assertEqual(P('c:a/Dot ending.').with_name('d.xml'), P('c:a/d.xml')) + self.assertEqual(P('c:/a/Dot ending.').with_name('d.xml'), P('c:/a/d.xml')) + self.assertRaises(ValueError, P('c:').with_name, 'd.xml') + self.assertRaises(ValueError, P('c:/').with_name, 'd.xml') + self.assertRaises(ValueError, P('//My/Share').with_name, 'd.xml') + self.assertEqual(str(P('a').with_name('d:')), '.\\d:') + self.assertEqual(str(P('a').with_name('d:e')), '.\\d:e') + self.assertEqual(P('c:a/b').with_name('d:'), P('c:a/d:')) + self.assertEqual(P('c:a/b').with_name('d:e'), P('c:a/d:e')) + self.assertRaises(ValueError, P('c:a/b').with_name, 'd:/e') + self.assertRaises(ValueError, P('c:a/b').with_name, '//My/Share') + + def test_with_name_empty(self): + P = self.cls + self.assertEqual(P('').with_name('d.xml'), P('d.xml')) + self.assertEqual(P('.').with_name('d.xml'), P('d.xml')) + self.assertEqual(P('/').with_name('d.xml'), P('/d.xml')) + self.assertEqual(P('a/b').with_name(''), P('a/')) + self.assertEqual(P('a/b').with_name('.'), P('a/.')) + + def test_with_name_seps(self): + P = self.cls + self.assertRaises(ValueError, P('a/b').with_name, '/c') + self.assertRaises(ValueError, P('a/b').with_name, 'c/') + self.assertRaises(ValueError, P('a/b').with_name, 'c/d') + + def test_with_stem_common(self): + P = self.cls + self.assertEqual(P('a/b').with_stem('d'), P('a/d')) + self.assertEqual(P('/a/b').with_stem('d'), P('/a/d')) + self.assertEqual(P('a/b.py').with_stem('d'), P('a/d.py')) + self.assertEqual(P('/a/b.py').with_stem('d'), P('/a/d.py')) + self.assertEqual(P('/a/b.tar.gz').with_stem('d'), P('/a/d.gz')) + self.assertEqual(P('a/Dot ending.').with_stem('d'), P('a/d')) + self.assertEqual(P('/a/Dot ending.').with_stem('d'), P('/a/d')) + + @needs_windows + def test_with_stem_windows(self): + P = self.cls + self.assertEqual(P('c:a/b').with_stem('d'), P('c:a/d')) + self.assertEqual(P('c:/a/b').with_stem('d'), P('c:/a/d')) + self.assertEqual(P('c:a/Dot ending.').with_stem('d'), P('c:a/d')) + self.assertEqual(P('c:/a/Dot ending.').with_stem('d'), P('c:/a/d')) + self.assertRaises(ValueError, P('c:').with_stem, 'd') + self.assertRaises(ValueError, P('c:/').with_stem, 'd') + self.assertRaises(ValueError, P('//My/Share').with_stem, 'd') + self.assertEqual(str(P('a').with_stem('d:')), '.\\d:') + self.assertEqual(str(P('a').with_stem('d:e')), '.\\d:e') + self.assertEqual(P('c:a/b').with_stem('d:'), P('c:a/d:')) + self.assertEqual(P('c:a/b').with_stem('d:e'), P('c:a/d:e')) + self.assertRaises(ValueError, P('c:a/b').with_stem, 'd:/e') + self.assertRaises(ValueError, P('c:a/b').with_stem, '//My/Share') + + def test_with_stem_empty(self): + P = self.cls + self.assertEqual(P('').with_stem('d'), P('d')) + self.assertEqual(P('.').with_stem('d'), P('d')) + self.assertEqual(P('/').with_stem('d'), P('/d')) + self.assertEqual(P('a/b').with_stem(''), P('a/')) + self.assertEqual(P('a/b').with_stem('.'), P('a/.')) + self.assertRaises(ValueError, P('foo.gz').with_stem, '') + self.assertRaises(ValueError, P('/a/b/foo.gz').with_stem, '') + + def test_with_stem_seps(self): + P = self.cls + self.assertRaises(ValueError, P('a/b').with_stem, '/c') + self.assertRaises(ValueError, P('a/b').with_stem, 'c/') + self.assertRaises(ValueError, P('a/b').with_stem, 'c/d') + + def test_with_suffix_common(self): + P = self.cls + self.assertEqual(P('a/b').with_suffix('.gz'), P('a/b.gz')) + self.assertEqual(P('/a/b').with_suffix('.gz'), P('/a/b.gz')) + self.assertEqual(P('a/b.py').with_suffix('.gz'), P('a/b.gz')) + self.assertEqual(P('/a/b.py').with_suffix('.gz'), P('/a/b.gz')) + # Stripping suffix. + self.assertEqual(P('a/b.py').with_suffix(''), P('a/b')) + self.assertEqual(P('/a/b').with_suffix(''), P('/a/b')) + + @needs_windows + def test_with_suffix_windows(self): + P = self.cls + self.assertEqual(P('c:a/b').with_suffix('.gz'), P('c:a/b.gz')) + self.assertEqual(P('c:/a/b').with_suffix('.gz'), P('c:/a/b.gz')) + self.assertEqual(P('c:a/b.py').with_suffix('.gz'), P('c:a/b.gz')) + self.assertEqual(P('c:/a/b.py').with_suffix('.gz'), P('c:/a/b.gz')) + # Path doesn't have a "filename" component. + self.assertRaises(ValueError, P('').with_suffix, '.gz') + self.assertRaises(ValueError, P('.').with_suffix, '.gz') + self.assertRaises(ValueError, P('/').with_suffix, '.gz') + self.assertRaises(ValueError, P('//My/Share').with_suffix, '.gz') + # Invalid suffix. + self.assertRaises(ValueError, P('c:a/b').with_suffix, 'gz') + self.assertRaises(ValueError, P('c:a/b').with_suffix, '/') + self.assertRaises(ValueError, P('c:a/b').with_suffix, '\\') + self.assertRaises(ValueError, P('c:a/b').with_suffix, 'c:') + self.assertRaises(ValueError, P('c:a/b').with_suffix, '/.gz') + self.assertRaises(ValueError, P('c:a/b').with_suffix, '\\.gz') + self.assertRaises(ValueError, P('c:a/b').with_suffix, 'c:.gz') + self.assertRaises(ValueError, P('c:a/b').with_suffix, 'c/d') + self.assertRaises(ValueError, P('c:a/b').with_suffix, 'c\\d') + self.assertRaises(ValueError, P('c:a/b').with_suffix, '.c/d') + self.assertRaises(ValueError, P('c:a/b').with_suffix, '.c\\d') + self.assertRaises(TypeError, P('c:a/b').with_suffix, None) + + def test_with_suffix_empty(self): + P = self.cls + # Path doesn't have a "filename" component. + self.assertRaises(ValueError, P('').with_suffix, '.gz') + self.assertRaises(ValueError, P('/').with_suffix, '.gz') + + def test_with_suffix_invalid(self): + P = self.cls + # Invalid suffix. + self.assertRaises(ValueError, P('a/b').with_suffix, 'gz') + self.assertRaises(ValueError, P('a/b').with_suffix, '/') + self.assertRaises(ValueError, P('a/b').with_suffix, '.') + self.assertRaises(ValueError, P('a/b').with_suffix, '/.gz') + self.assertRaises(ValueError, P('a/b').with_suffix, 'c/d') + self.assertRaises(ValueError, P('a/b').with_suffix, '.c/.d') + self.assertRaises(ValueError, P('a/b').with_suffix, './.d') + self.assertRaises(ValueError, P('a/b').with_suffix, '.d/.') + self.assertRaises(TypeError, P('a/b').with_suffix, None) + + def test_relative_to_common(self): + P = self.cls + p = P('a/b') + self.assertRaises(TypeError, p.relative_to) + self.assertRaises(TypeError, p.relative_to, b'a') + self.assertEqual(p.relative_to(P('')), P('a/b')) + self.assertEqual(p.relative_to(''), P('a/b')) + self.assertEqual(p.relative_to(P('a')), P('b')) + self.assertEqual(p.relative_to('a'), P('b')) + self.assertEqual(p.relative_to('a/'), P('b')) + self.assertEqual(p.relative_to(P('a/b')), P('')) + self.assertEqual(p.relative_to('a/b'), P('')) + self.assertEqual(p.relative_to(P(''), walk_up=True), P('a/b')) + self.assertEqual(p.relative_to('', walk_up=True), P('a/b')) + self.assertEqual(p.relative_to(P('a'), walk_up=True), P('b')) + self.assertEqual(p.relative_to('a', walk_up=True), P('b')) + self.assertEqual(p.relative_to('a/', walk_up=True), P('b')) + self.assertEqual(p.relative_to(P('a/b'), walk_up=True), P('')) + self.assertEqual(p.relative_to('a/b', walk_up=True), P('')) + self.assertEqual(p.relative_to(P('a/c'), walk_up=True), P('../b')) + self.assertEqual(p.relative_to('a/c', walk_up=True), P('../b')) + self.assertEqual(p.relative_to(P('a/b/c'), walk_up=True), P('..')) + self.assertEqual(p.relative_to('a/b/c', walk_up=True), P('..')) + self.assertEqual(p.relative_to(P('c'), walk_up=True), P('../a/b')) + self.assertEqual(p.relative_to('c', walk_up=True), P('../a/b')) + # Unrelated paths. + self.assertRaises(ValueError, p.relative_to, P('c')) + self.assertRaises(ValueError, p.relative_to, P('a/b/c')) + self.assertRaises(ValueError, p.relative_to, P('a/c')) + self.assertRaises(ValueError, p.relative_to, P('/a')) + self.assertRaises(ValueError, p.relative_to, P("../a")) + self.assertRaises(ValueError, p.relative_to, P("a/..")) + self.assertRaises(ValueError, p.relative_to, P("/a/..")) + self.assertRaises(ValueError, p.relative_to, P('/'), walk_up=True) + self.assertRaises(ValueError, p.relative_to, P('/a'), walk_up=True) + self.assertRaises(ValueError, p.relative_to, P("../a"), walk_up=True) + self.assertRaises(ValueError, p.relative_to, P("a/.."), walk_up=True) + self.assertRaises(ValueError, p.relative_to, P("/a/.."), walk_up=True) + p = P('/a/b') + self.assertEqual(p.relative_to(P('/')), P('a/b')) + self.assertEqual(p.relative_to('/'), P('a/b')) + self.assertEqual(p.relative_to(P('/a')), P('b')) + self.assertEqual(p.relative_to('/a'), P('b')) + self.assertEqual(p.relative_to('/a/'), P('b')) + self.assertEqual(p.relative_to(P('/a/b')), P('')) + self.assertEqual(p.relative_to('/a/b'), P('')) + self.assertEqual(p.relative_to(P('/'), walk_up=True), P('a/b')) + self.assertEqual(p.relative_to('/', walk_up=True), P('a/b')) + self.assertEqual(p.relative_to(P('/a'), walk_up=True), P('b')) + self.assertEqual(p.relative_to('/a', walk_up=True), P('b')) + self.assertEqual(p.relative_to('/a/', walk_up=True), P('b')) + self.assertEqual(p.relative_to(P('/a/b'), walk_up=True), P('')) + self.assertEqual(p.relative_to('/a/b', walk_up=True), P('')) + self.assertEqual(p.relative_to(P('/a/c'), walk_up=True), P('../b')) + self.assertEqual(p.relative_to('/a/c', walk_up=True), P('../b')) + self.assertEqual(p.relative_to(P('/a/b/c'), walk_up=True), P('..')) + self.assertEqual(p.relative_to('/a/b/c', walk_up=True), P('..')) + self.assertEqual(p.relative_to(P('/c'), walk_up=True), P('../a/b')) + self.assertEqual(p.relative_to('/c', walk_up=True), P('../a/b')) + # Unrelated paths. + self.assertRaises(ValueError, p.relative_to, P('/c')) + self.assertRaises(ValueError, p.relative_to, P('/a/b/c')) + self.assertRaises(ValueError, p.relative_to, P('/a/c')) + self.assertRaises(ValueError, p.relative_to, P('')) + self.assertRaises(ValueError, p.relative_to, '') + self.assertRaises(ValueError, p.relative_to, P('a')) + self.assertRaises(ValueError, p.relative_to, P("../a")) + self.assertRaises(ValueError, p.relative_to, P("a/..")) + self.assertRaises(ValueError, p.relative_to, P("/a/..")) + self.assertRaises(ValueError, p.relative_to, P(''), walk_up=True) + self.assertRaises(ValueError, p.relative_to, P('a'), walk_up=True) + self.assertRaises(ValueError, p.relative_to, P("../a"), walk_up=True) + self.assertRaises(ValueError, p.relative_to, P("a/.."), walk_up=True) + self.assertRaises(ValueError, p.relative_to, P("/a/.."), walk_up=True) + + @needs_windows + def test_relative_to_windows(self): + P = self.cls + p = P('C:Foo/Bar') + self.assertEqual(p.relative_to(P('c:')), P('Foo/Bar')) + self.assertEqual(p.relative_to('c:'), P('Foo/Bar')) + self.assertEqual(p.relative_to(P('c:foO')), P('Bar')) + self.assertEqual(p.relative_to('c:foO'), P('Bar')) + self.assertEqual(p.relative_to('c:foO/'), P('Bar')) + self.assertEqual(p.relative_to(P('c:foO/baR')), P()) + self.assertEqual(p.relative_to('c:foO/baR'), P()) + self.assertEqual(p.relative_to(P('c:'), walk_up=True), P('Foo/Bar')) + self.assertEqual(p.relative_to('c:', walk_up=True), P('Foo/Bar')) + self.assertEqual(p.relative_to(P('c:foO'), walk_up=True), P('Bar')) + self.assertEqual(p.relative_to('c:foO', walk_up=True), P('Bar')) + self.assertEqual(p.relative_to('c:foO/', walk_up=True), P('Bar')) + self.assertEqual(p.relative_to(P('c:foO/baR'), walk_up=True), P()) + self.assertEqual(p.relative_to('c:foO/baR', walk_up=True), P()) + self.assertEqual(p.relative_to(P('C:Foo/Bar/Baz'), walk_up=True), P('..')) + self.assertEqual(p.relative_to(P('C:Foo/Baz'), walk_up=True), P('../Bar')) + self.assertEqual(p.relative_to(P('C:Baz/Bar'), walk_up=True), P('../../Foo/Bar')) + # Unrelated paths. + self.assertRaises(ValueError, p.relative_to, P()) + self.assertRaises(ValueError, p.relative_to, '') + self.assertRaises(ValueError, p.relative_to, P('d:')) + self.assertRaises(ValueError, p.relative_to, P('/')) + self.assertRaises(ValueError, p.relative_to, P('Foo')) + self.assertRaises(ValueError, p.relative_to, P('/Foo')) + self.assertRaises(ValueError, p.relative_to, P('C:/Foo')) + self.assertRaises(ValueError, p.relative_to, P('C:Foo/Bar/Baz')) + self.assertRaises(ValueError, p.relative_to, P('C:Foo/Baz')) + self.assertRaises(ValueError, p.relative_to, P(), walk_up=True) + self.assertRaises(ValueError, p.relative_to, '', walk_up=True) + self.assertRaises(ValueError, p.relative_to, P('d:'), walk_up=True) + self.assertRaises(ValueError, p.relative_to, P('/'), walk_up=True) + self.assertRaises(ValueError, p.relative_to, P('Foo'), walk_up=True) + self.assertRaises(ValueError, p.relative_to, P('/Foo'), walk_up=True) + self.assertRaises(ValueError, p.relative_to, P('C:/Foo'), walk_up=True) + p = P('C:/Foo/Bar') + self.assertEqual(p.relative_to(P('c:/')), P('Foo/Bar')) + self.assertEqual(p.relative_to('c:/'), P('Foo/Bar')) + self.assertEqual(p.relative_to(P('c:/foO')), P('Bar')) + self.assertEqual(p.relative_to('c:/foO'), P('Bar')) + self.assertEqual(p.relative_to('c:/foO/'), P('Bar')) + self.assertEqual(p.relative_to(P('c:/foO/baR')), P()) + self.assertEqual(p.relative_to('c:/foO/baR'), P()) + self.assertEqual(p.relative_to(P('c:/'), walk_up=True), P('Foo/Bar')) + self.assertEqual(p.relative_to('c:/', walk_up=True), P('Foo/Bar')) + self.assertEqual(p.relative_to(P('c:/foO'), walk_up=True), P('Bar')) + self.assertEqual(p.relative_to('c:/foO', walk_up=True), P('Bar')) + self.assertEqual(p.relative_to('c:/foO/', walk_up=True), P('Bar')) + self.assertEqual(p.relative_to(P('c:/foO/baR'), walk_up=True), P()) + self.assertEqual(p.relative_to('c:/foO/baR', walk_up=True), P()) + self.assertEqual(p.relative_to('C:/Baz', walk_up=True), P('../Foo/Bar')) + self.assertEqual(p.relative_to('C:/Foo/Bar/Baz', walk_up=True), P('..')) + self.assertEqual(p.relative_to('C:/Foo/Baz', walk_up=True), P('../Bar')) + # Unrelated paths. + self.assertRaises(ValueError, p.relative_to, 'c:') + self.assertRaises(ValueError, p.relative_to, P('c:')) + self.assertRaises(ValueError, p.relative_to, P('C:/Baz')) + self.assertRaises(ValueError, p.relative_to, P('C:/Foo/Bar/Baz')) + self.assertRaises(ValueError, p.relative_to, P('C:/Foo/Baz')) + self.assertRaises(ValueError, p.relative_to, P('C:Foo')) + self.assertRaises(ValueError, p.relative_to, P('d:')) + self.assertRaises(ValueError, p.relative_to, P('d:/')) + self.assertRaises(ValueError, p.relative_to, P('/')) + self.assertRaises(ValueError, p.relative_to, P('/Foo')) + self.assertRaises(ValueError, p.relative_to, P('//C/Foo')) + self.assertRaises(ValueError, p.relative_to, 'c:', walk_up=True) + self.assertRaises(ValueError, p.relative_to, P('c:'), walk_up=True) + self.assertRaises(ValueError, p.relative_to, P('C:Foo'), walk_up=True) + self.assertRaises(ValueError, p.relative_to, P('d:'), walk_up=True) + self.assertRaises(ValueError, p.relative_to, P('d:/'), walk_up=True) + self.assertRaises(ValueError, p.relative_to, P('/'), walk_up=True) + self.assertRaises(ValueError, p.relative_to, P('/Foo'), walk_up=True) + self.assertRaises(ValueError, p.relative_to, P('//C/Foo'), walk_up=True) + # UNC paths. + p = P('//Server/Share/Foo/Bar') + self.assertEqual(p.relative_to(P('//sErver/sHare')), P('Foo/Bar')) + self.assertEqual(p.relative_to('//sErver/sHare'), P('Foo/Bar')) + self.assertEqual(p.relative_to('//sErver/sHare/'), P('Foo/Bar')) + self.assertEqual(p.relative_to(P('//sErver/sHare/Foo')), P('Bar')) + self.assertEqual(p.relative_to('//sErver/sHare/Foo'), P('Bar')) + self.assertEqual(p.relative_to('//sErver/sHare/Foo/'), P('Bar')) + self.assertEqual(p.relative_to(P('//sErver/sHare/Foo/Bar')), P()) + self.assertEqual(p.relative_to('//sErver/sHare/Foo/Bar'), P()) + self.assertEqual(p.relative_to(P('//sErver/sHare'), walk_up=True), P('Foo/Bar')) + self.assertEqual(p.relative_to('//sErver/sHare', walk_up=True), P('Foo/Bar')) + self.assertEqual(p.relative_to('//sErver/sHare/', walk_up=True), P('Foo/Bar')) + self.assertEqual(p.relative_to(P('//sErver/sHare/Foo'), walk_up=True), P('Bar')) + self.assertEqual(p.relative_to('//sErver/sHare/Foo', walk_up=True), P('Bar')) + self.assertEqual(p.relative_to('//sErver/sHare/Foo/', walk_up=True), P('Bar')) + self.assertEqual(p.relative_to(P('//sErver/sHare/Foo/Bar'), walk_up=True), P()) + self.assertEqual(p.relative_to('//sErver/sHare/Foo/Bar', walk_up=True), P()) + self.assertEqual(p.relative_to(P('//sErver/sHare/bar'), walk_up=True), P('../Foo/Bar')) + self.assertEqual(p.relative_to('//sErver/sHare/bar', walk_up=True), P('../Foo/Bar')) + # Unrelated paths. + self.assertRaises(ValueError, p.relative_to, P('/Server/Share/Foo')) + self.assertRaises(ValueError, p.relative_to, P('c:/Server/Share/Foo')) + self.assertRaises(ValueError, p.relative_to, P('//z/Share/Foo')) + self.assertRaises(ValueError, p.relative_to, P('//Server/z/Foo')) + self.assertRaises(ValueError, p.relative_to, P('/Server/Share/Foo'), walk_up=True) + self.assertRaises(ValueError, p.relative_to, P('c:/Server/Share/Foo'), walk_up=True) + self.assertRaises(ValueError, p.relative_to, P('//z/Share/Foo'), walk_up=True) + self.assertRaises(ValueError, p.relative_to, P('//Server/z/Foo'), walk_up=True) + + def test_is_relative_to_common(self): + P = self.cls + p = P('a/b') + self.assertRaises(TypeError, p.is_relative_to) + self.assertRaises(TypeError, p.is_relative_to, b'a') + self.assertTrue(p.is_relative_to(P(''))) + self.assertTrue(p.is_relative_to('')) + self.assertTrue(p.is_relative_to(P('a'))) + self.assertTrue(p.is_relative_to('a/')) + self.assertTrue(p.is_relative_to(P('a/b'))) + self.assertTrue(p.is_relative_to('a/b')) + # Unrelated paths. + self.assertFalse(p.is_relative_to(P('c'))) + self.assertFalse(p.is_relative_to(P('a/b/c'))) + self.assertFalse(p.is_relative_to(P('a/c'))) + self.assertFalse(p.is_relative_to(P('/a'))) + p = P('/a/b') + self.assertTrue(p.is_relative_to(P('/'))) + self.assertTrue(p.is_relative_to('/')) + self.assertTrue(p.is_relative_to(P('/a'))) + self.assertTrue(p.is_relative_to('/a')) + self.assertTrue(p.is_relative_to('/a/')) + self.assertTrue(p.is_relative_to(P('/a/b'))) + self.assertTrue(p.is_relative_to('/a/b')) + # Unrelated paths. + self.assertFalse(p.is_relative_to(P('/c'))) + self.assertFalse(p.is_relative_to(P('/a/b/c'))) + self.assertFalse(p.is_relative_to(P('/a/c'))) + self.assertFalse(p.is_relative_to(P(''))) + self.assertFalse(p.is_relative_to('')) + self.assertFalse(p.is_relative_to(P('a'))) + + @needs_windows + def test_is_relative_to_windows(self): + P = self.cls + p = P('C:Foo/Bar') + self.assertTrue(p.is_relative_to(P('c:'))) + self.assertTrue(p.is_relative_to('c:')) + self.assertTrue(p.is_relative_to(P('c:foO'))) + self.assertTrue(p.is_relative_to('c:foO')) + self.assertTrue(p.is_relative_to('c:foO/')) + self.assertTrue(p.is_relative_to(P('c:foO/baR'))) + self.assertTrue(p.is_relative_to('c:foO/baR')) + # Unrelated paths. + self.assertFalse(p.is_relative_to(P())) + self.assertFalse(p.is_relative_to('')) + self.assertFalse(p.is_relative_to(P('d:'))) + self.assertFalse(p.is_relative_to(P('/'))) + self.assertFalse(p.is_relative_to(P('Foo'))) + self.assertFalse(p.is_relative_to(P('/Foo'))) + self.assertFalse(p.is_relative_to(P('C:/Foo'))) + self.assertFalse(p.is_relative_to(P('C:Foo/Bar/Baz'))) + self.assertFalse(p.is_relative_to(P('C:Foo/Baz'))) + p = P('C:/Foo/Bar') + self.assertTrue(p.is_relative_to(P('c:/'))) + self.assertTrue(p.is_relative_to(P('c:/foO'))) + self.assertTrue(p.is_relative_to('c:/foO/')) + self.assertTrue(p.is_relative_to(P('c:/foO/baR'))) + self.assertTrue(p.is_relative_to('c:/foO/baR')) + # Unrelated paths. + self.assertFalse(p.is_relative_to('c:')) + self.assertFalse(p.is_relative_to(P('C:/Baz'))) + self.assertFalse(p.is_relative_to(P('C:/Foo/Bar/Baz'))) + self.assertFalse(p.is_relative_to(P('C:/Foo/Baz'))) + self.assertFalse(p.is_relative_to(P('C:Foo'))) + self.assertFalse(p.is_relative_to(P('d:'))) + self.assertFalse(p.is_relative_to(P('d:/'))) + self.assertFalse(p.is_relative_to(P('/'))) + self.assertFalse(p.is_relative_to(P('/Foo'))) + self.assertFalse(p.is_relative_to(P('//C/Foo'))) + # UNC paths. + p = P('//Server/Share/Foo/Bar') + self.assertTrue(p.is_relative_to(P('//sErver/sHare'))) + self.assertTrue(p.is_relative_to('//sErver/sHare')) + self.assertTrue(p.is_relative_to('//sErver/sHare/')) + self.assertTrue(p.is_relative_to(P('//sErver/sHare/Foo'))) + self.assertTrue(p.is_relative_to('//sErver/sHare/Foo')) + self.assertTrue(p.is_relative_to('//sErver/sHare/Foo/')) + self.assertTrue(p.is_relative_to(P('//sErver/sHare/Foo/Bar'))) + self.assertTrue(p.is_relative_to('//sErver/sHare/Foo/Bar')) + # Unrelated paths. + self.assertFalse(p.is_relative_to(P('/Server/Share/Foo'))) + self.assertFalse(p.is_relative_to(P('c:/Server/Share/Foo'))) + self.assertFalse(p.is_relative_to(P('//z/Share/Foo'))) + self.assertFalse(p.is_relative_to(P('//Server/z/Foo'))) + + @needs_posix + def test_is_absolute_posix(self): + P = self.cls + self.assertFalse(P('').is_absolute()) + self.assertFalse(P('a').is_absolute()) + self.assertFalse(P('a/b/').is_absolute()) + self.assertTrue(P('/').is_absolute()) + self.assertTrue(P('/a').is_absolute()) + self.assertTrue(P('/a/b/').is_absolute()) + self.assertTrue(P('//a').is_absolute()) + self.assertTrue(P('//a/b').is_absolute()) + + @needs_windows + def test_is_absolute_windows(self): + P = self.cls + # Under NT, only paths with both a drive and a root are absolute. + self.assertFalse(P().is_absolute()) + self.assertFalse(P('a').is_absolute()) + self.assertFalse(P('a/b/').is_absolute()) + self.assertFalse(P('/').is_absolute()) + self.assertFalse(P('/a').is_absolute()) + self.assertFalse(P('/a/b/').is_absolute()) + self.assertFalse(P('c:').is_absolute()) + self.assertFalse(P('c:a').is_absolute()) + self.assertFalse(P('c:a/b/').is_absolute()) + self.assertTrue(P('c:/').is_absolute()) + self.assertTrue(P('c:/a').is_absolute()) + self.assertTrue(P('c:/a/b/').is_absolute()) + # UNC paths are absolute by definition. + self.assertTrue(P('//').is_absolute()) + self.assertTrue(P('//a').is_absolute()) + self.assertTrue(P('//a/b').is_absolute()) + self.assertTrue(P('//a/b/').is_absolute()) + self.assertTrue(P('//a/b/c').is_absolute()) + self.assertTrue(P('//a/b/c/d').is_absolute()) + self.assertTrue(P('//?/UNC/').is_absolute()) + self.assertTrue(P('//?/UNC/spam').is_absolute()) + + +# +# Tests for the virtual classes. +# + +class PathBaseTest(PurePathBaseTest): + cls = PathBase + + def test_unsupported_operation(self): + P = self.cls + p = self.cls('') + e = UnsupportedOperation + self.assertRaises(e, p.stat) + self.assertRaises(e, p.lstat) + self.assertRaises(e, p.exists) + self.assertRaises(e, p.samefile, 'foo') + self.assertRaises(e, p.is_dir) + self.assertRaises(e, p.is_file) + self.assertRaises(e, p.is_mount) + self.assertRaises(e, p.is_symlink) + self.assertRaises(e, p.is_block_device) + self.assertRaises(e, p.is_char_device) + self.assertRaises(e, p.is_fifo) + self.assertRaises(e, p.is_socket) + self.assertRaises(e, p.open) + self.assertRaises(e, p.read_bytes) + self.assertRaises(e, p.read_text) + self.assertRaises(e, p.write_bytes, b'foo') + self.assertRaises(e, p.write_text, 'foo') + self.assertRaises(e, p.iterdir) + self.assertRaises(e, p.glob, '*') + self.assertRaises(e, p.rglob, '*') + self.assertRaises(e, lambda: list(p.walk())) + self.assertRaises(e, p.absolute) + self.assertRaises(e, P.cwd) + self.assertRaises(e, p.expanduser) + self.assertRaises(e, p.home) + self.assertRaises(e, p.readlink) + self.assertRaises(e, p.symlink_to, 'foo') + self.assertRaises(e, p.hardlink_to, 'foo') + self.assertRaises(e, p.mkdir) + self.assertRaises(e, p.touch) + self.assertRaises(e, p.rename, 'foo') + self.assertRaises(e, p.replace, 'foo') + self.assertRaises(e, p.chmod, 0o755) + self.assertRaises(e, p.lchmod, 0o755) + self.assertRaises(e, p.unlink) + self.assertRaises(e, p.rmdir) + self.assertRaises(e, p.owner) + self.assertRaises(e, p.group) + self.assertRaises(e, p.as_uri) + + def test_as_uri_common(self): + e = UnsupportedOperation + self.assertRaises(e, self.cls('').as_uri) + + def test_fspath_common(self): + self.assertRaises(TypeError, os.fspath, self.cls('')) + + def test_as_bytes_common(self): + self.assertRaises(TypeError, bytes, self.cls('')) + + +class DummyPathIO(io.BytesIO): + """ + Used by DummyPath to implement `open('w')` + """ + + def __init__(self, files, path): + super().__init__() + self.files = files + self.path = path + + def close(self): + self.files[self.path] = self.getvalue() + super().close() + + +DummyPathStatResult = collections.namedtuple( + 'DummyPathStatResult', + 'st_mode st_ino st_dev st_nlink st_uid st_gid st_size st_atime st_mtime st_ctime') + + +class DummyPath(PathBase): + """ + Simple implementation of PathBase that keeps files and directories in + memory. + """ + __slots__ = () + parser = posixpath + + _files = {} + _directories = {} + _symlinks = {} + + def __eq__(self, other): + if not isinstance(other, DummyPath): + return NotImplemented + return str(self) == str(other) + + def __hash__(self): + return hash(str(self)) + + def __repr__(self): + return "{}({!r})".format(self.__class__.__name__, self.as_posix()) + + def stat(self, *, follow_symlinks=True): + if follow_symlinks or self.name in ('', '.', '..'): + path = str(self.resolve(strict=True)) + else: + path = str(self.parent.resolve(strict=True) / self.name) + if path in self._files: + st_mode = stat.S_IFREG + elif path in self._directories: + st_mode = stat.S_IFDIR + elif path in self._symlinks: + st_mode = stat.S_IFLNK + else: + raise FileNotFoundError(errno.ENOENT, "Not found", str(self)) + return DummyPathStatResult(st_mode, hash(str(self)), 0, 0, 0, 0, 0, 0, 0, 0) + + def open(self, mode='r', buffering=-1, encoding=None, + errors=None, newline=None): + if buffering != -1: + raise NotImplementedError + path_obj = self.resolve() + path = str(path_obj) + name = path_obj.name + parent = str(path_obj.parent) + if path in self._directories: + raise IsADirectoryError(errno.EISDIR, "Is a directory", path) + + text = 'b' not in mode + mode = ''.join(c for c in mode if c not in 'btU') + if mode == 'r': + if path not in self._files: + raise FileNotFoundError(errno.ENOENT, "File not found", path) + stream = io.BytesIO(self._files[path]) + elif mode == 'w': + if parent not in self._directories: + raise FileNotFoundError(errno.ENOENT, "File not found", parent) + stream = DummyPathIO(self._files, path) + self._files[path] = b'' + self._directories[parent].add(name) + else: + raise NotImplementedError + if text: + stream = io.TextIOWrapper(stream, encoding=encoding, errors=errors, newline=newline) + return stream + + def iterdir(self): + path = str(self.resolve()) + if path in self._files: + raise NotADirectoryError(errno.ENOTDIR, "Not a directory", path) + elif path in self._directories: + return (self / name for name in self._directories[path]) + else: + raise FileNotFoundError(errno.ENOENT, "File not found", path) + + def mkdir(self, mode=0o777, parents=False, exist_ok=False): + path = str(self.resolve()) + if path in self._directories: + if exist_ok: + return + else: + raise FileExistsError(errno.EEXIST, "File exists", path) + try: + if self.name: + self._directories[str(self.parent)].add(self.name) + self._directories[path] = set() + except KeyError: + if not parents: + raise FileNotFoundError(errno.ENOENT, "File not found", str(self.parent)) from None + self.parent.mkdir(parents=True, exist_ok=True) + self.mkdir(mode, parents=False, exist_ok=exist_ok) + + +class DummyPathTest(DummyPurePathTest): + """Tests for PathBase methods that use stat(), open() and iterdir().""" + + cls = DummyPath + can_symlink = False + + # (self.base) + # | + # |-- brokenLink -> non-existing + # |-- dirA + # | `-- linkC -> ../dirB + # |-- dirB + # | |-- fileB + # | `-- linkD -> ../dirB + # |-- dirC + # | |-- dirD + # | | `-- fileD + # | `-- fileC + # | `-- novel.txt + # |-- dirE # No permissions + # |-- fileA + # |-- linkA -> fileA + # |-- linkB -> dirB + # `-- brokenLinkLoop -> brokenLinkLoop + # + + def setUp(self): + super().setUp() + name = self.id().split('.')[-1] + if name in _tests_needing_symlinks and not self.can_symlink: + self.skipTest('requires symlinks') + parser = self.cls.parser + p = self.cls(self.base) + p.mkdir(parents=True) + p.joinpath('dirA').mkdir() + p.joinpath('dirB').mkdir() + p.joinpath('dirC').mkdir() + p.joinpath('dirC', 'dirD').mkdir() + p.joinpath('dirE').mkdir() + with p.joinpath('fileA').open('wb') as f: + f.write(b"this is file A\n") + with p.joinpath('dirB', 'fileB').open('wb') as f: + f.write(b"this is file B\n") + with p.joinpath('dirC', 'fileC').open('wb') as f: + f.write(b"this is file C\n") + with p.joinpath('dirC', 'novel.txt').open('wb') as f: + f.write(b"this is a novel\n") + with p.joinpath('dirC', 'dirD', 'fileD').open('wb') as f: + f.write(b"this is file D\n") + if self.can_symlink: + p.joinpath('linkA').symlink_to('fileA') + p.joinpath('brokenLink').symlink_to('non-existing') + p.joinpath('linkB').symlink_to('dirB') + p.joinpath('dirA', 'linkC').symlink_to(parser.join('..', 'dirB')) + p.joinpath('dirB', 'linkD').symlink_to(parser.join('..', 'dirB')) + p.joinpath('brokenLinkLoop').symlink_to('brokenLinkLoop') + + def tearDown(self): + cls = self.cls + cls._files.clear() + cls._directories.clear() + cls._symlinks.clear() + + def tempdir(self): + path = self.cls(self.base).with_name('tmp-dirD') + path.mkdir() + return path + + def assertFileNotFound(self, func, *args, **kwargs): + with self.assertRaises(FileNotFoundError) as cm: + func(*args, **kwargs) + self.assertEqual(cm.exception.errno, errno.ENOENT) + + def assertEqualNormCase(self, path_a, path_b): + normcase = self.parser.normcase + self.assertEqual(normcase(path_a), normcase(path_b)) + + def test_samefile(self): + parser = self.parser + fileA_path = parser.join(self.base, 'fileA') + fileB_path = parser.join(self.base, 'dirB', 'fileB') + p = self.cls(fileA_path) + pp = self.cls(fileA_path) + q = self.cls(fileB_path) + self.assertTrue(p.samefile(fileA_path)) + self.assertTrue(p.samefile(pp)) + self.assertFalse(p.samefile(fileB_path)) + self.assertFalse(p.samefile(q)) + # Test the non-existent file case + non_existent = parser.join(self.base, 'foo') + r = self.cls(non_existent) + self.assertRaises(FileNotFoundError, p.samefile, r) + self.assertRaises(FileNotFoundError, p.samefile, non_existent) + self.assertRaises(FileNotFoundError, r.samefile, p) + self.assertRaises(FileNotFoundError, r.samefile, non_existent) + self.assertRaises(FileNotFoundError, r.samefile, r) + self.assertRaises(FileNotFoundError, r.samefile, non_existent) + + def test_exists(self): + P = self.cls + p = P(self.base) + self.assertIs(True, p.exists()) + self.assertIs(True, (p / 'dirA').exists()) + self.assertIs(True, (p / 'fileA').exists()) + self.assertIs(False, (p / 'fileA' / 'bah').exists()) + if self.can_symlink: + self.assertIs(True, (p / 'linkA').exists()) + self.assertIs(True, (p / 'linkB').exists()) + self.assertIs(True, (p / 'linkB' / 'fileB').exists()) + self.assertIs(False, (p / 'linkA' / 'bah').exists()) + self.assertIs(False, (p / 'brokenLink').exists()) + self.assertIs(True, (p / 'brokenLink').exists(follow_symlinks=False)) + self.assertIs(False, (p / 'foo').exists()) + self.assertIs(False, P('/xyzzy').exists()) + self.assertIs(False, P(self.base + '\udfff').exists()) + self.assertIs(False, P(self.base + '\x00').exists()) + + def test_open_common(self): + p = self.cls(self.base) + with (p / 'fileA').open('r') as f: + self.assertIsInstance(f, io.TextIOBase) + self.assertEqual(f.read(), "this is file A\n") + with (p / 'fileA').open('rb') as f: + self.assertIsInstance(f, io.BufferedIOBase) + self.assertEqual(f.read().strip(), b"this is file A") + + def test_read_write_bytes(self): + p = self.cls(self.base) + (p / 'fileA').write_bytes(b'abcdefg') + self.assertEqual((p / 'fileA').read_bytes(), b'abcdefg') + # Check that trying to write str does not truncate the file. + self.assertRaises(TypeError, (p / 'fileA').write_bytes, 'somestr') + self.assertEqual((p / 'fileA').read_bytes(), b'abcdefg') + + def test_read_write_text(self): + p = self.cls(self.base) + (p / 'fileA').write_text('äbcdefg', encoding='latin-1') + self.assertEqual((p / 'fileA').read_text( + encoding='utf-8', errors='ignore'), 'bcdefg') + # Check that trying to write bytes does not truncate the file. + self.assertRaises(TypeError, (p / 'fileA').write_text, b'somebytes') + self.assertEqual((p / 'fileA').read_text(encoding='latin-1'), 'äbcdefg') + + def test_read_text_with_newlines(self): + p = self.cls(self.base) + # Check that `\n` character change nothing + (p / 'fileA').write_bytes(b'abcde\r\nfghlk\n\rmnopq') + self.assertEqual((p / 'fileA').read_text(newline='\n'), + 'abcde\r\nfghlk\n\rmnopq') + # Check that `\r` character replaces `\n` + (p / 'fileA').write_bytes(b'abcde\r\nfghlk\n\rmnopq') + self.assertEqual((p / 'fileA').read_text(newline='\r'), + 'abcde\r\nfghlk\n\rmnopq') + # Check that `\r\n` character replaces `\n` + (p / 'fileA').write_bytes(b'abcde\r\nfghlk\n\rmnopq') + self.assertEqual((p / 'fileA').read_text(newline='\r\n'), + 'abcde\r\nfghlk\n\rmnopq') + + def test_write_text_with_newlines(self): + p = self.cls(self.base) + # Check that `\n` character change nothing + (p / 'fileA').write_text('abcde\r\nfghlk\n\rmnopq', newline='\n') + self.assertEqual((p / 'fileA').read_bytes(), + b'abcde\r\nfghlk\n\rmnopq') + # Check that `\r` character replaces `\n` + (p / 'fileA').write_text('abcde\r\nfghlk\n\rmnopq', newline='\r') + self.assertEqual((p / 'fileA').read_bytes(), + b'abcde\r\rfghlk\r\rmnopq') + # Check that `\r\n` character replaces `\n` + (p / 'fileA').write_text('abcde\r\nfghlk\n\rmnopq', newline='\r\n') + self.assertEqual((p / 'fileA').read_bytes(), + b'abcde\r\r\nfghlk\r\n\rmnopq') + # Check that no argument passed will change `\n` to `os.linesep` + os_linesep_byte = bytes(os.linesep, encoding='ascii') + (p / 'fileA').write_text('abcde\nfghlk\n\rmnopq') + self.assertEqual((p / 'fileA').read_bytes(), + b'abcde' + os_linesep_byte + b'fghlk' + os_linesep_byte + b'\rmnopq') + + def test_iterdir(self): + P = self.cls + p = P(self.base) + it = p.iterdir() + paths = set(it) + expected = ['dirA', 'dirB', 'dirC', 'dirE', 'fileA'] + if self.can_symlink: + expected += ['linkA', 'linkB', 'brokenLink', 'brokenLinkLoop'] + self.assertEqual(paths, { P(self.base, q) for q in expected }) + + @needs_symlinks + def test_iterdir_symlink(self): + # __iter__ on a symlink to a directory. + P = self.cls + p = P(self.base, 'linkB') + paths = set(p.iterdir()) + expected = { P(self.base, 'linkB', q) for q in ['fileB', 'linkD'] } + self.assertEqual(paths, expected) + + def test_iterdir_nodir(self): + # __iter__ on something that is not a directory. + p = self.cls(self.base, 'fileA') + with self.assertRaises(OSError) as cm: + p.iterdir() + # ENOENT or EINVAL under Windows, ENOTDIR otherwise + # (see issue #12802). + self.assertIn(cm.exception.errno, (errno.ENOTDIR, + errno.ENOENT, errno.EINVAL)) + + def test_glob_common(self): + def _check(glob, expected): + self.assertEqual(set(glob), { P(self.base, q) for q in expected }) + P = self.cls + p = P(self.base) + it = p.glob("fileA") + self.assertIsInstance(it, collections.abc.Iterator) + _check(it, ["fileA"]) + _check(p.glob("fileB"), []) + _check(p.glob("dir*/file*"), ["dirB/fileB", "dirC/fileC"]) + if not self.can_symlink: + _check(p.glob("*A"), ['dirA', 'fileA']) + else: + _check(p.glob("*A"), ['dirA', 'fileA', 'linkA']) + if not self.can_symlink: + _check(p.glob("*B/*"), ['dirB/fileB']) + else: + _check(p.glob("*B/*"), ['dirB/fileB', 'dirB/linkD', + 'linkB/fileB', 'linkB/linkD']) + if not self.can_symlink: + _check(p.glob("*/fileB"), ['dirB/fileB']) + else: + _check(p.glob("*/fileB"), ['dirB/fileB', 'linkB/fileB']) + if self.can_symlink: + _check(p.glob("brokenLink"), ['brokenLink']) + + if not self.can_symlink: + _check(p.glob("*/"), ["dirA/", "dirB/", "dirC/", "dirE/"]) + else: + _check(p.glob("*/"), ["dirA/", "dirB/", "dirC/", "dirE/", "linkB/"]) + + @needs_posix + def test_glob_posix(self): + P = self.cls + p = P(self.base) + q = p / "FILEa" + given = set(p.glob("FILEa")) + expect = {q} if q.exists() else set() + self.assertEqual(given, expect) + self.assertEqual(set(p.glob("FILEa*")), set()) + + @needs_windows + def test_glob_windows(self): + P = self.cls + p = P(self.base) + self.assertEqual(set(p.glob("FILEa")), { P(self.base, "fileA") }) + self.assertEqual(set(p.glob("*a\\")), { P(self.base, "dirA/") }) + self.assertEqual(set(p.glob("F*a")), { P(self.base, "fileA") }) + + def test_glob_empty_pattern(self): + P = self.cls + p = P(self.base) + self.assertEqual(list(p.glob("")), [p]) + self.assertEqual(list(p.glob(".")), [p / "."]) + self.assertEqual(list(p.glob("./")), [p / "./"]) + + def test_glob_case_sensitive(self): + P = self.cls + def _check(path, pattern, case_sensitive, expected): + actual = {str(q) for q in path.glob(pattern, case_sensitive=case_sensitive)} + expected = {str(P(self.base, q)) for q in expected} + self.assertEqual(actual, expected) + path = P(self.base) + _check(path, "DIRB/FILE*", True, []) + _check(path, "DIRB/FILE*", False, ["dirB/fileB"]) + _check(path, "dirb/file*", True, []) + _check(path, "dirb/file*", False, ["dirB/fileB"]) + + @needs_symlinks + def test_glob_recurse_symlinks_common(self): + def _check(path, glob, expected): + actual = {path for path in path.glob(glob, recurse_symlinks=True) + if path.parts.count("linkD") <= 1} # exclude symlink loop. + self.assertEqual(actual, { P(self.base, q) for q in expected }) + P = self.cls + p = P(self.base) + _check(p, "fileB", []) + _check(p, "dir*/file*", ["dirB/fileB", "dirC/fileC"]) + _check(p, "*A", ["dirA", "fileA", "linkA"]) + _check(p, "*B/*", ["dirB/fileB", "dirB/linkD", "linkB/fileB", "linkB/linkD"]) + _check(p, "*/fileB", ["dirB/fileB", "linkB/fileB"]) + _check(p, "*/", ["dirA/", "dirB/", "dirC/", "dirE/", "linkB/"]) + _check(p, "dir*/*/..", ["dirC/dirD/..", "dirA/linkC/..", "dirB/linkD/.."]) + _check(p, "dir*/**", [ + "dirA/", "dirA/linkC", "dirA/linkC/fileB", "dirA/linkC/linkD", "dirA/linkC/linkD/fileB", + "dirB/", "dirB/fileB", "dirB/linkD", "dirB/linkD/fileB", + "dirC/", "dirC/fileC", "dirC/dirD", "dirC/dirD/fileD", "dirC/novel.txt", + "dirE/"]) + _check(p, "dir*/**/", ["dirA/", "dirA/linkC/", "dirA/linkC/linkD/", "dirB/", "dirB/linkD/", + "dirC/", "dirC/dirD/", "dirE/"]) + _check(p, "dir*/**/..", ["dirA/..", "dirA/linkC/..", "dirB/..", + "dirB/linkD/..", "dirA/linkC/linkD/..", + "dirC/..", "dirC/dirD/..", "dirE/.."]) + _check(p, "dir*/*/**", [ + "dirA/linkC/", "dirA/linkC/linkD", "dirA/linkC/fileB", "dirA/linkC/linkD/fileB", + "dirB/linkD/", "dirB/linkD/fileB", + "dirC/dirD/", "dirC/dirD/fileD"]) + _check(p, "dir*/*/**/", ["dirA/linkC/", "dirA/linkC/linkD/", "dirB/linkD/", "dirC/dirD/"]) + _check(p, "dir*/*/**/..", ["dirA/linkC/..", "dirA/linkC/linkD/..", + "dirB/linkD/..", "dirC/dirD/.."]) + _check(p, "dir*/**/fileC", ["dirC/fileC"]) + _check(p, "dir*/*/../dirD/**/", ["dirC/dirD/../dirD/"]) + _check(p, "*/dirD/**", ["dirC/dirD/", "dirC/dirD/fileD"]) + _check(p, "*/dirD/**/", ["dirC/dirD/"]) + + def test_rglob_recurse_symlinks_false(self): + def _check(path, glob, expected): + actual = set(path.rglob(glob, recurse_symlinks=False)) + self.assertEqual(actual, { P(self.base, q) for q in expected }) + P = self.cls + p = P(self.base) + it = p.rglob("fileA") + self.assertIsInstance(it, collections.abc.Iterator) + _check(p, "fileA", ["fileA"]) + _check(p, "fileB", ["dirB/fileB"]) + _check(p, "**/fileB", ["dirB/fileB"]) + _check(p, "*/fileA", []) + + if self.can_symlink: + _check(p, "*/fileB", ["dirB/fileB", "dirB/linkD/fileB", + "linkB/fileB", "dirA/linkC/fileB"]) + _check(p, "*/", [ + "dirA/", "dirA/linkC/", "dirB/", "dirB/linkD/", "dirC/", + "dirC/dirD/", "dirE/", "linkB/"]) + else: + _check(p, "*/fileB", ["dirB/fileB"]) + _check(p, "*/", ["dirA/", "dirB/", "dirC/", "dirC/dirD/", "dirE/"]) + + _check(p, "file*", ["fileA", "dirB/fileB", "dirC/fileC", "dirC/dirD/fileD"]) + _check(p, "", ["", "dirA/", "dirB/", "dirC/", "dirE/", "dirC/dirD/"]) + p = P(self.base, "dirC") + _check(p, "*", ["dirC/fileC", "dirC/novel.txt", + "dirC/dirD", "dirC/dirD/fileD"]) + _check(p, "file*", ["dirC/fileC", "dirC/dirD/fileD"]) + _check(p, "**/file*", ["dirC/fileC", "dirC/dirD/fileD"]) + _check(p, "dir*/**", ["dirC/dirD/", "dirC/dirD/fileD"]) + _check(p, "dir*/**/", ["dirC/dirD/"]) + _check(p, "*/*", ["dirC/dirD/fileD"]) + _check(p, "*/", ["dirC/dirD/"]) + _check(p, "", ["dirC/", "dirC/dirD/"]) + _check(p, "**", ["dirC/", "dirC/fileC", "dirC/dirD", "dirC/dirD/fileD", "dirC/novel.txt"]) + _check(p, "**/", ["dirC/", "dirC/dirD/"]) + # gh-91616, a re module regression + _check(p, "*.txt", ["dirC/novel.txt"]) + _check(p, "*.*", ["dirC/novel.txt"]) + + @needs_posix + def test_rglob_posix(self): + P = self.cls + p = P(self.base, "dirC") + q = p / "dirD" / "FILEd" + given = set(p.rglob("FILEd")) + expect = {q} if q.exists() else set() + self.assertEqual(given, expect) + self.assertEqual(set(p.rglob("FILEd*")), set()) + + @needs_windows + def test_rglob_windows(self): + P = self.cls + p = P(self.base, "dirC") + self.assertEqual(set(p.rglob("FILEd")), { P(self.base, "dirC/dirD/fileD") }) + self.assertEqual(set(p.rglob("*\\")), { P(self.base, "dirC/dirD/") }) + + @needs_symlinks + def test_rglob_recurse_symlinks_common(self): + def _check(path, glob, expected): + actual = {path for path in path.rglob(glob, recurse_symlinks=True) + if path.parts.count("linkD") <= 1} # exclude symlink loop. + self.assertEqual(actual, { P(self.base, q) for q in expected }) + P = self.cls + p = P(self.base) + _check(p, "fileB", ["dirB/fileB", "dirA/linkC/fileB", "linkB/fileB", + "dirA/linkC/linkD/fileB", "dirB/linkD/fileB", "linkB/linkD/fileB"]) + _check(p, "*/fileA", []) + _check(p, "*/fileB", ["dirB/fileB", "dirA/linkC/fileB", "linkB/fileB", + "dirA/linkC/linkD/fileB", "dirB/linkD/fileB", "linkB/linkD/fileB"]) + _check(p, "file*", ["fileA", "dirA/linkC/fileB", "dirB/fileB", + "dirA/linkC/linkD/fileB", "dirB/linkD/fileB", "linkB/linkD/fileB", + "dirC/fileC", "dirC/dirD/fileD", "linkB/fileB"]) + _check(p, "*/", ["dirA/", "dirA/linkC/", "dirA/linkC/linkD/", "dirB/", "dirB/linkD/", + "dirC/", "dirC/dirD/", "dirE/", "linkB/", "linkB/linkD/"]) + _check(p, "", ["", "dirA/", "dirA/linkC/", "dirA/linkC/linkD/", "dirB/", "dirB/linkD/", + "dirC/", "dirE/", "dirC/dirD/", "linkB/", "linkB/linkD/"]) + + p = P(self.base, "dirC") + _check(p, "*", ["dirC/fileC", "dirC/novel.txt", + "dirC/dirD", "dirC/dirD/fileD"]) + _check(p, "file*", ["dirC/fileC", "dirC/dirD/fileD"]) + _check(p, "*/*", ["dirC/dirD/fileD"]) + _check(p, "*/", ["dirC/dirD/"]) + _check(p, "", ["dirC/", "dirC/dirD/"]) + # gh-91616, a re module regression + _check(p, "*.txt", ["dirC/novel.txt"]) + _check(p, "*.*", ["dirC/novel.txt"]) + + @needs_symlinks + def test_rglob_symlink_loop(self): + # Don't get fooled by symlink loops (Issue #26012). + P = self.cls + p = P(self.base) + given = set(p.rglob('*', recurse_symlinks=False)) + expect = {'brokenLink', + 'dirA', 'dirA/linkC', + 'dirB', 'dirB/fileB', 'dirB/linkD', + 'dirC', 'dirC/dirD', 'dirC/dirD/fileD', + 'dirC/fileC', 'dirC/novel.txt', + 'dirE', + 'fileA', + 'linkA', + 'linkB', + 'brokenLinkLoop', + } + self.assertEqual(given, {p / x for x in expect}) + + # See https://github.com/WebAssembly/wasi-filesystem/issues/26 + @unittest.skipIf(is_wasi, "WASI resolution of '..' parts doesn't match POSIX") + def test_glob_dotdot(self): + # ".." is not special in globs. + P = self.cls + p = P(self.base) + self.assertEqual(set(p.glob("..")), { P(self.base, "..") }) + self.assertEqual(set(p.glob("../..")), { P(self.base, "..", "..") }) + self.assertEqual(set(p.glob("dirA/..")), { P(self.base, "dirA", "..") }) + self.assertEqual(set(p.glob("dirA/../file*")), { P(self.base, "dirA/../fileA") }) + self.assertEqual(set(p.glob("dirA/../file*/..")), set()) + self.assertEqual(set(p.glob("../xyzzy")), set()) + if self.cls.parser is posixpath: + self.assertEqual(set(p.glob("xyzzy/..")), set()) + else: + # ".." segments are normalized first on Windows, so this path is stat()able. + self.assertEqual(set(p.glob("xyzzy/..")), { P(self.base, "xyzzy", "..") }) + self.assertEqual(set(p.glob("/".join([".."] * 50))), { P(self.base, *[".."] * 50)}) + + @needs_symlinks + def test_glob_permissions(self): + # See bpo-38894 + P = self.cls + base = P(self.base) / 'permissions' + base.mkdir() + + for i in range(100): + link = base / f"link{i}" + if i % 2: + link.symlink_to(P(self.base, "dirE", "nonexistent")) + else: + link.symlink_to(P(self.base, "dirC")) + + self.assertEqual(len(set(base.glob("*"))), 100) + self.assertEqual(len(set(base.glob("*/"))), 50) + self.assertEqual(len(set(base.glob("*/fileC"))), 50) + self.assertEqual(len(set(base.glob("*/file*"))), 50) + + @needs_symlinks + def test_glob_long_symlink(self): + # See gh-87695 + base = self.cls(self.base) / 'long_symlink' + base.mkdir() + bad_link = base / 'bad_link' + bad_link.symlink_to("bad" * 200) + self.assertEqual(sorted(base.glob('**/*')), [bad_link]) + + @needs_symlinks + def test_readlink(self): + P = self.cls(self.base) + self.assertEqual((P / 'linkA').readlink(), self.cls('fileA')) + self.assertEqual((P / 'brokenLink').readlink(), + self.cls('non-existing')) + self.assertEqual((P / 'linkB').readlink(), self.cls('dirB')) + self.assertEqual((P / 'linkB' / 'linkD').readlink(), self.cls('../dirB')) + with self.assertRaises(OSError): + (P / 'fileA').readlink() + + @unittest.skipIf(hasattr(os, "readlink"), "os.readlink() is present") + def test_readlink_unsupported(self): + P = self.cls(self.base) + p = P / 'fileA' + with self.assertRaises(UnsupportedOperation): + q.readlink(p) + + def _check_resolve(self, p, expected, strict=True): + q = p.resolve(strict) + self.assertEqual(q, expected) + + # This can be used to check both relative and absolute resolutions. + _check_resolve_relative = _check_resolve_absolute = _check_resolve + + @needs_symlinks + def test_resolve_common(self): + P = self.cls + p = P(self.base, 'foo') + with self.assertRaises(OSError) as cm: + p.resolve(strict=True) + self.assertEqual(cm.exception.errno, errno.ENOENT) + # Non-strict + parser = self.parser + self.assertEqualNormCase(str(p.resolve(strict=False)), + parser.join(self.base, 'foo')) + p = P(self.base, 'foo', 'in', 'spam') + self.assertEqualNormCase(str(p.resolve(strict=False)), + parser.join(self.base, 'foo', 'in', 'spam')) + p = P(self.base, '..', 'foo', 'in', 'spam') + self.assertEqualNormCase(str(p.resolve(strict=False)), + parser.join(parser.dirname(self.base), 'foo', 'in', 'spam')) + # These are all relative symlinks. + p = P(self.base, 'dirB', 'fileB') + self._check_resolve_relative(p, p) + p = P(self.base, 'linkA') + self._check_resolve_relative(p, P(self.base, 'fileA')) + p = P(self.base, 'dirA', 'linkC', 'fileB') + self._check_resolve_relative(p, P(self.base, 'dirB', 'fileB')) + p = P(self.base, 'dirB', 'linkD', 'fileB') + self._check_resolve_relative(p, P(self.base, 'dirB', 'fileB')) + # Non-strict + p = P(self.base, 'dirA', 'linkC', 'fileB', 'foo', 'in', 'spam') + self._check_resolve_relative(p, P(self.base, 'dirB', 'fileB', 'foo', 'in', + 'spam'), False) + p = P(self.base, 'dirA', 'linkC', '..', 'foo', 'in', 'spam') + if self.cls.parser is not posixpath: + # In Windows, if linkY points to dirB, 'dirA\linkY\..' + # resolves to 'dirA' without resolving linkY first. + self._check_resolve_relative(p, P(self.base, 'dirA', 'foo', 'in', + 'spam'), False) + else: + # In Posix, if linkY points to dirB, 'dirA/linkY/..' + # resolves to 'dirB/..' first before resolving to parent of dirB. + self._check_resolve_relative(p, P(self.base, 'foo', 'in', 'spam'), False) + # Now create absolute symlinks. + d = self.tempdir() + P(self.base, 'dirA', 'linkX').symlink_to(d) + P(self.base, str(d), 'linkY').symlink_to(self.parser.join(self.base, 'dirB')) + p = P(self.base, 'dirA', 'linkX', 'linkY', 'fileB') + self._check_resolve_absolute(p, P(self.base, 'dirB', 'fileB')) + # Non-strict + p = P(self.base, 'dirA', 'linkX', 'linkY', 'foo', 'in', 'spam') + self._check_resolve_relative(p, P(self.base, 'dirB', 'foo', 'in', 'spam'), + False) + p = P(self.base, 'dirA', 'linkX', 'linkY', '..', 'foo', 'in', 'spam') + if self.cls.parser is not posixpath: + # In Windows, if linkY points to dirB, 'dirA\linkY\..' + # resolves to 'dirA' without resolving linkY first. + self._check_resolve_relative(p, P(d, 'foo', 'in', 'spam'), False) + else: + # In Posix, if linkY points to dirB, 'dirA/linkY/..' + # resolves to 'dirB/..' first before resolving to parent of dirB. + self._check_resolve_relative(p, P(self.base, 'foo', 'in', 'spam'), False) + + @needs_symlinks + def test_resolve_dot(self): + # See http://web.archive.org/web/20200623062557/https://bitbucket.org/pitrou/pathlib/issues/9/ + parser = self.parser + p = self.cls(self.base) + p.joinpath('0').symlink_to('.', target_is_directory=True) + p.joinpath('1').symlink_to(parser.join('0', '0'), target_is_directory=True) + p.joinpath('2').symlink_to(parser.join('1', '1'), target_is_directory=True) + q = p / '2' + self.assertEqual(q.resolve(strict=True), p) + r = q / '3' / '4' + self.assertRaises(FileNotFoundError, r.resolve, strict=True) + # Non-strict + self.assertEqual(r.resolve(strict=False), p / '3' / '4') + + def _check_symlink_loop(self, *args): + path = self.cls(*args) + with self.assertRaises(OSError) as cm: + path.resolve(strict=True) + self.assertEqual(cm.exception.errno, errno.ELOOP) + + @needs_posix + @needs_symlinks + def test_resolve_loop(self): + # Loops with relative symlinks. + self.cls(self.base, 'linkX').symlink_to('linkX/inside') + self._check_symlink_loop(self.base, 'linkX') + self.cls(self.base, 'linkY').symlink_to('linkY') + self._check_symlink_loop(self.base, 'linkY') + self.cls(self.base, 'linkZ').symlink_to('linkZ/../linkZ') + self._check_symlink_loop(self.base, 'linkZ') + # Non-strict + p = self.cls(self.base, 'linkZ', 'foo') + self.assertEqual(p.resolve(strict=False), p) + # Loops with absolute symlinks. + self.cls(self.base, 'linkU').symlink_to(self.parser.join(self.base, 'linkU/inside')) + self._check_symlink_loop(self.base, 'linkU') + self.cls(self.base, 'linkV').symlink_to(self.parser.join(self.base, 'linkV')) + self._check_symlink_loop(self.base, 'linkV') + self.cls(self.base, 'linkW').symlink_to(self.parser.join(self.base, 'linkW/../linkW')) + self._check_symlink_loop(self.base, 'linkW') + # Non-strict + q = self.cls(self.base, 'linkW', 'foo') + self.assertEqual(q.resolve(strict=False), q) + + def test_stat(self): + statA = self.cls(self.base).joinpath('fileA').stat() + statB = self.cls(self.base).joinpath('dirB', 'fileB').stat() + statC = self.cls(self.base).joinpath('dirC').stat() + # st_mode: files are the same, directory differs. + self.assertIsInstance(statA.st_mode, int) + self.assertEqual(statA.st_mode, statB.st_mode) + self.assertNotEqual(statA.st_mode, statC.st_mode) + self.assertNotEqual(statB.st_mode, statC.st_mode) + # st_ino: all different, + self.assertIsInstance(statA.st_ino, int) + self.assertNotEqual(statA.st_ino, statB.st_ino) + self.assertNotEqual(statA.st_ino, statC.st_ino) + self.assertNotEqual(statB.st_ino, statC.st_ino) + # st_dev: all the same. + self.assertIsInstance(statA.st_dev, int) + self.assertEqual(statA.st_dev, statB.st_dev) + self.assertEqual(statA.st_dev, statC.st_dev) + # other attributes not used by pathlib. + + @needs_symlinks + def test_stat_no_follow_symlinks(self): + p = self.cls(self.base) / 'linkA' + st = p.stat() + self.assertNotEqual(st, p.stat(follow_symlinks=False)) + + def test_stat_no_follow_symlinks_nosymlink(self): + p = self.cls(self.base) / 'fileA' + st = p.stat() + self.assertEqual(st, p.stat(follow_symlinks=False)) + + @needs_symlinks + def test_lstat(self): + p = self.cls(self.base)/ 'linkA' + st = p.stat() + self.assertNotEqual(st, p.lstat()) + + def test_lstat_nosymlink(self): + p = self.cls(self.base) / 'fileA' + st = p.stat() + self.assertEqual(st, p.lstat()) + + def test_is_dir(self): + P = self.cls(self.base) + self.assertTrue((P / 'dirA').is_dir()) + self.assertFalse((P / 'fileA').is_dir()) + self.assertFalse((P / 'non-existing').is_dir()) + self.assertFalse((P / 'fileA' / 'bah').is_dir()) + if self.can_symlink: + self.assertFalse((P / 'linkA').is_dir()) + self.assertTrue((P / 'linkB').is_dir()) + self.assertFalse((P/ 'brokenLink').is_dir()) + self.assertFalse((P / 'dirA\udfff').is_dir()) + self.assertFalse((P / 'dirA\x00').is_dir()) + + def test_is_dir_no_follow_symlinks(self): + P = self.cls(self.base) + self.assertTrue((P / 'dirA').is_dir(follow_symlinks=False)) + self.assertFalse((P / 'fileA').is_dir(follow_symlinks=False)) + self.assertFalse((P / 'non-existing').is_dir(follow_symlinks=False)) + self.assertFalse((P / 'fileA' / 'bah').is_dir(follow_symlinks=False)) + if self.can_symlink: + self.assertFalse((P / 'linkA').is_dir(follow_symlinks=False)) + self.assertFalse((P / 'linkB').is_dir(follow_symlinks=False)) + self.assertFalse((P/ 'brokenLink').is_dir(follow_symlinks=False)) + self.assertFalse((P / 'dirA\udfff').is_dir(follow_symlinks=False)) + self.assertFalse((P / 'dirA\x00').is_dir(follow_symlinks=False)) + + def test_is_file(self): + P = self.cls(self.base) + self.assertTrue((P / 'fileA').is_file()) + self.assertFalse((P / 'dirA').is_file()) + self.assertFalse((P / 'non-existing').is_file()) + self.assertFalse((P / 'fileA' / 'bah').is_file()) + if self.can_symlink: + self.assertTrue((P / 'linkA').is_file()) + self.assertFalse((P / 'linkB').is_file()) + self.assertFalse((P/ 'brokenLink').is_file()) + self.assertFalse((P / 'fileA\udfff').is_file()) + self.assertFalse((P / 'fileA\x00').is_file()) + + def test_is_file_no_follow_symlinks(self): + P = self.cls(self.base) + self.assertTrue((P / 'fileA').is_file(follow_symlinks=False)) + self.assertFalse((P / 'dirA').is_file(follow_symlinks=False)) + self.assertFalse((P / 'non-existing').is_file(follow_symlinks=False)) + self.assertFalse((P / 'fileA' / 'bah').is_file(follow_symlinks=False)) + if self.can_symlink: + self.assertFalse((P / 'linkA').is_file(follow_symlinks=False)) + self.assertFalse((P / 'linkB').is_file(follow_symlinks=False)) + self.assertFalse((P/ 'brokenLink').is_file(follow_symlinks=False)) + self.assertFalse((P / 'fileA\udfff').is_file(follow_symlinks=False)) + self.assertFalse((P / 'fileA\x00').is_file(follow_symlinks=False)) + + def test_is_mount(self): + P = self.cls(self.base) + self.assertFalse((P / 'fileA').is_mount()) + self.assertFalse((P / 'dirA').is_mount()) + self.assertFalse((P / 'non-existing').is_mount()) + self.assertFalse((P / 'fileA' / 'bah').is_mount()) + if self.can_symlink: + self.assertFalse((P / 'linkA').is_mount()) + + def test_is_symlink(self): + P = self.cls(self.base) + self.assertFalse((P / 'fileA').is_symlink()) + self.assertFalse((P / 'dirA').is_symlink()) + self.assertFalse((P / 'non-existing').is_symlink()) + self.assertFalse((P / 'fileA' / 'bah').is_symlink()) + if self.can_symlink: + self.assertTrue((P / 'linkA').is_symlink()) + self.assertTrue((P / 'linkB').is_symlink()) + self.assertTrue((P/ 'brokenLink').is_symlink()) + self.assertIs((P / 'fileA\udfff').is_file(), False) + self.assertIs((P / 'fileA\x00').is_file(), False) + if self.can_symlink: + self.assertIs((P / 'linkA\udfff').is_file(), False) + self.assertIs((P / 'linkA\x00').is_file(), False) + + def test_is_junction_false(self): + P = self.cls(self.base) + self.assertFalse((P / 'fileA').is_junction()) + self.assertFalse((P / 'dirA').is_junction()) + self.assertFalse((P / 'non-existing').is_junction()) + self.assertFalse((P / 'fileA' / 'bah').is_junction()) + self.assertFalse((P / 'fileA\udfff').is_junction()) + self.assertFalse((P / 'fileA\x00').is_junction()) + + def test_is_fifo_false(self): + P = self.cls(self.base) + self.assertFalse((P / 'fileA').is_fifo()) + self.assertFalse((P / 'dirA').is_fifo()) + self.assertFalse((P / 'non-existing').is_fifo()) + self.assertFalse((P / 'fileA' / 'bah').is_fifo()) + self.assertIs((P / 'fileA\udfff').is_fifo(), False) + self.assertIs((P / 'fileA\x00').is_fifo(), False) + + def test_is_socket_false(self): + P = self.cls(self.base) + self.assertFalse((P / 'fileA').is_socket()) + self.assertFalse((P / 'dirA').is_socket()) + self.assertFalse((P / 'non-existing').is_socket()) + self.assertFalse((P / 'fileA' / 'bah').is_socket()) + self.assertIs((P / 'fileA\udfff').is_socket(), False) + self.assertIs((P / 'fileA\x00').is_socket(), False) + + def test_is_block_device_false(self): + P = self.cls(self.base) + self.assertFalse((P / 'fileA').is_block_device()) + self.assertFalse((P / 'dirA').is_block_device()) + self.assertFalse((P / 'non-existing').is_block_device()) + self.assertFalse((P / 'fileA' / 'bah').is_block_device()) + self.assertIs((P / 'fileA\udfff').is_block_device(), False) + self.assertIs((P / 'fileA\x00').is_block_device(), False) + + def test_is_char_device_false(self): + P = self.cls(self.base) + self.assertFalse((P / 'fileA').is_char_device()) + self.assertFalse((P / 'dirA').is_char_device()) + self.assertFalse((P / 'non-existing').is_char_device()) + self.assertFalse((P / 'fileA' / 'bah').is_char_device()) + self.assertIs((P / 'fileA\udfff').is_char_device(), False) + self.assertIs((P / 'fileA\x00').is_char_device(), False) + + def _check_complex_symlinks(self, link0_target): + # Test solving a non-looping chain of symlinks (issue #19887). + parser = self.parser + P = self.cls(self.base) + P.joinpath('link1').symlink_to(parser.join('link0', 'link0'), target_is_directory=True) + P.joinpath('link2').symlink_to(parser.join('link1', 'link1'), target_is_directory=True) + P.joinpath('link3').symlink_to(parser.join('link2', 'link2'), target_is_directory=True) + P.joinpath('link0').symlink_to(link0_target, target_is_directory=True) + + # Resolve absolute paths. + p = (P / 'link0').resolve() + self.assertEqual(p, P) + self.assertEqualNormCase(str(p), self.base) + p = (P / 'link1').resolve() + self.assertEqual(p, P) + self.assertEqualNormCase(str(p), self.base) + p = (P / 'link2').resolve() + self.assertEqual(p, P) + self.assertEqualNormCase(str(p), self.base) + p = (P / 'link3').resolve() + self.assertEqual(p, P) + self.assertEqualNormCase(str(p), self.base) + + # Resolve relative paths. + try: + self.cls('').absolute() + except UnsupportedOperation: + return + old_path = os.getcwd() + os.chdir(self.base) + try: + p = self.cls('link0').resolve() + self.assertEqual(p, P) + self.assertEqualNormCase(str(p), self.base) + p = self.cls('link1').resolve() + self.assertEqual(p, P) + self.assertEqualNormCase(str(p), self.base) + p = self.cls('link2').resolve() + self.assertEqual(p, P) + self.assertEqualNormCase(str(p), self.base) + p = self.cls('link3').resolve() + self.assertEqual(p, P) + self.assertEqualNormCase(str(p), self.base) + finally: + os.chdir(old_path) + + @needs_symlinks + def test_complex_symlinks_absolute(self): + self._check_complex_symlinks(self.base) + + @needs_symlinks + def test_complex_symlinks_relative(self): + self._check_complex_symlinks('.') + + @needs_symlinks + def test_complex_symlinks_relative_dot_dot(self): + self._check_complex_symlinks(self.parser.join('dirA', '..')) + + def setUpWalk(self): + # Build: + # TESTFN/ + # TEST1/ a file kid and two directory kids + # tmp1 + # SUB1/ a file kid and a directory kid + # tmp2 + # SUB11/ no kids + # SUB2/ a file kid and a dirsymlink kid + # tmp3 + # link/ a symlink to TEST2 + # broken_link + # broken_link2 + # TEST2/ + # tmp4 a lone file + self.walk_path = self.cls(self.base, "TEST1") + self.sub1_path = self.walk_path / "SUB1" + self.sub11_path = self.sub1_path / "SUB11" + self.sub2_path = self.walk_path / "SUB2" + tmp1_path = self.walk_path / "tmp1" + tmp2_path = self.sub1_path / "tmp2" + tmp3_path = self.sub2_path / "tmp3" + self.link_path = self.sub2_path / "link" + t2_path = self.cls(self.base, "TEST2") + tmp4_path = self.cls(self.base, "TEST2", "tmp4") + broken_link_path = self.sub2_path / "broken_link" + broken_link2_path = self.sub2_path / "broken_link2" + + self.sub11_path.mkdir(parents=True) + self.sub2_path.mkdir(parents=True) + t2_path.mkdir(parents=True) + + for path in tmp1_path, tmp2_path, tmp3_path, tmp4_path: + with path.open("w", encoding='utf-8') as f: + f.write(f"I'm {path} and proud of it. Blame test_pathlib.\n") + + if self.can_symlink: + self.link_path.symlink_to(t2_path) + broken_link_path.symlink_to('broken') + broken_link2_path.symlink_to(self.cls('tmp3', 'broken')) + self.sub2_tree = (self.sub2_path, [], ["broken_link", "broken_link2", "link", "tmp3"]) + else: + self.sub2_tree = (self.sub2_path, [], ["tmp3"]) + + def test_walk_topdown(self): + self.setUpWalk() + walker = self.walk_path.walk() + entry = next(walker) + entry[1].sort() # Ensure we visit SUB1 before SUB2 + self.assertEqual(entry, (self.walk_path, ["SUB1", "SUB2"], ["tmp1"])) + entry = next(walker) + self.assertEqual(entry, (self.sub1_path, ["SUB11"], ["tmp2"])) + entry = next(walker) + self.assertEqual(entry, (self.sub11_path, [], [])) + entry = next(walker) + entry[1].sort() + entry[2].sort() + self.assertEqual(entry, self.sub2_tree) + with self.assertRaises(StopIteration): + next(walker) + + def test_walk_prune(self): + self.setUpWalk() + # Prune the search. + all = [] + for root, dirs, files in self.walk_path.walk(): + all.append((root, dirs, files)) + if 'SUB1' in dirs: + # Note that this also mutates the dirs we appended to all! + dirs.remove('SUB1') + + self.assertEqual(len(all), 2) + self.assertEqual(all[0], (self.walk_path, ["SUB2"], ["tmp1"])) + + all[1][-1].sort() + all[1][1].sort() + self.assertEqual(all[1], self.sub2_tree) + + def test_walk_bottom_up(self): + self.setUpWalk() + seen_testfn = seen_sub1 = seen_sub11 = seen_sub2 = False + for path, dirnames, filenames in self.walk_path.walk(top_down=False): + if path == self.walk_path: + self.assertFalse(seen_testfn) + self.assertTrue(seen_sub1) + self.assertTrue(seen_sub2) + self.assertEqual(sorted(dirnames), ["SUB1", "SUB2"]) + self.assertEqual(filenames, ["tmp1"]) + seen_testfn = True + elif path == self.sub1_path: + self.assertFalse(seen_testfn) + self.assertFalse(seen_sub1) + self.assertTrue(seen_sub11) + self.assertEqual(dirnames, ["SUB11"]) + self.assertEqual(filenames, ["tmp2"]) + seen_sub1 = True + elif path == self.sub11_path: + self.assertFalse(seen_sub1) + self.assertFalse(seen_sub11) + self.assertEqual(dirnames, []) + self.assertEqual(filenames, []) + seen_sub11 = True + elif path == self.sub2_path: + self.assertFalse(seen_testfn) + self.assertFalse(seen_sub2) + self.assertEqual(sorted(dirnames), sorted(self.sub2_tree[1])) + self.assertEqual(sorted(filenames), sorted(self.sub2_tree[2])) + seen_sub2 = True + else: + raise AssertionError(f"Unexpected path: {path}") + self.assertTrue(seen_testfn) + + @needs_symlinks + def test_walk_follow_symlinks(self): + self.setUpWalk() + walk_it = self.walk_path.walk(follow_symlinks=True) + for root, dirs, files in walk_it: + if root == self.link_path: + self.assertEqual(dirs, []) + self.assertEqual(files, ["tmp4"]) + break + else: + self.fail("Didn't follow symlink with follow_symlinks=True") + + @needs_symlinks + def test_walk_symlink_location(self): + self.setUpWalk() + # Tests whether symlinks end up in filenames or dirnames depending + # on the `follow_symlinks` argument. + walk_it = self.walk_path.walk(follow_symlinks=False) + for root, dirs, files in walk_it: + if root == self.sub2_path: + self.assertIn("link", files) + break + else: + self.fail("symlink not found") + + walk_it = self.walk_path.walk(follow_symlinks=True) + for root, dirs, files in walk_it: + if root == self.sub2_path: + self.assertIn("link", dirs) + break + else: + self.fail("symlink not found") + + +class DummyPathWithSymlinks(DummyPath): + __slots__ = () + + # Reduce symlink traversal limit to make tests run faster. + _max_symlinks = 20 + + def readlink(self): + path = str(self.parent.resolve() / self.name) + if path in self._symlinks: + return self.with_segments(self._symlinks[path]) + elif path in self._files or path in self._directories: + raise OSError(errno.EINVAL, "Not a symlink", path) + else: + raise FileNotFoundError(errno.ENOENT, "File not found", path) + + def symlink_to(self, target, target_is_directory=False): + self._directories[str(self.parent)].add(self.name) + self._symlinks[str(self)] = str(target) + + +class DummyPathWithSymlinksTest(DummyPathTest): + cls = DummyPathWithSymlinks + can_symlink = True + + +if __name__ == "__main__": + unittest.main() diff --git a/Lib/test/test_urllib.py b/Lib/test/test_urllib.py index 82f1d9dc2e7..aee9fb78017 100644 --- a/Lib/test/test_urllib.py +++ b/Lib/test/test_urllib.py @@ -10,7 +10,9 @@ from unittest.mock import patch from test import support from test.support import os_helper +from test.support import socket_helper from test.support import warnings_helper +from test.support.testcase import ExtraAssertions import os try: import ssl @@ -18,12 +20,15 @@ ssl = None import sys import tempfile -from nturl2path import url2pathname, pathname2url from base64 import b64encode import collections +if not socket_helper.has_gethostname: + raise unittest.SkipTest("test requires gethostname()") + + def hexescape(char): """Escape char as RFC 2396 specifies""" hex_repr = hex(ord(char))[2:].upper() @@ -135,7 +140,7 @@ def unfakeftp(self): urllib.request.ftpwrapper = self._ftpwrapper_class -class urlopen_FileTests(unittest.TestCase): +class urlopen_FileTests(unittest.TestCase, ExtraAssertions): """Test urlopen() opening a temporary file. Try to test as much functionality as possible so as to cut down on reliance @@ -153,7 +158,7 @@ def setUp(self): finally: f.close() self.pathname = os_helper.TESTFN - self.quoted_pathname = urllib.parse.quote(self.pathname) + self.quoted_pathname = urllib.parse.quote(os.fsencode(self.pathname)) self.returned_obj = urlopen("file:%s" % self.quoted_pathname) def tearDown(self): @@ -165,9 +170,7 @@ def test_interface(self): # Make sure object returned by urlopen() has the specified methods for attr in ("read", "readline", "readlines", "fileno", "close", "info", "geturl", "getcode", "__iter__"): - self.assertTrue(hasattr(self.returned_obj, attr), - "object returned by urlopen() lacks %s attribute" % - attr) + self.assertHasAttr(self.returned_obj, attr) def test_read(self): self.assertEqual(self.text, self.returned_obj.read()) @@ -232,17 +235,12 @@ class ProxyTests(unittest.TestCase): def setUp(self): # Records changes to env vars - self.env = os_helper.EnvironmentVarGuard() + self.env = self.enterContext(os_helper.EnvironmentVarGuard()) # Delete all proxy related env vars for k in list(os.environ): if 'proxy' in k.lower(): self.env.unset(k) - def tearDown(self): - # Restore all proxy related env vars - self.env.__exit__() - del self.env - def test_getproxies_environment_keep_no_proxies(self): self.env.set('NO_PROXY', 'localhost') proxies = urllib.request.getproxies_environment() @@ -477,7 +475,9 @@ def test_read_bogus(self): Content-Type: text/html; charset=iso-8859-1 ''', mock_close=True) try: - self.assertRaises(OSError, urlopen, "http://python.org/") + with self.assertRaises(urllib.error.HTTPError) as cm: + urllib.request.urlopen("http://python.org/") + cm.exception.close() finally: self.unfakehttp() @@ -492,8 +492,9 @@ def test_invalid_redirect(self): ''', mock_close=True) try: msg = "Redirection to url 'file:" - with self.assertRaisesRegex(urllib.error.HTTPError, msg): - urlopen("http://python.org/") + with self.assertRaisesRegex(urllib.error.HTTPError, msg) as cm: + urllib.request.urlopen("http://python.org/") + cm.exception.close() finally: self.unfakehttp() @@ -506,8 +507,9 @@ def test_redirect_limit_independent(self): Connection: close ''', mock_close=True) try: - self.assertRaises(urllib.error.HTTPError, urlopen, - "http://something") + with self.assertRaises(urllib.error.HTTPError) as cm: + urllib.request.urlopen("http://something") + cm.exception.close() finally: self.unfakehttp() @@ -597,17 +599,8 @@ def test_URLopener_deprecation(self): with warnings_helper.check_warnings(('',DeprecationWarning)): urllib.request.URLopener() - @unittest.skipUnless(ssl, "ssl module required") - def test_cafile_and_context(self): - context = ssl.create_default_context() - with warnings_helper.check_warnings(('', DeprecationWarning)): - with self.assertRaises(ValueError): - urllib.request.urlopen( - "https://localhost", cafile="/nonexistent/path", context=context - ) - -class urlopen_DataTests(unittest.TestCase): +class urlopen_DataTests(unittest.TestCase, ExtraAssertions): """Test urlopen() opening a data URL.""" def setUp(self): @@ -636,18 +629,17 @@ def setUp(self): "QOjdAAAAAXNSR0IArs4c6QAAAA9JREFUCNdj%0AYGBg%2BP//PwAGAQL%2BCm8 " "vHgAAAABJRU5ErkJggg%3D%3D%0A%20") - self.text_url_resp = urllib.request.urlopen(self.text_url) - self.text_url_base64_resp = urllib.request.urlopen( - self.text_url_base64) - self.image_url_resp = urllib.request.urlopen(self.image_url) + self.text_url_resp = self.enterContext( + urllib.request.urlopen(self.text_url)) + self.text_url_base64_resp = self.enterContext( + urllib.request.urlopen(self.text_url_base64)) + self.image_url_resp = self.enterContext(urllib.request.urlopen(self.image_url)) def test_interface(self): # Make sure object returned by urlopen() has the specified methods for attr in ("read", "readline", "readlines", "close", "info", "geturl", "getcode", "__iter__"): - self.assertTrue(hasattr(self.text_url_resp, attr), - "object returned by urlopen() lacks %s attribute" % - attr) + self.assertHasAttr(self.text_url_resp, attr) def test_info(self): self.assertIsInstance(self.text_url_resp.info(), email.message.Message) @@ -655,8 +647,10 @@ def test_info(self): [('text/plain', ''), ('charset', 'ISO-8859-1')]) self.assertEqual(self.image_url_resp.info()['content-length'], str(len(self.image))) - self.assertEqual(urllib.request.urlopen("data:,").info().get_params(), + r = urllib.request.urlopen("data:,") + self.assertEqual(r.info().get_params(), [('text/plain', ''), ('charset', 'US-ASCII')]) + r.close() def test_geturl(self): self.assertEqual(self.text_url_resp.geturl(), self.text_url) @@ -719,10 +713,6 @@ def tearDown(self): def constructLocalFileUrl(self, filePath): filePath = os.path.abspath(filePath) - try: - filePath.encode("utf-8") - except UnicodeEncodeError: - raise unittest.SkipTest("filePath is not encodable to utf8") return "file://%s" % urllib.request.pathname2url(filePath) def createNewTempFile(self, data=b""): @@ -1104,6 +1094,8 @@ def test_unquoting(self): self.assertEqual(result.count('%'), 1, "using unquote(): not all characters escaped: " "%s" % result) + + def test_unquote_rejects_none_and_tuple(self): self.assertRaises((TypeError, AttributeError), urllib.parse.unquote, None) self.assertRaises((TypeError, AttributeError), urllib.parse.unquote, ()) @@ -1527,39 +1519,121 @@ def test_quoting(self): (expect, result)) @unittest.skipUnless(sys.platform == 'win32', - 'test specific to the nturl2path functions.') - def test_prefixes(self): + 'test specific to Windows pathnames.') + def test_pathname2url_win(self): # Test special prefixes are correctly handled in pathname2url() - given = '\\\\?\\C:\\dir' - expect = '///C:/dir' - result = urllib.request.pathname2url(given) - self.assertEqual(expect, result, - "pathname2url() failed; %s != %s" % - (expect, result)) - given = '\\\\?\\unc\\server\\share\\dir' - expect = '/server/share/dir' - result = urllib.request.pathname2url(given) - self.assertEqual(expect, result, - "pathname2url() failed; %s != %s" % - (expect, result)) - + fn = urllib.request.pathname2url + self.assertEqual(fn('\\\\?\\C:\\dir'), '///C:/dir') + self.assertEqual(fn('\\\\?\\unc\\server\\share\\dir'), '//server/share/dir') + self.assertEqual(fn("C:"), '///C:') + self.assertEqual(fn("C:\\"), '///C:/') + self.assertEqual(fn('C:\\a\\b.c'), '///C:/a/b.c') + self.assertEqual(fn('C:\\a\\b.c\\'), '///C:/a/b.c/') + self.assertEqual(fn('C:\\a\\\\b.c'), '///C:/a//b.c') + self.assertEqual(fn('C:\\a\\b%#c'), '///C:/a/b%25%23c') + self.assertEqual(fn('C:\\a\\b\xe9'), '///C:/a/b%C3%A9') + self.assertEqual(fn('C:\\foo\\bar\\spam.foo'), "///C:/foo/bar/spam.foo") + # Long drive letter + self.assertRaises(IOError, fn, "XX:\\") + # No drive letter + self.assertEqual(fn("\\folder\\test\\"), '/folder/test/') + self.assertEqual(fn("\\\\folder\\test\\"), '//folder/test/') + self.assertEqual(fn("\\\\\\folder\\test\\"), '///folder/test/') + self.assertEqual(fn('\\\\some\\share\\'), '//some/share/') + self.assertEqual(fn('\\\\some\\share\\a\\b.c'), '//some/share/a/b.c') + self.assertEqual(fn('\\\\some\\share\\a\\b%#c\xe9'), '//some/share/a/b%25%23c%C3%A9') + # Alternate path separator + self.assertEqual(fn('C:/a/b.c'), '///C:/a/b.c') + self.assertEqual(fn('//some/share/a/b.c'), '//some/share/a/b.c') + self.assertEqual(fn('//?/C:/dir'), '///C:/dir') + self.assertEqual(fn('//?/unc/server/share/dir'), '//server/share/dir') + # Round-tripping + urls = ['///C:', + '/folder/test/', + '///C:/foo/bar/spam.foo'] + for url in urls: + self.assertEqual(fn(urllib.request.url2pathname(url)), url) + + @unittest.skipIf(sys.platform == 'win32', + 'test specific to POSIX pathnames') + @unittest.expectedFailure # AssertionError: '//a/b.c' != '////a/b.c' + def test_pathname2url_posix(self): + fn = urllib.request.pathname2url + self.assertEqual(fn('/'), '/') + self.assertEqual(fn('/a/b.c'), '/a/b.c') + self.assertEqual(fn('//a/b.c'), '////a/b.c') + self.assertEqual(fn('///a/b.c'), '/////a/b.c') + self.assertEqual(fn('////a/b.c'), '//////a/b.c') + self.assertEqual(fn('/a/b%#c'), '/a/b%25%23c') + + @unittest.skipUnless(os_helper.FS_NONASCII, 'need os_helper.FS_NONASCII') + def test_pathname2url_nonascii(self): + encoding = sys.getfilesystemencoding() + errors = sys.getfilesystemencodeerrors() + url = urllib.parse.quote(os_helper.FS_NONASCII, encoding=encoding, errors=errors) + self.assertEqual(urllib.request.pathname2url(os_helper.FS_NONASCII), url) @unittest.skipUnless(sys.platform == 'win32', - 'test specific to the urllib.url2path function.') - def test_ntpath(self): - given = ('/C:/', '///C:/', '/C|//') - expect = 'C:\\' - for url in given: - result = urllib.request.url2pathname(url) - self.assertEqual(expect, result, - 'urllib.request..url2pathname() failed; %s != %s' % - (expect, result)) - given = '///C|/path' - expect = 'C:\\path' - result = urllib.request.url2pathname(given) - self.assertEqual(expect, result, - 'urllib.request.url2pathname() failed; %s != %s' % - (expect, result)) + 'test specific to Windows pathnames.') + def test_url2pathname_win(self): + fn = urllib.request.url2pathname + self.assertEqual(fn('/C:/'), 'C:\\') + self.assertEqual(fn("///C|"), 'C:') + self.assertEqual(fn("///C:"), 'C:') + self.assertEqual(fn('///C:/'), 'C:\\') + self.assertEqual(fn('/C|//'), 'C:\\\\') + self.assertEqual(fn('///C|/path'), 'C:\\path') + # No DOS drive + self.assertEqual(fn("///C/test/"), '\\C\\test\\') + self.assertEqual(fn("////C/test/"), '\\\\C\\test\\') + # DOS drive paths + self.assertEqual(fn('C:/path/to/file'), 'C:\\path\\to\\file') + self.assertEqual(fn('C:/path/to/file/'), 'C:\\path\\to\\file\\') + self.assertEqual(fn('C:/path/to//file'), 'C:\\path\\to\\\\file') + self.assertEqual(fn('C|/path/to/file'), 'C:\\path\\to\\file') + self.assertEqual(fn('/C|/path/to/file'), 'C:\\path\\to\\file') + self.assertEqual(fn('///C|/path/to/file'), 'C:\\path\\to\\file') + self.assertEqual(fn("///C|/foo/bar/spam.foo"), 'C:\\foo\\bar\\spam.foo') + # Non-ASCII drive letter + self.assertRaises(IOError, fn, "///\u00e8|/") + # UNC paths + self.assertEqual(fn('//server/path/to/file'), '\\\\server\\path\\to\\file') + self.assertEqual(fn('////server/path/to/file'), '\\\\server\\path\\to\\file') + self.assertEqual(fn('/////server/path/to/file'), '\\\\server\\path\\to\\file') + # Localhost paths + self.assertEqual(fn('//localhost/C:/path/to/file'), 'C:\\path\\to\\file') + self.assertEqual(fn('//localhost/C|/path/to/file'), 'C:\\path\\to\\file') + self.assertEqual(fn('//localhost/path/to/file'), '\\path\\to\\file') + self.assertEqual(fn('//localhost//server/path/to/file'), '\\\\server\\path\\to\\file') + # Percent-encoded forward slashes are preserved for backwards compatibility + self.assertEqual(fn('C:/foo%2fbar'), 'C:\\foo/bar') + self.assertEqual(fn('//server/share/foo%2fbar'), '\\\\server\\share\\foo/bar') + # Round-tripping + paths = ['C:', + r'\C\test\\', + r'C:\foo\bar\spam.foo'] + for path in paths: + self.assertEqual(fn(urllib.request.pathname2url(path)), path) + + @unittest.skipIf(sys.platform == 'win32', + 'test specific to POSIX pathnames') + @unittest.expectedFailure # AssertionError: '///foo/bar' != '/foo/bar' + def test_url2pathname_posix(self): + fn = urllib.request.url2pathname + self.assertEqual(fn('/foo/bar'), '/foo/bar') + self.assertEqual(fn('//foo/bar'), '//foo/bar') + self.assertEqual(fn('///foo/bar'), '/foo/bar') + self.assertEqual(fn('////foo/bar'), '//foo/bar') + self.assertEqual(fn('//localhost/foo/bar'), '/foo/bar') + + @unittest.skipUnless(os_helper.FS_NONASCII, 'need os_helper.FS_NONASCII') + def test_url2pathname_nonascii(self): + encoding = sys.getfilesystemencoding() + errors = sys.getfilesystemencodeerrors() + url = os_helper.FS_NONASCII + self.assertEqual(urllib.request.url2pathname(url), os_helper.FS_NONASCII) + url = urllib.parse.quote(url, encoding=encoding, errors=errors) + self.assertEqual(urllib.request.url2pathname(url), os_helper.FS_NONASCII) class Utility_Tests(unittest.TestCase): """Testcase to test the various utility functions in the urllib.""" @@ -1643,60 +1717,5 @@ def test_with_method_arg(self): self.assertEqual(request.get_method(), 'HEAD') -class URL2PathNameTests(unittest.TestCase): - - def test_converting_drive_letter(self): - self.assertEqual(url2pathname("///C|"), 'C:') - self.assertEqual(url2pathname("///C:"), 'C:') - self.assertEqual(url2pathname("///C|/"), 'C:\\') - - def test_converting_when_no_drive_letter(self): - # cannot end a raw string in \ - self.assertEqual(url2pathname("///C/test/"), r'\\\C\test' '\\') - self.assertEqual(url2pathname("////C/test/"), r'\\C\test' '\\') - - def test_simple_compare(self): - self.assertEqual(url2pathname("///C|/foo/bar/spam.foo"), - r'C:\foo\bar\spam.foo') - - def test_non_ascii_drive_letter(self): - self.assertRaises(IOError, url2pathname, "///\u00e8|/") - - def test_roundtrip_url2pathname(self): - list_of_paths = ['C:', - r'\\\C\test\\', - r'C:\foo\bar\spam.foo' - ] - for path in list_of_paths: - self.assertEqual(url2pathname(pathname2url(path)), path) - -class PathName2URLTests(unittest.TestCase): - - def test_converting_drive_letter(self): - self.assertEqual(pathname2url("C:"), '///C:') - self.assertEqual(pathname2url("C:\\"), '///C:') - - def test_converting_when_no_drive_letter(self): - self.assertEqual(pathname2url(r"\\\folder\test" "\\"), - '/////folder/test/') - self.assertEqual(pathname2url(r"\\folder\test" "\\"), - '////folder/test/') - self.assertEqual(pathname2url(r"\folder\test" "\\"), - '/folder/test/') - - def test_simple_compare(self): - self.assertEqual(pathname2url(r'C:\foo\bar\spam.foo'), - "///C:/foo/bar/spam.foo" ) - - def test_long_drive_letter(self): - self.assertRaises(IOError, pathname2url, "XX:\\") - - def test_roundtrip_pathname2url(self): - list_of_paths = ['///C:', - '/////folder/test/', - '///C:/foo/bar/spam.foo'] - for path in list_of_paths: - self.assertEqual(pathname2url(url2pathname(path)), path) - if __name__ == '__main__': unittest.main()