path: root/debian/lib/python/debian_linux/debian.py

from __future__ import annotations

import collections
import collections.abc
import dataclasses
import enum
import itertools
import os.path
import re
import typing
import warnings
from typing import (
    Iterable,
    Self,
    TypeAlias,
)


class Changelog(list):
    _top_rules = r"""
^
(?P<source>
    \w[-+0-9a-z.]+
)
[ ]
\(
(?P<version>
    [^\(\)\ \t]+
)
\)
\s+
(?P<distribution>
    [-+0-9a-zA-Z.]+
)
\;\s+urgency=
(?P<urgency>
    \w+
)
(?:,|\n)
"""
    _top_re = re.compile(_top_rules, re.X)
    _bottom_rules = r"""
^
[ ]--[ ]
(?P<maintainer>
    \S(?:[ ]?\S)*
)
[ ]{2}
(?P<date>
    (.*)
)
\n
"""
    _bottom_re = re.compile(_bottom_rules, re.X)
    _ignore_re = re.compile(r'^(?:  |\s*\n)')

    class Entry(object):
        __slot__ = ('distribution', 'source', 'version', 'urgency',
                    'maintainer', 'date')

        def __init__(self, **kwargs):
            for key, value in kwargs.items():
                setattr(self, key, value)

    def __init__(self, dir='', version=None, file=None) -> None:
        if version is None:
            version = Version
        if file:
            self._parse(version, file)
        else:
            with open(os.path.join(dir, "debian/changelog"),
                      encoding="UTF-8") as f:
                self._parse(version, f)

    def _parse(self, version, f) -> None:
        top_match = None
        line_no = 0

        for line in f:
            line_no += 1

            if self._ignore_re.match(line):
                pass
            elif top_match is None:
                top_match = self._top_re.match(line)
                if not top_match:
                    raise Exception('invalid top line %d in changelog' %
                                    line_no)
                try:
                    v = version(top_match.group('version'))
                except Exception:
                    if not len(self):
                        raise
                    v = Version(top_match.group('version'))
            else:
                bottom_match = self._bottom_re.match(line)
                if not bottom_match:
                    raise Exception('invalid bottom line %d in changelog' %
                                    line_no)

                self.append(self.Entry(
                    distribution=top_match.group('distribution'),
                    source=top_match.group('source'),
                    version=v,
                    urgency=top_match.group('urgency'),
                    maintainer=bottom_match.group('maintainer'),
                    date=bottom_match.group('date')))
                top_match = bottom_match = None


class Version(object):
    revision: str | None

    _epoch_re = re.compile(r'\d+$')
    _upstream_re = re.compile(r'[0-9][A-Za-z0-9.+\-:~]*$')
    _revision_re = re.compile(r'[A-Za-z0-9+.~]+$')

    def __init__(self, version) -> None:
        try:
            split = version.index(':')
        except ValueError:
            epoch, rest = None, version
        else:
            epoch, rest = version[0:split], version[split+1:]
        try:
            split = rest.rindex('-')
        except ValueError:
            upstream, revision = rest, None
        else:
            upstream, revision = rest[0:split], rest[split+1:]
        if (epoch is not None and not self._epoch_re.match(epoch)) or \
           not self._upstream_re.match(upstream) or \
           (revision is not None and not self._revision_re.match(revision)):
            raise RuntimeError(u"Invalid debian version")
        self.epoch = epoch and int(epoch)
        self.upstream = upstream
        self.revision = revision

    def __str__(self) -> str:
        return self.complete

    @property
    def complete(self) -> str:
        if self.epoch is not None:
            return u"%d:%s" % (self.epoch, self.complete_noepoch)
        return self.complete_noepoch

    @property
    def complete_noepoch(self) -> str:
        if self.revision is not None:
            return u"%s-%s" % (self.upstream, self.revision)
        return self.upstream

    @property
    def debian(self) -> str | None:
        from warnings import warn
        warn(u"debian argument was replaced by revision", DeprecationWarning,
             stacklevel=2)
        return self.revision


class VersionLinux(Version):
    _upstream_re = re.compile(r"""
(?P<version>
    \d+\.\d+
)
(?P<update>
    (?:\.\d+)?
    (?:-[a-z]+\d+)?
)
(?:
    ~
    (?P<modifier>
        .+?
    )
)?
(?:
    \.dfsg\.
    (?P<dfsg>
        \d+
    )
)?
$
    """, re.X)
    _revision_re = re.compile(r"""
\d+
(\.\d+)?
(?:
    (?P<revision_experimental>
        ~exp\d+
    )
    |
    (?P<revision_security>
        (?:[~+]deb\d+u\d+)+
    )?
    (?P<revision_backports>
        ~bpo\d+\+\d+
    )?
    |
    (?P<revision_other>
        .+?
    )
)
(?:\+b\d+)?
$
    """, re.X)

    def __init__(self, version) -> None:
        super(VersionLinux, self).__init__(version)
        up_match = self._upstream_re.match(self.upstream)
        assert self.revision is not None
        rev_match = self._revision_re.match(self.revision)
        if up_match is None or rev_match is None:
            raise RuntimeError(u"Invalid debian linux version")
        d = up_match.groupdict()
        self.linux_modifier = d['modifier']
        self.linux_version = d['version']
        if d['modifier'] is not None:
            assert not d['update']
            self.linux_upstream = '-'.join((d['version'], d['modifier']))
        else:
            self.linux_upstream = d['version']
        self.linux_upstream_full = self.linux_upstream + d['update']
        self.linux_dfsg = d['dfsg']
        d = rev_match.groupdict()
        self.linux_revision_experimental = d['revision_experimental'] and True
        self.linux_revision_security = d['revision_security'] and True
        self.linux_revision_backports = d['revision_backports'] and True
        self.linux_revision_other = d['revision_other'] and True


class PackageArchitecture(set[str]):
    def __init__(
        self,
        v: str | Iterable[str] | None = None,
        /,
    ) -> None:
        if v:
            if isinstance(v, str):
                v = re.split(r'\s+', v.strip())
            self |= frozenset(v)

    def __str__(self) -> str:
        return ' '.join(sorted(self))


class PackageDescription:
    short: list[str]
    long: list[str]

    def __init__(
        self,
        v: str | Self | None = None,
        /,
    ) -> None:
        self.short = []
        self.long = []

        if v:
            if isinstance(v, str):
                desc_split = v.split('\n', 1)
                self.append_short(desc_split[0])
                if len(desc_split) == 2:
                    self.append(desc_split[1])
            else:
                self.short.extend(v.short)
                self.long.extend(v.long)

    def __str__(self) -> str:
        from .utils import TextWrapper
        wrap = TextWrapper(width=74, fix_sentence_endings=True).wrap
        short = ', '.join(self.short)
        long_pars = []
        for i in self.long:
            long_pars.append(wrap(i))
        long = '\n .\n '.join('\n '.join(i) for i in long_pars)
        return short + '\n ' + long if long else short

    def append(self, long: str) -> None:
        long = long.strip()
        if long:
            self.long.extend(long.split('\n.\n'))

    def append_short(self, short: str) -> None:
        for i in [i.strip() for i in short.split(',')]:
            if i:
                self.short.append(i)

    def extend(self, desc: PackageDescription) -> None:
        self.short.extend(desc.short)
        self.long.extend(desc.long)


class PackageRelationEntryOperator(enum.StrEnum):
    OP_LT = '<<'
    OP_LE = '<='
    OP_EQ = '='
    OP_NE = '!='
    OP_GE = '>='
    OP_GT = '>>'

    def __neg__(self) -> PackageRelationEntryOperator:
        return typing.cast(PackageRelationEntryOperator, {
            self.OP_LT: self.OP_GE,
            self.OP_LE: self.OP_GT,
            self.OP_EQ: self.OP_NE,
            self.OP_NE: self.OP_EQ,
            self.OP_GE: self.OP_LT,
            self.OP_GT: self.OP_LE,
        }[self])


class PackageRelationEntry:
    name: str
    operator: typing.Optional[PackageRelationEntryOperator]
    version: typing.Optional[str]
    arches: PackageArchitecture
    restrictions: PackageBuildprofile

    __re = re.compile(
        r'^(?P<name>\S+)'
        r'(?: \((?P<operator><<|<=|=|!=|>=|>>)\s*(?P<version>[^)]+)\))?'
        r'(?: \[(?P<arches>[^]]+)\])?'
        r'(?P<restrictions>(?: <[^>]+>)*)$'
    )

    def __init__(
        self,
        v: str | Self,
        /, *,
        name: str | None = None,
        arches: set[str] | None = None,
        restrictions: PackageBuildprofile | str | None = None,
    ) -> None:
        if isinstance(v, str):
            match = self.__re.match(v)
            if not match:
                raise RuntimeError('Unable to parse dependency "%s"' % v)

            self.name = name or match['name']

            if operator := match['operator']:
                self.operator = PackageRelationEntryOperator(operator)
            else:
                self.operator = None

            self.version = match['version']
            self.arches = PackageArchitecture(arches or match['arches'])
            if isinstance(restrictions, PackageBuildprofile):
                self.restrictions = restrictions.copy()
            else:
                self.restrictions = PackageBuildprofile.parse(
                    restrictions or match['restrictions'],
                )

        else:
            self.name = name or v.name
            self.operator = v.operator
            self.version = v.version
            self.arches = PackageArchitecture(arches or v.arches)
            if isinstance(restrictions, str):
                self.restrictions = PackageBuildprofile.parse(restrictions)
            else:
                self.restrictions = (restrictions or v.restrictions).copy()

    def __str__(self):
        ret = [self.name]
        if self.operator and self.version:
            ret.append(f'({self.operator} {self.version})')
        if self.arches:
            ret.append(f'[{self.arches}]')
        if self.restrictions:
            ret.append(str(self.restrictions))
        return ' '.join(ret)


class PackageRelationGroup(list[PackageRelationEntry]):
    def __init__(
        self,
        v: Iterable[PackageRelationEntry | str] | str | Self | None = None,
        /, *,
        arches: set[str] | None = None,
    ) -> None:
        if v:
            if isinstance(v, str):
                v = (i.strip() for i in re.split(r'\|', v.strip()))
            self.extend(PackageRelationEntry(i, arches=arches) for i in v if i)

    def __str__(self) -> str:
        return ' | '.join(str(i) for i in self)

    def _merge_eq(self, v: PackageRelationGroup) -> typing.Optional[PackageRelationGroup]:
        if all(
            (
                i.name == j.name and i.operator == j.operator
                and i.version == j.version
            ) for i, j in zip(self, v)
        ):
            return self
        return None


class PackageRelation(list[PackageRelationGroup]):
    Init: TypeAlias = PackageRelationGroup | Iterable[PackageRelationEntry] | str

    def __init__(
        self,
        v: Iterable[Init] | str | Self | None = None,
        /, *,
        arches: set[str] | None = None,
    ) -> None:
        if v:
            if isinstance(v, str):
                v = (i.strip() for i in re.split(r',', v.strip()))
            self.extend(PackageRelationGroup(i, arches=arches) for i in v if i)

    def __str__(self) -> str:
        return ', '.join(str(i) for i in self)

    def _merge_eq(self, v: PackageRelationGroup) -> typing.Optional[PackageRelationGroup]:
        for i in self:
            if i._merge_eq(v):
                return i
        return None

    def merge(
        self,
        v: Init | str,
        /,
    ) -> None:
        v = PackageRelationGroup(v)
        if g := self._merge_eq(v):
            for i, j in zip(g, v):
                i.arches |= j.arches
                i.restrictions.update(j.restrictions)
        else:
            super().append(v)


@dataclasses.dataclass
class PackageBuildprofileEntry:
    pos: set[str] = dataclasses.field(default_factory=set)
    neg: set[str] = dataclasses.field(default_factory=set)

    __re = re.compile(r'^<(?P<profiles>[a-z0-9. !-]+)>$')

    def copy(self) -> Self:
        return self.__class__(
            pos=set(self.pos),
            neg=set(self.neg),
        )

    @classmethod
    def parse(cls, v: str, /) -> Self:
        match = cls.__re.match(v)
        if not match:
            raise RuntimeError('Unable to parse build profile "%s"' % v)

        ret = cls()
        for i in re.split(r' ', match.group('profiles')):
            if i:
                if i[0] == '!':
                    ret.neg.add(i[1:])
                else:
                    ret.pos.add(i)
        return ret

    def __eq__(self, other: object, /) -> bool:
        if not isinstance(other, PackageBuildprofileEntry):
            return NotImplemented
        return self.pos == other.pos and self.neg == other.neg

    def isdisjoint(self, other: Self, /) -> bool:
        return not (self.issubset(other)) and not (self.issuperset(other))

    def issubset(self, other: Self, /) -> bool:
        '''
        Test wether this build profile would select a subset of packages.

        For positive profile matches: Ading profiles will select a subset.
        For negative profile matches: Removing profiles will select a subset.
        '''
        return self.pos >= other.pos and self.neg <= other.neg
    __le__ = issubset

    def issuperset(self, other: Self, /) -> bool:
        '''
        Test wether this build profile would select a superset of packages.

        For positive profile matches: Removing profiles will select a superset.
        For negative profile matches: Adding profiles will select a superset.
        '''
        return self.pos <= other.pos and self.neg >= other.neg
    __ge__ = issuperset

    def update(self, other: Self, /) -> None:
        '''
        Update the build profiles, adding entries from other, merging if possible.

        Negating entries (profile vs !profile) are completely removed.
        All others remain if they are used on both sides.
        '''
        diff = (self.pos & other.neg) | (self.neg & other.pos)
        self.pos &= other.pos - diff
        self.neg &= other.neg - diff
    __ior__ = update

    def __str__(self) -> str:
        s = ' '.join(itertools.chain(
            sorted(self.pos),
            (f'!{i}' for i in sorted(self.neg)),
        ))
        return f'<{s}>'


class PackageBuildprofile(list[PackageBuildprofileEntry]):
    __re = re.compile(r' *(<[^>]+>)(?: +|$)')

    def copy(self) -> Self:
        return self.__class__(i.copy() for i in self)

    @classmethod
    def parse(cls, v: str, /) -> Self:
        ret = cls()
        for match in cls.__re.finditer(v):
            ret.append(PackageBuildprofileEntry.parse(match.group(1)))
        return ret

    def update(self, v: Self, /) -> None:
        for i in v:
            for j in self:
                if not j.isdisjoint(i):
                    j.update(i)
                    break
            else:
                self.append(i)
    __ior__ = update

    def __str__(self) -> str:
        return ' '.join(str(i) for i in self)


class _ControlFileDict(collections.abc.MutableMapping):
    def __init__(self):
        self.__data = {}
        self.meta = {}

    def __getitem__(self, key):
        return self.__data[key]

    def __setitem__(self, key, value):
        if key.lower().startswith('meta-'):
            self.meta[key.lower()[5:]] = value
            return

        try:
            cls = self._fields[key]
            if not isinstance(value, cls):
                if f := getattr(cls, 'parse', None):
                    value = f(value)
                else:
                    value = cls(value)
        except KeyError:
            warnings.warn(
                f'setting unknown field { key } in { type(self).__name__ }',
                stacklevel=2)
        self.__data[key] = value

    def __delitem__(self, key):
        del self.__data[key]

    def __iter__(self):
        keys = set(self.__data.keys())
        for key in self._fields.keys():
            if key in self.__data:
                keys.remove(key)
                yield key
        for key in sorted(keys):
            yield key

    def __len__(self):
        return len(self.__data)

    def setdefault(self, key):
        try:
            return self[key]
        except KeyError:
            try:
                ret = self[key] = self._fields[key]()
            except KeyError:
                warnings.warn(
                    f'setting unknown field { key } in { type(self).__name__ }',
                    stacklevel=2)
                ret = self[key] = ''
            return ret

    def copy(self):
        ret = self.__class__()
        ret.__data = self.__data.copy()
        ret.meta = self.meta.copy()
        return ret

    @classmethod
    def read_rfc822(cls, f):
        entries = []
        eof = False

        while not eof:
            e = cls()
            last = None
            lines = []
            while True:
                line = f.readline()
                if not line:
                    eof = True
                    break
                # Strip comments rather than trying to preserve them
                if line[0] == '#':
                    continue
                line = line.strip('\n')
                if not line:
                    break
                if line[0] in ' \t':
                    if not last:
                        raise ValueError(
                            'Continuation line seen before first header')
                    lines.append(line.lstrip())
                    continue
                if last:
                    e[last] = '\n'.join(lines)
                i = line.find(':')
                if i < 0:
                    raise ValueError(u"Not a header, not a continuation: ``%s''" %
                                     line)
                last = line[:i]
                lines = [line[i + 1:].lstrip()]
            if last:
                e[last] = '\n'.join(lines)
            if e:
                entries.append(e)

        return entries


class SourcePackage(_ControlFileDict):
    _fields = collections.OrderedDict((
        ('Source', str),
        ('Architecture', PackageArchitecture),
        ('Section', str),
        ('Priority', str),
        ('Maintainer', str),
        ('Uploaders', str),
        ('Standards-Version', str),
        ('Build-Depends', PackageRelation),
        ('Build-Depends-Arch', PackageRelation),
        ('Build-Depends-Indep', PackageRelation),
        ('Rules-Requires-Root', str),
        ('Homepage', str),
        ('Vcs-Browser', str),
        ('Vcs-Git', str),
        ('XS-Autobuild', str),
    ))


class BinaryPackage(_ControlFileDict):
    _fields = collections.OrderedDict((
        ('Package', str),
        ('Package-Type', str),  # for udeb only
        ('Architecture', PackageArchitecture),
        ('Section', str),
        ('Priority', str),
        # Build-Depends* fields aren't allowed for binary packages in
        # the real control file, but we move them to the source
        # package
        ('Build-Depends', PackageRelation),
        ('Build-Depends-Arch', PackageRelation),
        ('Build-Depends-Indep', PackageRelation),
        ('Build-Profiles', PackageBuildprofile),
        ('Built-Using', PackageRelation),
        ('Provides', PackageRelation),
        ('Pre-Depends', PackageRelation),
        ('Depends', PackageRelation),
        ('Recommends', PackageRelation),
        ('Suggests', PackageRelation),
        ('Replaces', PackageRelation),
        ('Breaks', PackageRelation),
        ('Conflicts', PackageRelation),
        ('Multi-Arch', str),
        ('Kernel-Version', str),  # for udeb only
        ('Description', PackageDescription),
        ('Homepage', str),
    ))


class TestsControl(_ControlFileDict):
    _fields = collections.OrderedDict((
        ('Tests', str),
        ('Test-Command', str),
        ('Architecture', PackageArchitecture),
        ('Restrictions', str),
        ('Features', str),
        ('Depends', PackageRelation),
        ('Tests-Directory', str),
        ('Classes', str),
    ))