typeddfs.utils.checksum_models.ChecksumFile.delete() - Code Metrics - dmyersturnbull/typed-dfs - Measure and Improve Code Quality continuously with Scrutinizer

ChecksumFile.delete() A
last analyzed 2023-10-24 08:21 UTC

↳ Parent: typeddfs.utils.checksum_models

Complexity

Conditions

Size

Total Lines	8
Code Lines	2

Duplication

Lines	0
Ratio	0 %

Importance

Changes

Metric	Value
cc	1
eloc	2
nop	1
dl	0
loc	8
rs	10
c	0
b	0
f	0

# SPDX-FileCopyrightText: Copyright 2020-2023, Contributors to typed-dfs
# SPDX-PackageHomePage: https://github.com/dmyersturnbull/typed-dfs
# SPDX-License-Identifier: Apache-2.0
"""
Models for shasum-like files.
"""
from __future__ import annotations

from dataclasses import dataclass
from pathlib import Path, PurePath
from typing import TYPE_CHECKING

import regex

from typeddfs.df_errors import (
    HashContradictsExistingError,
    HashDidNotValidateError,
    HashExistsError,
    HashFileMissingError,
    HashFilenameMissingError,
    PathNotRelativeError,
)

if TYPE_CHECKING:
    from collections.abc import Callable, Iterable, Mapping, Sequence, ValuesView

    from typeddfs.utils._utils import PathLike

_hex_pattern = regex.compile(r"[A-Ha-h0-9]+", flags=regex.V1)
_hashsum_file_sep = regex.compile(r" [ *]", flags=regex.V1)


@dataclass(frozen=True, slots=True)
class _ChecksumMapping:
    hash_path: Path
    _dct: Mapping[Path, str]

    def __post_init__(self):
        try:
            for p in self._dct:
                # will error if it's not
                p.relative_to(self.directory)
        except ValueError as e:
            msg = f"{e}: Full contents are {self._dct}"
            raise PathNotRelativeError(msg) from e

    def lines(self) -> Sequence[str]:
        """
        Returns the text that would be written for this .shasum-like file.
        Calls :meth:`unresolve` first.
        """
        unresolved = self.unresolve()
        return [f"{v} *{p.name}" for p, v in unresolved._dct.items()]

    def line(self, path: PathLike) -> str:
        """
        Returns the text that would be written for a single path in a .shasum-like file.
        """
        path = Path(path)
        v = self._dct[path]
        return f"{v} *{path.name}"

    @property
    def directory(self) -> Path:
        return self.hash_path.parent

    def resolve(self) -> __qualname__:
        """
        Calls ``pathlib.Path.resolve()`` on all paths.
        This will follow symlinks, etc.
        """
        # noinspection PyArgumentList
        return self.__class__(
            self.hash_path.resolve(),
            {k.resolve(): v for k, v in self._dct.items()},
        )

    def unresolve(self) -> __qualname__:
        """
        Each path becomes its filename under ``self.directory``.
        This means that the parent nodes of each path are discarded in favor of ``self.directory``.

        Raises:
            ValueError: If two paths are "un-resolved" to the same path

        .. note::
            This will not work correctly with subdirectories
        """
        dct = {}
        for k, v in self._dct.items():
            k = self.directory / k.name
            if k in dct:
                msg = f"At least 2 paths resolve to {k}"
                raise ValueError(msg)
            dct[k] = v
        # noinspection PyArgumentList
        return self.__class__(self.hash_path, dct)

    @classmethod
    def _parse(
        cls,
        path: Path,
        *,
        lines: Sequence[str] | None = None,
        missing_ok: bool = False,
        subdirs: bool = False,
    ) -> __qualname__:
        path = Path(path)
        if lines is None and path.exists():
            lines = path.read_text(encoding="utf-8").splitlines()
        elif missing_ok and lines is None:
            lines = []
        elif lines is None:
            msg = f"Hash file {path} not found"
            raise HashFileMissingError(msg)

        # ignore spaces -- editors often add an extra line break, and it's probably fine anyway
        read = [_hashsum_file_sep.split(s, 1) for s in lines if len(s) > 0]
        # obviously this means that / can't appear in a node
        # this is consistent with the commonly accepted spec for shasum
        # does not handle root (beginning with /)
        kv = {
            Path(*[n for n in r[1].strip().split("/") if n != "."]): r[0]
            for r in read
            if len(r[0]) != 0
        }
        if not subdirs:
            slashed = {k for k in kv if len(k.parts) > 1}
            if len(slashed) > 0:
                msg = f"Subdirectory (containing /): {slashed} in {path}"
                raise ValueError(msg)
        kv = {Path(path.parent, p): v for p, v in kv.items()}
        return cls(path, kv)

    def _get_updated(
        self,
        *,
        path: PathLike,
        new: str | None,
        overwrite: bool | None,
        missing_ok: bool,
    ) -> str | None:
        path = Path(path)
        z = self._dct.get(path)
        if z is None and not missing_ok:
            msg = f"{path} not found ({len(self._dct)} are)"
            raise HashFilenameMissingError(msg)
        if z is not None:
            err = None
            if overwrite is None and z != new:
                err = (HashContradictsExistingError, f"Hash for {path} exists but does not match")
            elif overwrite is False:
                err = (
                    HashExistsError,
                    f"Hash for {path} exists ({'matches' if z == new else 'differs'})",
                )
            if err is not None:
                raise err[0](err[1], key=str(path), original=z, new=new)
        return new


@dataclass(frozen=True, slots=True)
class ChecksumFile(_ChecksumMapping):
    def load(self) -> __qualname__:
        """
        Returns a copy of ``self`` read from :attr:`hash_path`.
        """
        return self.__class__.parse(self.hash_path)

    @classmethod
    def parse(
        cls,
        path: Path,
        *,
        lines: Sequence[str] | None = None,
    ) -> __qualname__:
        """
        Reads hash file contents.

        Args:
            path: The path of the checksum file; required to resolve paths relative to its parent
            lines: The lines in the checksum file; reads ``path`` if None

        Returns:
            A ChecksumFile
        """
        return cls._parse(path, lines=lines, missing_ok=False, subdirs=False)

    @classmethod
    def new(
        cls,
        hash_path: PathLike,
        file_path: PathLike,
        hash_value: str,
    ) -> ChecksumFile:
        """
        Use this as a constructor.
        """
        hash_path = Path(hash_path)
        return cls(hash_path, {Path(file_path): hash_value})

    def rename(self, path: Path) -> __qualname__:
        """
        Replaces :attr:`self.file_path` with ``path``.
        This will affect the filename written in a .shasum-like file.
        No OS operations are performed.
        """
        return self.new(self.hash_path, file_path=path, hash_value=self.hash_value)

    def update(self, value: str, overwrite: bool | None = True) -> __qualname__:
        """
        Modifies the hash.

        Args:
            value: The new hex-encoded hash
            overwrite: If ``None``, requires that the value is the same as before
                       (no operation is performed).
                       If ``False``, this method will always raise an error.
        """
        x = self._get_updated(
            path=self.file_path,
            new=value,
            missing_ok=False,
            overwrite=overwrite,
        )
        return self.new(self.hash_path, file_path=self.file_path, hash_value=x)

    def delete(self) -> None:
        """
        Deletes the hash file by calling ``pathlib.Path.unlink(self.hash_path)``.

        Raises:
            OSError: Accordingly
        """
        self.hash_path.unlink(missing_ok=True)

    def write(self) -> None:
        """
        Writes the hash file.

        Raises:
            OsError: Accordingly
        """
        self.directory.mkdir(exist_ok=True, parents=True)
        self.hash_path.write_text("\n".join(self.lines()), encoding="utf-8")

    @property
    def file_path(self) -> Path:
        if len(self._dct) != 1:
            msg = f"{self.hash_path} contains {len(self._dct)} (!= 1) items"
            raise AssertionError(msg)
        return next(iter(self._dct.keys()))

    @property
    def hash_value(self) -> str:
        if len(self._dct) != 1:
            msg = f"{self.hash_path} contains {len(self._dct)} (!= 1) items"
            raise AssertionError(msg)
        return next(iter(self._dct.values()))

    def verify(self, computed: str) -> None:
        """
        Verifies the checksum.

        Args:
            computed: A pre-computed hex-encoded hash

        Raises:
            HashDidNotValidateError: If the hashes are not equal
        """
        if computed != self.hash_value:
            msg = f"Hash for {self.file_path}: calculated {computed} != expected {self.hash_value}"
            raise HashDidNotValidateError(
                msg,
                actual=computed,
                expected=self.hash_value,
            )


@dataclass(frozen=True, slots=True)
class ChecksumMapping(_ChecksumMapping):
    def load(self, missing_ok: bool = False) -> __qualname__:
        """
        Replaces this map with one read from the hash file.

        Args:
            missing_ok: If the hash path does not exist, treat it has having no items
        """
        return self.__class__.parse(self.hash_path, missing_ok=missing_ok)

    @classmethod
    def parse(
        cls,
        path: Path,
        *,
        lines: Sequence[str] | None = None,
        missing_ok: bool = False,
        subdirs: bool = False,
    ) -> __qualname__:
        """
        Reads hash file contents.

        Args:
            path: The path of the checksum file; required to resolve paths relative to its parent
            lines: The lines in the checksum file; reads ``path`` if None
            missing_ok: If ``path`` does not exist, assume it contains no items
            subdirs: Permit files within subdirectories specified with ``/``
                     Most tools do not support these.

        Returns:
            A mapping from raw string filenames to their hex hashes.
            Any node called ``./`` in the path is stripped.
        """
        return cls._parse(path, lines=lines, missing_ok=missing_ok, subdirs=subdirs)

    @classmethod
    def new(
        cls,
        hash_path: PathLike,
        dct: Mapping[PathLike, str],
    ) -> ChecksumMapping:
        """
        Use this as the constructor.
        """
        hash_path = Path(hash_path)
        return cls(hash_path, {Path(k): v for k, v in dct.items()})

    def write(
        self,
        *,
        sort: bool | Callable[[Sequence[Path]], Sequence[Path]] = False,
        rm_if_empty: bool = False,
    ) -> None:
        """
        Writes to the hash (.shasum-like) file.

        Args:
            sort: Sort with this function, or ``sorted`` if True
            rm_if_empty: Delete with ``pathlib.Path.unlink`` if this contains no items

        Raises:
            OSError: Accordingly
        """
        if sort is True:
            sort = sorted
        if rm_if_empty and len(self._dct) == 0:
            self.hash_path.unlink(missing_ok=True)
        else:
            lines = self.lines()
            if callable(sort):
                lines = sort(lines)
            self.directory.mkdir(exist_ok=True, parents=True)
            self.hash_path.write_text("\n".join(lines), encoding="utf-8")

    @property
    def entries(self) -> Mapping[Path, str]:
        return dict(self._dct)

    def keys(self) -> set[Path]:
        return self._dct.keys()

    def values(self) -> ValuesView[str]:
        return self._dct.values()

    def items(self) -> set[tuple[Path, str]]:
        return self._dct.items()

    def get(self, key: Path, default: str | None = None) -> str | None:
        return self._dct.get(key, default)

    def __contains__(self, path: Path) -> bool:
        return path in self._dct

    def __getitem__(self, path: Path) -> str:
        return self._dct[path]

    def __len__(self) -> int:
        return len(self._dct)

    def __add__(
        self,
        other: ChecksumMapping | Mapping[PathLike, str] | __qualname__,
    ) -> __qualname__:
        """
        Performs a symmetric addition.

        Raises:
            ValueError: If ``other`` intersects (shares keys) with ``self``

        See Also:
            :meth:`append`
        """
        if isinstance(other, ChecksumMapping):
            other = other._dct
        other = {Path(k): v for k, v in other.items()}
        intersection = set(self._dct).intersection(other)
        if len(intersection) > 0:
            msg = f"Cannot merge with intersection: {intersection}"
            raise ValueError(msg)
        return ChecksumMapping(self.hash_path, {**self, **other})

    def __sub__(self, other: PathLike | Iterable[PathLike] | ChecksumMapping) -> __qualname__:
        """
        Removes entries.

        See Also:
            :meth:`remove`
        """
        if isinstance(other, ChecksumMapping):
            other = other._dct
        if isinstance(other, PurePath | str):
            other = {other}
        other = {Path(p) for p in other}
        return self.new(self.hash_path, {k: v for k, v in self.items() if k not in other})

    def remove(
        self,
        remove: PathLike | Iterable[PathLike],
        *,
        missing_ok: bool = False,
    ) -> __qualname__:
        """
        Strips paths from this hash collection.
        Like :meth:`update` but less flexible and only for removing paths.

        Raises:
            :class:`typeddfs.df_errors.PathNotRelativeError`: To avoid, try calling ``resolve`` first
        """
        if isinstance(remove, str | PurePath):
            remove = [remove]
        return self.update({p: None for p in remove}, missing_ok=missing_ok, overwrite=True)

    def append(
        self,
        append: Mapping[PathLike, str],
        *,
        overwrite: bool | None = False,
    ) -> __qualname__:
        """
        Append paths to a dir hash file.
        Like :meth:`update` but less flexible and only for adding paths.
        """
        return self.update(append, missing_ok=True, overwrite=overwrite)

    def update(
        self,
        update: Callable[[Path], PathLike | None] | Mapping[PathLike, PathLike | None],
        *,
        missing_ok: bool = True,
        overwrite: bool | None = True,
    ) -> __qualname__:
        """
        Returns updated hashes from a dir hash file.

        Args:
            update: Values to overwrite.
                    May be a function or a dictionary from paths to values.
                    If ``None`` is returned, the entry will be removed;
                    otherwise, updates with the returned hex hash.
            missing_ok: Require that the path is already listed
            overwrite: Allow overwriting an existing value.
                       If ``None``, only allow if the hash is the same.
        """
        fixed = {}
        # update existing items:
        for p, v in self.items():
            v_new = update(p) if callable(update) else update.get(p, v)
            if v == v_new:
                # avoid an error about overwriting if we're not changing values
                fixed[p] = v
            else:
                fixed[p] = self._get_updated(
                    path=p,
                    new=v_new,
                    missing_ok=missing_ok,
                    overwrite=overwrite,
                )
        # add new items:
        if not callable(update):
            for p, v in update.items():
                p = Path(p)
                fixed[p] = self._get_updated(
                    path=p,
                    new=v,
                    missing_ok=missing_ok,
                    overwrite=overwrite,
                )
        fixed = {k: v for k, v in fixed.items() if v is not None}
        return self.new(self.hash_path, fixed)

    def verify(
        self,
        path: PathLike,
        computed: str,
        *,
        resolve: bool = False,
        exist: bool = False,
    ) -> None:
        """
        Verifies a checksum.
        The file ``path`` must be listed.

        Args:
            path: The file to look for
            computed: A pre-computed hex-encoded hash; if set, do not calculate from ``path``
            resolve: Resolve paths before comparison
            exist: Require that ``path`` exists

        Raises:
            FileNotFoundError: If ``path`` does not exist
            HashFileMissingError: If the hash file does not exist
            HashDidNotValidateError: If the hashes are not equal
            HashVerificationError`: Superclass of ``HashDidNotValidateError`` if
                                    the filename is not listed, etc.
        """
        path = Path(path)
        if resolve:
            path = path.resolve()
        elif not path.is_absolute():
            path = self.directory / path
        if exist and not path.exists():
            msg = f"Path {path} does not exist"
            raise FileNotFoundError(msg)
        found = self.get(path)
        if found is None:
            msg = f"Path {path} not listed in {self.hash_path}"
            raise FileNotFoundError(msg)
        if computed != found:
            msg = f"Hash for {path}: calculated {computed} != expected {found}"
            raise HashDidNotValidateError(
                msg,
                actual=computed,
                expected=found,
            )


__all__ = ["ChecksumFile", "ChecksumMapping"]


1			# SPDX-FileCopyrightText: Copyright 2020-2023, Contributors to typed-dfs
2			# SPDX-PackageHomePage: https://github.com/dmyersturnbull/typed-dfs
3			# SPDX-License-Identifier: Apache-2.0
4			"""
5			Models for shasum-like files.
6			"""
7			from __future__ import annotations
8
9			from dataclasses import dataclass
10			from pathlib import Path, PurePath
11			from typing import TYPE_CHECKING
12
13			import regex
14
15			from typeddfs.df_errors import (
16			HashContradictsExistingError,
17			HashDidNotValidateError,
18			HashExistsError,
19			HashFileMissingError,
20			HashFilenameMissingError,
21			PathNotRelativeError,
22			)
23
24			if TYPE_CHECKING:
25			from collections.abc import Callable, Iterable, Mapping, Sequence, ValuesView
26
27			from typeddfs.utils._utils import PathLike
28
29			_hex_pattern = regex.compile(r"[A-Ha-h0-9]+", flags=regex.V1)
30			_hashsum_file_sep = regex.compile(r" [ *]", flags=regex.V1)
31
32
33			@dataclass(frozen=True, slots=True)
34			class _ChecksumMapping:
35			hash_path: Path
36			_dct: Mapping[Path, str]
37
38			def __post_init__(self):
39			try:
40			for p in self._dct:
41			# will error if it's not
42			p.relative_to(self.directory)
43			except ValueError as e:
44			msg = f"{e}: Full contents are {self._dct}"
45			raise PathNotRelativeError(msg) from e
46
47			def lines(self) -> Sequence[str]:
48			"""
49			Returns the text that would be written for this .shasum-like file.
50			Calls :meth:`unresolve` first.
51			"""
52			unresolved = self.unresolve()
53			return [f"{v} *{p.name}" for p, v in unresolved._dct.items()]
54
55			def line(self, path: PathLike) -> str:
56			"""
57			Returns the text that would be written for a single path in a .shasum-like file.
58			"""
59			path = Path(path)
60			v = self._dct[path]
61			return f"{v} *{path.name}"
62
63			@property
64			def directory(self) -> Path:
65			return self.hash_path.parent
66
67			def resolve(self) -> __qualname__:
68			"""
69			Calls ``pathlib.Path.resolve()`` on all paths.
70			This will follow symlinks, etc.
71			"""
72			# noinspection PyArgumentList
73			return self.__class__(
74			self.hash_path.resolve(),
75			{k.resolve(): v for k, v in self._dct.items()},
76			)
77
78			def unresolve(self) -> __qualname__:
79			"""
80			Each path becomes its filename under ``self.directory``.
81			This means that the parent nodes of each path are discarded in favor of ``self.directory``.
82
83			Raises:
84			ValueError: If two paths are "un-resolved" to the same path
85
86			.. note::
87			This will not work correctly with subdirectories
88			"""
89			dct = {}
90			for k, v in self._dct.items():
91			k = self.directory / k.name
92			if k in dct:
93			msg = f"At least 2 paths resolve to {k}"
94			raise ValueError(msg)
95			dct[k] = v
96			# noinspection PyArgumentList
97			return self.__class__(self.hash_path, dct)
98
99			@classmethod
100			def _parse(
101			cls,
102			path: Path,
103			*,
104			lines: Sequence[str] \| None = None,
105			missing_ok: bool = False,
106			subdirs: bool = False,
107			) -> __qualname__:
108			path = Path(path)
109			if lines is None and path.exists():
110			lines = path.read_text(encoding="utf-8").splitlines()
111			elif missing_ok and lines is None:
112			lines = []
113			elif lines is None:
114			msg = f"Hash file {path} not found"
115			raise HashFileMissingError(msg)
116
117			# ignore spaces -- editors often add an extra line break, and it's probably fine anyway
118			read = [_hashsum_file_sep.split(s, 1) for s in lines if len(s) > 0]
119			# obviously this means that / can't appear in a node
120			# this is consistent with the commonly accepted spec for shasum
121			# does not handle root (beginning with /)
122			kv = {
123			Path(*[n for n in r[1].strip().split("/") if n != "."]): r[0]
124			for r in read
125			if len(r[0]) != 0
126			}
127			if not subdirs:
128			slashed = {k for k in kv if len(k.parts) > 1}
129			if len(slashed) > 0:
130			msg = f"Subdirectory (containing /): {slashed} in {path}"
131			raise ValueError(msg)
132			kv = {Path(path.parent, p): v for p, v in kv.items()}
133			return cls(path, kv)
134
135			def _get_updated(
136			self,
137			*,
138			path: PathLike,
139			new: str \| None,
140			overwrite: bool \| None,
141			missing_ok: bool,
142			) -> str \| None:
143			path = Path(path)
144			z = self._dct.get(path)
145			if z is None and not missing_ok:
146			msg = f"{path} not found ({len(self._dct)} are)"
147			raise HashFilenameMissingError(msg)
148			if z is not None:
149			err = None
150			if overwrite is None and z != new:
151			err = (HashContradictsExistingError, f"Hash for {path} exists but does not match")
152			elif overwrite is False:
153			err = (
154			HashExistsError,
155			f"Hash for {path} exists ({'matches' if z == new else 'differs'})",
156			)
157			if err is not None:
158			raise err[0](err[1], key=str(path), original=z, new=new)
159			return new
160
161
162			@dataclass(frozen=True, slots=True)
163			class ChecksumFile(_ChecksumMapping):
164			def load(self) -> __qualname__:
165			"""
166			Returns a copy of ``self`` read from :attr:`hash_path`.
167			"""
168			return self.__class__.parse(self.hash_path)
169
170			@classmethod
171			def parse(
172			cls,
173			path: Path,
174			*,
175			lines: Sequence[str] \| None = None,
176			) -> __qualname__:
177			"""
178			Reads hash file contents.
179
180			Args:
181			path: The path of the checksum file; required to resolve paths relative to its parent
182			lines: The lines in the checksum file; reads ``path`` if None
183
184			Returns:
185			A ChecksumFile
186			"""
187			return cls._parse(path, lines=lines, missing_ok=False, subdirs=False)
188
189			@classmethod
190			def new(
191			cls,
192			hash_path: PathLike,
193			file_path: PathLike,
194			hash_value: str,
195			) -> ChecksumFile:
196			"""
197			Use this as a constructor.
198			"""
199			hash_path = Path(hash_path)
200			return cls(hash_path, {Path(file_path): hash_value})
201
202			def rename(self, path: Path) -> __qualname__:
203			"""
204			Replaces :attr:`self.file_path` with ``path``.
205			This will affect the filename written in a .shasum-like file.
206			No OS operations are performed.
207			"""
208			return self.new(self.hash_path, file_path=path, hash_value=self.hash_value)
209
210			def update(self, value: str, overwrite: bool \| None = True) -> __qualname__:
211			"""
212			Modifies the hash.
213
214			Args:
215			value: The new hex-encoded hash
216			overwrite: If ``None``, requires that the value is the same as before
217			(no operation is performed).
218			If ``False``, this method will always raise an error.
219			"""
220			x = self._get_updated(
221			path=self.file_path,
222			new=value,
223			missing_ok=False,
224			overwrite=overwrite,
225			)
226			return self.new(self.hash_path, file_path=self.file_path, hash_value=x)
227
228			def delete(self) -> None:
229			"""
230			Deletes the hash file by calling ``pathlib.Path.unlink(self.hash_path)``.
231
232			Raises:
233			OSError: Accordingly
234			"""
235			self.hash_path.unlink(missing_ok=True)
236
237			def write(self) -> None:
238			"""
239			Writes the hash file.
240
241			Raises:
242			OsError: Accordingly
243			"""
244			self.directory.mkdir(exist_ok=True, parents=True)
245			self.hash_path.write_text("\n".join(self.lines()), encoding="utf-8")
246
247			@property
248			def file_path(self) -> Path:
249			if len(self._dct) != 1:
250			msg = f"{self.hash_path} contains {len(self._dct)} (!= 1) items"
251			raise AssertionError(msg)
252			return next(iter(self._dct.keys()))
253
254			@property
255			def hash_value(self) -> str:
256			if len(self._dct) != 1:
257			msg = f"{self.hash_path} contains {len(self._dct)} (!= 1) items"
258			raise AssertionError(msg)
259			return next(iter(self._dct.values()))
260
261			def verify(self, computed: str) -> None:
262			"""
263			Verifies the checksum.
264
265			Args:
266			computed: A pre-computed hex-encoded hash
267
268			Raises:
269			HashDidNotValidateError: If the hashes are not equal
270			"""
271			if computed != self.hash_value:
272			msg = f"Hash for {self.file_path}: calculated {computed} != expected {self.hash_value}"
273			raise HashDidNotValidateError(
274			msg,
275			actual=computed,
276			expected=self.hash_value,
277			)
278
279
280			@dataclass(frozen=True, slots=True)
281			class ChecksumMapping(_ChecksumMapping):
282			def load(self, missing_ok: bool = False) -> __qualname__:
283			"""
284			Replaces this map with one read from the hash file.
285
286			Args:
287			missing_ok: If the hash path does not exist, treat it has having no items
288			"""
289			return self.__class__.parse(self.hash_path, missing_ok=missing_ok)
290
291			@classmethod
292			def parse(
293			cls,
294			path: Path,
295			*,
296			lines: Sequence[str] \| None = None,
297			missing_ok: bool = False,
298			subdirs: bool = False,
299			) -> __qualname__:
300			"""
301			Reads hash file contents.
302
303			Args:
304			path: The path of the checksum file; required to resolve paths relative to its parent
305			lines: The lines in the checksum file; reads ``path`` if None
306			missing_ok: If ``path`` does not exist, assume it contains no items
307			subdirs: Permit files within subdirectories specified with ``/``
308			Most tools do not support these.
309
310			Returns:
311			A mapping from raw string filenames to their hex hashes.
312			Any node called ``./`` in the path is stripped.
313			"""
314			return cls._parse(path, lines=lines, missing_ok=missing_ok, subdirs=subdirs)
315
316			@classmethod
317			def new(
318			cls,
319			hash_path: PathLike,
320			dct: Mapping[PathLike, str],
321			) -> ChecksumMapping:
322			"""
323			Use this as the constructor.
324			"""
325			hash_path = Path(hash_path)
326			return cls(hash_path, {Path(k): v for k, v in dct.items()})
327
328			def write(
329			self,
330			*,
331			sort: bool \| Callable[[Sequence[Path]], Sequence[Path]] = False,
332			rm_if_empty: bool = False,
333			) -> None:
334			"""
335			Writes to the hash (.shasum-like) file.
336
337			Args:
338			sort: Sort with this function, or ``sorted`` if True
339			rm_if_empty: Delete with ``pathlib.Path.unlink`` if this contains no items
340
341			Raises:
342			OSError: Accordingly
343			"""
344			if sort is True:
345			sort = sorted
346			if rm_if_empty and len(self._dct) == 0:
347			self.hash_path.unlink(missing_ok=True)
348			else:
349			lines = self.lines()
350			if callable(sort):
351			lines = sort(lines)
352			self.directory.mkdir(exist_ok=True, parents=True)
353			self.hash_path.write_text("\n".join(lines), encoding="utf-8")
354
355			@property
356			def entries(self) -> Mapping[Path, str]:
357			return dict(self._dct)
358
359			def keys(self) -> set[Path]:
360			return self._dct.keys()
361
362			def values(self) -> ValuesView[str]:
363			return self._dct.values()
364
365			def items(self) -> set[tuple[Path, str]]:
366			return self._dct.items()
367
368			def get(self, key: Path, default: str \| None = None) -> str \| None:
369			return self._dct.get(key, default)
370
371			def __contains__(self, path: Path) -> bool:
372			return path in self._dct
373
374			def __getitem__(self, path: Path) -> str:
375			return self._dct[path]
376
377			def __len__(self) -> int:
378			return len(self._dct)
379
380			def __add__(
381			self,
382			other: ChecksumMapping \| Mapping[PathLike, str] \| __qualname__,
383			) -> __qualname__:
384			"""
385			Performs a symmetric addition.
386
387			Raises:
388			ValueError: If ``other`` intersects (shares keys) with ``self``
389
390			See Also:
391			:meth:`append`
392			"""
393			if isinstance(other, ChecksumMapping):
394			other = other._dct
395			other = {Path(k): v for k, v in other.items()}
396			intersection = set(self._dct).intersection(other)
397			if len(intersection) > 0:
398			msg = f"Cannot merge with intersection: {intersection}"
399			raise ValueError(msg)
400			return ChecksumMapping(self.hash_path, {self, other})
401
402			def __sub__(self, other: PathLike \| Iterable[PathLike] \| ChecksumMapping) -> __qualname__:
403			"""
404			Removes entries.
405
406			See Also:
407			:meth:`remove`
408			"""
409			if isinstance(other, ChecksumMapping):
410			other = other._dct
411			if isinstance(other, PurePath \| str):
412			other = {other}
413			other = {Path(p) for p in other}
414			return self.new(self.hash_path, {k: v for k, v in self.items() if k not in other})
415
416			def remove(
417			self,
418			remove: PathLike \| Iterable[PathLike],
419			*,
420			missing_ok: bool = False,
421			) -> __qualname__:
422			"""
423			Strips paths from this hash collection.
424			Like :meth:`update` but less flexible and only for removing paths.
425
426			Raises:
427			:class:`typeddfs.df_errors.PathNotRelativeError`: To avoid, try calling ``resolve`` first
428			"""
429			if isinstance(remove, str \| PurePath):
430			remove = [remove]
431			return self.update({p: None for p in remove}, missing_ok=missing_ok, overwrite=True)
432
433			def append(
434			self,
435			append: Mapping[PathLike, str],
436			*,
437			overwrite: bool \| None = False,
438			) -> __qualname__:
439			"""
440			Append paths to a dir hash file.
441			Like :meth:`update` but less flexible and only for adding paths.
442			"""
443			return self.update(append, missing_ok=True, overwrite=overwrite)
444
445			def update(
446			self,
447			update: Callable[[Path], PathLike \| None] \| Mapping[PathLike, PathLike \| None],
448			*,
449			missing_ok: bool = True,
450			overwrite: bool \| None = True,
451			) -> __qualname__:
452			"""
453			Returns updated hashes from a dir hash file.
454
455			Args:
456			update: Values to overwrite.
457			May be a function or a dictionary from paths to values.
458			If ``None`` is returned, the entry will be removed;
459			otherwise, updates with the returned hex hash.
460			missing_ok: Require that the path is already listed
461			overwrite: Allow overwriting an existing value.
462			If ``None``, only allow if the hash is the same.
463			"""
464			fixed = {}
465			# update existing items:
466			for p, v in self.items():
467			v_new = update(p) if callable(update) else update.get(p, v)
468			if v == v_new:
469			# avoid an error about overwriting if we're not changing values
470			fixed[p] = v
471			else:
472			fixed[p] = self._get_updated(
473			path=p,
474			new=v_new,
475			missing_ok=missing_ok,
476			overwrite=overwrite,
477			)
478			# add new items:
479			if not callable(update):
480			for p, v in update.items():
481			p = Path(p)
482			fixed[p] = self._get_updated(
483			path=p,
484			new=v,
485			missing_ok=missing_ok,
486			overwrite=overwrite,
487			)
488			fixed = {k: v for k, v in fixed.items() if v is not None}
489			return self.new(self.hash_path, fixed)
490
491			def verify(
492			self,
493			path: PathLike,
494			computed: str,
495			*,
496			resolve: bool = False,
497			exist: bool = False,
498			) -> None:
499			"""
500			Verifies a checksum.
501			The file ``path`` must be listed.
502
503			Args:
504			path: The file to look for
505			computed: A pre-computed hex-encoded hash; if set, do not calculate from ``path``
506			resolve: Resolve paths before comparison
507			exist: Require that ``path`` exists
508
509			Raises:
510			FileNotFoundError: If ``path`` does not exist
511			HashFileMissingError: If the hash file does not exist
512			HashDidNotValidateError: If the hashes are not equal
513			HashVerificationError`: Superclass of ``HashDidNotValidateError`` if
514			the filename is not listed, etc.
515			"""
516			path = Path(path)
517			if resolve:
518			path = path.resolve()
519			elif not path.is_absolute():
520			path = self.directory / path
521			if exist and not path.exists():
522			msg = f"Path {path} does not exist"
523			raise FileNotFoundError(msg)
524			found = self.get(path)
525			if found is None:
526			msg = f"Path {path} not listed in {self.hash_path}"
527			raise FileNotFoundError(msg)
528			if computed != found:
529			msg = f"Hash for {path}: calculated {computed} != expected {found}"
530			raise HashDidNotValidateError(
531			msg,
532			actual=computed,
533			expected=found,
534			)
535
536
537			__all__ = ["ChecksumFile", "ChecksumMapping"]
538

dmyersturnbull / typed-dfs

ChecksumFile.delete() A last analyzed 2023-10-24 08:21 UTC

Complexity

Size

Duplication

Importance

Duplication Side-by-Side

Filter issues like

ChecksumFile.delete() A
last analyzed 2023-10-24 08:21 UTC