pkg: Support upstream metadata

Fully supported only for Portage, pkgcore can return only remote-ids.

gentoopm/basepm/pkg.py

@@ -1,7 +1,9 @@
 # (c) 2011-2024 Michał Górny <[email protected]>
+# (c) 2024 Anna <[email protected]>
 # SPDX-License-Identifier: GPL-2.0-or-later
 
 import os.path
+import typing
 from abc import abstractmethod, abstractproperty
 
 from ..util import (
@@ -14,6 +16,7 @@
 
 from .atom import PMAtom, PMPackageKey
 from .environ import PMPackageEnvironment
+from gentoopm.basepm.upstream import PMUpstream
 
 PMPackageState = EnumTuple("PMPackageState", "installable", "installed")
 
@@ -417,6 +420,13 @@ def maintainers(self):
         """
         pass
 
+    @property
+    @abstractmethod
+    def upstream(self) -> typing.Optional[PMUpstream]:
+        """
+        Get the package upstream metadata (or ``None`` if unavailable).
+        """
+
     @abstractproperty
     def repo_masked(self):
         """

gentoopm/basepm/upstream.py

@@ -0,0 +1,72 @@
+# (c) 2024 Anna <[email protected]>
+# SPDX-License-Identifier: GPL-2.0-or-later
+
+import typing
+from collections.abc import Iterable
+from dataclasses import dataclass
+from enum import Enum, auto
+
+
+class PMUpstreamMaintainerStatus(Enum):
+    """
+    Maintainer status enumeration.
+    """
+
+    NONE = auto()
+    ACTIVE = auto()
+    INACTIVE = auto()
+
+
+@dataclass(frozen=True, order=True)
+class PMUpstreamMaintainer:
+    """
+    Representation of an upstream maintainer.
+    """
+
+    name: str
+    email: typing.Optional[str] = None
+    status: PMUpstreamMaintainerStatus = PMUpstreamMaintainerStatus.NONE
+
+    def __str__(self) -> str:
+        if self.name and self.email:
+            return f"{self.name} <{self.email}>"
+        return self.name
+
+
+@dataclass(frozen=True)
+class PMUpstreamDoc:
+    """
+    Representation of a documentation link.
+    """
+
+    url: str
+    lang: typing.Optional[str] = None
+
+    def __str__(self) -> str:
+        return self.url
+
+
+@dataclass(frozen=True)
+class PMUpstreamRemoteID:
+    """
+    Representation of a remote-id.
+    """
+
+    name: str
+    site: str
+
+    def __str__(self) -> str:
+        return f"{self.name}: {self.site}"
+
+
+@dataclass
+class PMUpstream:
+    """
+    Representation of upstream metadata.
+    """
+
+    bugs_to: typing.Optional[str] = None
+    changelog: typing.Optional[str] = None
+    docs: typing.Optional[Iterable[PMUpstreamDoc]] = None
+    maintainers: typing.Optional[Iterable[PMUpstreamMaintainer]] = None
+    remote_ids: typing.Optional[Iterable[PMUpstreamRemoteID]] = None

gentoopm/pkgcorepm/pkg.py

@@ -1,6 +1,9 @@
 # (c) 2011-2024 Michał Górny <[email protected]>
+# (c) 2024 Anna <[email protected]>
 # SPDX-License-Identifier: GPL-2.0-or-later
 
+import typing
+
 from ..basepm.pkg import (
     PMPackage,
     PMPackageDescription,
@@ -12,6 +15,11 @@
     PMPackageMaintainer,
 )
 from ..basepm.pkgset import PMPackageSet, PMFilteredPackageSet
+from gentoopm.basepm.upstream import (
+    PMUpstream,
+    PMUpstreamMaintainer,
+    PMUpstreamRemoteID,
+)
 from ..util import SpaceSepTuple, SpaceSepFrozenSet
 
 from .atom import PkgCoreAtom, PkgCorePackageKey
@@ -228,6 +236,13 @@ def restrict(self):
     def maintainers(self):
         return PkgCoreMaintainerTuple(self._pkg.maintainers)
 
+    @property
+    def upstream(self) -> typing.Optional[PMUpstream]:
+        result = PMUpstream()
+        result.remote_ids = tuple(PMUpstreamRemoteID(r.type, r.name)
+                                  for r in self._pkg.upstreams)
+        return result
+
     @property
     def repo_masked(self):
         for m in self._pkg.repo.masked:

gentoopm/portagepm/pkg.py

@@ -1,8 +1,11 @@
 # (c) 2017-2024 Michał Górny <[email protected]>
+# (c) 2024 Anna <[email protected]>
 # SPDX-License-Identifier: GPL-2.0-or-later
 
 import errno
-import os.path
+import typing
+from collections.abc import Iterable
+from pathlib import Path
 
 from portage.versions import cpv_getkey
 from portage.xml.metadata import MetaDataXML
@@ -18,6 +21,12 @@
     PMUseFlag,
     PMPackageMaintainer,
 )
+from gentoopm.basepm.upstream import (
+    PMUpstream,
+    PMUpstreamDoc,
+    PMUpstreamMaintainer,
+    PMUpstreamRemoteID,
+)
 from ..basepm.pkgset import PMPackageSet, PMFilteredPackageSet
 from ..util import SpaceSepTuple, SpaceSepFrozenSet
 
@@ -298,6 +307,18 @@ def __init__(self, cpv, dbapi, tree, repo_prio):
         self._tree = tree
         self._repo_prio = repo_prio
 
+    @property
+    def _metadata(self) -> typing.Optional[MetaDataXML]:
+        # yes, seriously, the only API portage has is direct parser
+        # for the XML file
+        xml_path = Path(self.path).parent / "metadata.xml"
+        try:
+            return MetaDataXML(xml_path, None)
+        except (IOError, OSError) as e:
+            if e.errno == errno.ENOENT:
+                return None
+            raise e
+
     @property
     def path(self):
         return self._dbapi.findname(self._cpv, self._tree)
@@ -307,19 +328,39 @@ def repository(self):
         return self._dbapi.getRepositoryName(self._tree)
 
     @property
-    def maintainers(self):
-        # yes, seriously, the only API portage has is direct parser
-        # for the XML file
-        xml_path = os.path.join(os.path.dirname(self.path), "metadata.xml")
-        try:
-            meta = MetaDataXML(xml_path, None)
-        except (IOError, OSError) as e:
-            if e.errno == errno.ENOENT:
-                return ()
-            raise
-
+    def maintainers(self) -> typing.Optional[Iterable[PortagePackageMaintainer]]:
+        if (meta := self._metadata) is None:
+            return None
         return tuple(PortagePackageMaintainer(m) for m in meta.maintainers())
 
+    @property
+    def upstream(self) -> typing.Optional[PMUpstream]:
+        if (meta := self._metadata) is None:
+            return None
+
+        upstreams = meta.upstream()
+        if len(upstreams) == 0:
+            return None
+        upstream = upstreams[0]
+
+        result = PMUpstream()
+        if len(upstream.bugtrackers) != 0:
+            result.bugs_to = upstream.bugtrackers[0]
+        if len(upstream.changelogs) != 0:
+            result.changelog = upstream.changelogs[0]
+
+        result.docs = tuple(PMUpstreamDoc(url, lang)
+                            for url, lang in upstream.docs) or None
+
+        result.maintainers = tuple(PMUpstreamMaintainer(m.name, m.email)
+                                   for m in upstream.maintainers
+                                   if m.name) or None
+
+        result.remote_ids = tuple(PMUpstreamRemoteID(name, site)
+                                  for site, name in upstream.remoteids
+                                  if name and site) or None
+        return result
+
     @property
     def repo_masked(self):
         raise NotImplementedError(".repo_masked is not implemented for Portage")

gentoopm/tests/test_pkg.py

@@ -1,4 +1,5 @@
 # (c) 2011-2024 Michał Górny <[email protected]>
+# (c) 2024 Anna <[email protected]>
 # SPDX-License-Identifier: GPL-2.0-or-later
 
 import pytest
@@ -125,6 +126,14 @@ def test_no_maintainers(subslotted_pkg):
     assert list(subslotted_pkg.maintainers) == []
 
 
+def test_upstream_remote_id(stack_pkg):
+    assert (upstream := stack_pkg.upstream) is not None
+    assert [(r.name, r.site) for r in upstream.remote_ids] == [
+        ("github", "projg2/gentoopm"),
+        ("pypi", "gentoopm")
+    ]
+
+
 def test_repo_masked(pm):
     pkg = pm.stack.select(PackageNames.pmasked)
     try:

test-root/usr/portage/a/single/metadata.xml

@@ -9,4 +9,8 @@
 		<email>[email protected]</email>
 		<name>Michał Górny</name>
 	</maintainer>
+	<upstream>
+		<remote-id type="github">projg2/gentoopm</remote-id>
+		<remote-id type="pypi">gentoopm</remote-id>
+	</upstream>
 </pkgmetadata>