Check project against valid and deprecated trove classifiers. (#2881)

Inspects trove classifiers on `check` CLI command calls, and look for
unrecognized and deprecated categories.

Adds dependency https://github.com/pypa/trove-classifiers, a package
published and maintained by the PyPA that is cataloguing all classifiers.
This is the canonical source of all trove definitions.

Resolves: #2579

pyproject.toml

@@ -67,6 +67,7 @@ requests-toolbelt = "^0.9.1"
 shellingham = "^1.5"
 shellingham = "^1.5"
 # exclude 0.11.2 and 0.11.3 due to https://github.com/sdispater/tomlkit/issues/225
 # exclude 0.11.2 and 0.11.3 due to https://github.com/sdispater/tomlkit/issues/225
 tomlkit = ">=0.11.1,<1.0.0,!=0.11.2,!=0.11.3"
 tomlkit = ">=0.11.1,<1.0.0,!=0.11.2,!=0.11.3"
+trove-classifiers = "^2022.5.19"
 # exclude 20.4.5 - 20.4.6 due to https://github.com/pypa/pip/issues/9953
 # exclude 20.4.5 - 20.4.6 due to https://github.com/pypa/pip/issues/9953
 virtualenv = "^20.4.3,!=20.4.5,!=20.4.6"
 virtualenv = "^20.4.3,!=20.4.5,!=20.4.6"
 xattr = { version = "^0.9.7", markers = "sys_platform == 'darwin'" }
 xattr = { version = "^0.9.7", markers = "sys_platform == 'darwin'" }

src/poetry/console/commands/check.py

@@ -9,6 +9,52 @@ class CheckCommand(Command):
     name = "check"
     name = "check"
     description = "Checks the validity of the <comment>pyproject.toml</comment> file."
     description = "Checks the validity of the <comment>pyproject.toml</comment> file."
 
 
+    def validate_classifiers(
+        self, project_classifiers: set[str]
+    ) -> tuple[list[str], list[str]]:
+        """Identify unrecognized and deprecated trove classifiers.
+
+        A fully-qualified classifier is a string delimited by `` :: `` separators. To
+        make the error message more readable we need to have visual clues to
+        materialize the start and end of a classifier string. That way the user can
+        easily copy and paste it from the messages while reducing mistakes because of
+        extra spaces.
+
+        We use ``!r`` (``repr()``) for classifiers and list of classifiers for
+        consistency. That way all strings will be rendered with the same kind of quotes
+        (i.e. simple tick: ``'``).
+        """
+        from trove_classifiers import classifiers
+        from trove_classifiers import deprecated_classifiers
+
+        errors = []
+        warnings = []
+
+        unrecognized = sorted(
+            project_classifiers - set(classifiers) - set(deprecated_classifiers)
+        )
+        if unrecognized:
+            errors.append(f"Unrecognized classifiers: {unrecognized!r}.")
+
+        deprecated = sorted(
+            project_classifiers.intersection(set(deprecated_classifiers))
+        )
+        if deprecated:
+            for old_classifier in deprecated:
+                new_classifiers = deprecated_classifiers[old_classifier]
+                if new_classifiers:
+                    message = (
+                        f"Deprecated classifier {old_classifier!r}. "
+                        f"Must be replaced by {new_classifiers!r}."
+                    )
+                else:
+                    message = (
+                        f"Deprecated classifier {old_classifier!r}. Must be removed."
+                    )
+                warnings.append(message)
+
+        return errors, warnings
+
     def handle(self) -> int:
     def handle(self) -> int:
         from poetry.core.pyproject.toml import PyProjectTOML
         from poetry.core.pyproject.toml import PyProjectTOML
 
 
@@ -18,6 +64,13 @@ class CheckCommand(Command):
         poetry_file = Factory.locate(Path.cwd())
         poetry_file = Factory.locate(Path.cwd())
         config = PyProjectTOML(poetry_file).poetry_config
         config = PyProjectTOML(poetry_file).poetry_config
         check_result = Factory.validate(config, strict=True)
         check_result = Factory.validate(config, strict=True)
+
+        # Validate trove classifiers
+        project_classifiers = set(config.get("classifiers", []))
+        errors, warnings = self.validate_classifiers(project_classifiers)
+        check_result["errors"].extend(errors)
+        check_result["warnings"].extend(warnings)
+
         if not check_result["errors"] and not check_result["warnings"]:
         if not check_result["errors"] and not check_result["warnings"]:
             self.info("All set!")
             self.info("All set!")
 
 

tests/console/commands/test_check.py

@@ -41,10 +41,16 @@ def test_check_invalid(mocker: MockerFixture, tester: CommandTester):
 
 
     expected = """\
     expected = """\
 Error: 'description' is a required property
 Error: 'description' is a required property
+Error: Unrecognized classifiers: ['Intended Audience :: Clowns'].
 Warning: A wildcard Python dependency is ambiguous.\
 Warning: A wildcard Python dependency is ambiguous.\
  Consider specifying a more explicit one.
  Consider specifying a more explicit one.
 Warning: The "pendulum" dependency specifies the "allows-prereleases" property,\
 Warning: The "pendulum" dependency specifies the "allows-prereleases" property,\
  which is deprecated. Use "allow-prereleases" instead.
  which is deprecated. Use "allow-prereleases" instead.
+Warning: Deprecated classifier 'Natural Language :: Ukranian'.\
+ Must be replaced by ['Natural Language :: Ukrainian'].
+Warning: Deprecated classifier\
+ 'Topic :: Communications :: Chat :: AOL Instant Messenger'.\
+ Must be removed.
 """
 """
 
 
     assert tester.io.fetch_error() == expected
     assert tester.io.fetch_error() == expected

tests/fixtures/invalid_pyproject/pyproject.toml

@@ -5,6 +5,12 @@ authors = [
     "Foo <foo@bar.com>"
     "Foo <foo@bar.com>"
 ]
 ]
 license = "INVALID"
 license = "INVALID"
+classifiers = [
+    "Environment :: Console",
+    "Intended Audience :: Clowns",
+    "Natural Language :: Ukranian",
+    "Topic :: Communications :: Chat :: AOL Instant Messenger",
+]
 
 
 [tool.poetry.dependencies]
 [tool.poetry.dependencies]
 python = "*"
 python = "*"