Fix pycroscope findings

taxonomy/applicator/generic.py

@@ -1317,6 +1317,10 @@ def build_plan(
                                 f"change {index}: field {field_name!r} is not an "
                                 "ADT tag field"
                             )
+                        # Work around https://github.com/JelleZijlstra/pycroscope/issues/520.
+                        field = cast(  # type: ignore[redundant-cast]
+                            ADTField[Any], field
+                        )
                         if field_key in planned_values:
                             raise RecommendationError(
                                 f"change {index}: remove_raw must be the first change "

taxonomy/applicator/location.py

@@ -1481,7 +1481,7 @@ def add_virtual_models(plan: RecommendationPlan, builder: ProposalBuilder) -> No
                         change.field, change.new_value, expected_label=change.new_label
                     ),
                 )
-            tags = tuple(proposal.tags or ())
+            tags = tuple(proposal.tags)
             tags = tuple(tag for tag in tags if tag not in row.remove_tags)
             tags = (*tags, *(tag for tag in row.add_tags if tag not in tags))
             proposal.tags = tuple(sorted(set(tags)))  # type: ignore[assignment]

taxonomy/applicator/taxon.py

@@ -1,9 +1,10 @@
 """Atomic Taxon/base-Name recommendations for manifest-created classifications."""
 
+import enum
 from collections import Counter
 from collections.abc import Callable, Iterable, Mapping
 from dataclasses import dataclass
-from typing import Any
+from typing import Any, TypeVar
 
 from taxonomy.applicator import generic
 from taxonomy.applicator.proposals import ProposalBuilder
@@ -72,7 +73,10 @@ def _string(data: Mapping[str, Any], key: str, line: int) -> str:
     return value
 
 
-def _enum(data: Mapping[str, Any], key: str, cls: type[Any], line: int) -> Any:
+EnumT = TypeVar("EnumT", bound=enum.Enum)
+
+
+def _enum(data: Mapping[str, Any], key: str, cls: type[EnumT], line: int) -> EnumT:
     value = _string(data, key, line)
     try:
         return cls[value]

taxonomy/applicator/test_article.py

@@ -1,6 +1,7 @@
 import copy
 import hashlib
 from collections.abc import Mapping
+from dataclasses import dataclass
 from pathlib import Path
 from types import SimpleNamespace
 from typing import Any, cast
@@ -14,6 +15,12 @@
 from taxonomy.db.models.person import VirtualPerson
 
 
+@dataclass(frozen=True)
+class _Options:
+    new_path: Path
+    library_path: Path
+
+
 def _pdf_bytes() -> bytes:
     return b"%PDF-1.4\nminimal test fixture\n%%EOF\n"
 
@@ -138,7 +145,7 @@ def _build_volume_and_chapter(tmp_path: Path) -> recommendations.RecommendationP
     )
     return recommendations.build_plan(
         rows,
-        options=SimpleNamespace(new_path=new_path, library_path=library_path),
+        options=_Options(new_path=new_path, library_path=library_path),
         get_article=lambda _name: None,
         articles_with_doi=lambda _doi: (),
         is_catalog_folder=lambda _path: True,
@@ -157,7 +164,7 @@ def _build(tmp_path: Path) -> recommendations.RecommendationPlan:
     row = recommendations.parse_recommendation(_row(pdf), 1)
     return recommendations.build_plan(
         (row,),
-        options=SimpleNamespace(new_path=new_path, library_path=library_path),
+        options=_Options(new_path=new_path, library_path=library_path),
         get_article=lambda _name: None,
         articles_with_doi=lambda _doi: (),
         is_catalog_folder=lambda _path: True,
@@ -216,7 +223,7 @@ def test_build_plan_canonicalizes_tags_and_infers_jstor(tmp_path: Path) -> None:
 
     plan = recommendations.build_plan(
         (row,),
-        options=SimpleNamespace(new_path=new_path, library_path=library_path),
+        options=_Options(new_path=new_path, library_path=library_path),
         get_article=lambda _name: None,
         articles_with_doi=lambda _doi: (),
         is_catalog_folder=lambda _path: True,
@@ -280,7 +287,7 @@ def test_existing_person_author_is_guarded_and_reused_virtually(tmp_path: Path)
     )
     plan = recommendations.build_plan(
         (recommendations.parse_recommendation(row_data, 1),),
-        options=SimpleNamespace(new_path=new_path, library_path=library_path),
+        options=_Options(new_path=new_path, library_path=library_path),
         get_article=lambda _name: None,
         get_person_by_id=lambda _id: person,
         articles_with_doi=lambda _doi: (),
@@ -328,7 +335,7 @@ def test_existing_person_author_rejects_changed_family_name(tmp_path: Path) -> N
     ):
         recommendations.build_plan(
             (recommendations.parse_recommendation(row_data, 1),),
-            options=SimpleNamespace(new_path=new_path, library_path=library_path),
+            options=_Options(new_path=new_path, library_path=library_path),
             get_article=lambda _name: None,
             get_person_by_id=lambda _id: person,
             articles_with_doi=lambda _doi: (),
@@ -350,7 +357,7 @@ def test_build_plan_rejects_changed_staged_file(tmp_path: Path) -> None:
     with pytest.raises(recommendations.RecommendationError, match="size changed"):
         recommendations.build_plan(
             (row,),
-            options=SimpleNamespace(new_path=new_path, library_path=library_path),
+            options=_Options(new_path=new_path, library_path=library_path),
             get_article=lambda _name: None,
             articles_with_doi=lambda _doi: (),
             is_catalog_folder=lambda _path: True,
@@ -424,7 +431,7 @@ def test_inline_citation_group_is_planned_and_created(tmp_path: Path) -> None:
             recommendations.parse_recommendation(row_data, 1),
             recommendations.parse_recommendation(second_row_data, 2),
         ),
-        options=SimpleNamespace(new_path=new_path, library_path=library_path),
+        options=_Options(new_path=new_path, library_path=library_path),
         get_article=lambda _name: None,
         articles_with_doi=lambda _doi: (),
         is_catalog_folder=lambda _path: True,
@@ -494,7 +501,7 @@ def test_completed_exact_state_is_idempotent(tmp_path: Path) -> None:
 
     plan = recommendations.build_plan(
         (recommendations.parse_recommendation(_row(pdf), 1),),
-        options=SimpleNamespace(new_path=new_path, library_path=library_path),
+        options=_Options(new_path=new_path, library_path=library_path),
         get_article=lambda _name: existing,
         articles_with_doi=lambda _doi: (existing,),
         is_catalog_folder=lambda _path: True,
@@ -578,7 +585,7 @@ def test_build_plan_rejects_forward_planned_parent_reference(tmp_path: Path) ->
     with pytest.raises(recommendations.RecommendationError, match="must be an earlier"):
         recommendations.build_plan(
             rows,
-            options=SimpleNamespace(new_path=new_path, library_path=library_path),
+            options=_Options(new_path=new_path, library_path=library_path),
             get_article=lambda _name: None,
             articles_with_doi=lambda _doi: (),
             is_catalog_folder=lambda _path: True,
@@ -610,7 +617,7 @@ def test_build_plan_orders_forward_typed_parent_reference(tmp_path: Path) -> Non
 
     plan = recommendations.build_plan(
         rows,
-        options=SimpleNamespace(new_path=new_path, library_path=library_path),
+        options=_Options(new_path=new_path, library_path=library_path),
         get_article=lambda _name: None,
         articles_with_doi=lambda _doi: (),
         is_catalog_folder=lambda _path: True,

taxonomy/applicator/test_generic.py

@@ -74,6 +74,7 @@ def create_object(
         model: type[BaseModel], values: Mapping[str, object]
     ) -> BaseModel:
         obj = model.virtual(**values)
+        assert isinstance(obj, Location)
         created.append(obj)
         return obj
 
@@ -118,6 +119,7 @@ def create_object(
         model: type[BaseModel], values: Mapping[str, object]
     ) -> BaseModel:
         obj = model.virtual(**values)
+        assert isinstance(obj, Location)
         created.append(obj)
         return obj
 

taxonomy/db/models/base.py

@@ -1530,13 +1530,17 @@ def serialize(self, value: Sequence[ADTT]) -> str | None:
     def prepare_virtual(
         self, value: Sequence[ADTT]
     ) -> tuple[str | None, Sequence[ADTT]]:
-        raw_value = self.serialize(value)
+        # Work around https://github.com/JelleZijlstra/pycroscope/issues/521.
+        raw_value = self.serialize(value)  # static analysis: ignore[incompatible_call]
         if isinstance(value, str):
             return raw_value, self.deserialize(raw_value)
         return raw_value, tuple(value or ())
 
     def validate_persistent(self, value: Sequence[ADTT]) -> None:
-        super().validate_persistent(value)
+        # Work around https://github.com/JelleZijlstra/pycroscope/issues/521.
+        super().validate_persistent(
+            value  # static analysis: ignore[incompatible_argument]
+        )
 
         def iter_models(item: Any) -> Iterable[Model]:
             if isinstance(item, Model):

taxonomy/db/models/classification_entry/lint.py

@@ -94,13 +94,13 @@ def check_tags(ce: ClassificationEntry, cfg: LintConfig) -> Iterable[LintResult]
         yield "unnecessary CorrectedName tag"
     if counts[ClassificationEntryTag.ReferencedUsage] > 1:
         yield "multiple ReferencedUsage tags"
-    if counts[ClassificationEntryTag.AuxiliaryName] > 1:
+    if counts[type(ClassificationEntryTag.AuxiliaryName)] > 1:
         yield "multiple AuxiliaryName tags"
     if counts[ClassificationEntryTag.VerbatimParent] > 1:
         yield "multiple VerbatimParent tags"
     if counts[ClassificationEntryTag.Materialize] > 1:
         yield "multiple Materialize tags"
-    if counts[ClassificationEntryTag.OriginalCitation] > 1:
+    if counts[type(ClassificationEntryTag.OriginalCitation)] > 1:
         yield "multiple OriginalCitation tags"
     base_name_author_tags = list(
         ce.get_tags(ce.tags, ClassificationEntryTag.MaterializeBaseNameAuthor)
@@ -2146,6 +2146,7 @@ def check_page(ce: ClassificationEntry, cfg: LintConfig) -> Iterable[LintResult]
         new_page = yield from models.name.page.check_page(
             ce.page, get_raw_page_regex=ce.article.get_raw_page_regex
         )
+        assert new_page is not None
         if new_page != ce.page:
             fixes = [field_fix(ce, "page", new_page)]
             for tag in ce.get_tags(ce.tags, ClassificationEntryTag.PageLink):

taxonomy/db/models/classification_entry/test_lint.py

@@ -765,6 +765,21 @@ def test_mapped_entry_removes_orphaned_base_name_instructions() -> None:
     assert ce.tags == (retained,)
 
 
+def test_check_tags_counts_parameterless_tags() -> None:
+    ce = _make_ce(parent=None, auxiliary=False)
+
+    for tag, expected in (
+        (ClassificationEntryTag.AuxiliaryName, "multiple AuxiliaryName tags"),
+        (ClassificationEntryTag.OriginalCitation, "multiple OriginalCitation tags"),
+    ):
+        ce.tags = (tag, tag)  # type: ignore[assignment]
+
+        assert any(
+            expected in str(issue)
+            for issue in check_tags(ce, LintConfig(autofix=False, interactive=False))
+        )
+
+
 def test_etymology_detail_is_transferred_to_mapped_name() -> None:
     taxon = Taxon.virtual(valid_name="Endodonta", rank=Rank.genus, age=AgeClass.extant)
     name = Name.virtual(

taxonomy/db/models/name/lint.py

@@ -4092,6 +4092,8 @@ def check_required_fields(nam: Name, cfg: LintConfig) -> Iterable[LintResult]:
     if nam.verbatim_citation and not nam.citation_group:
         yield "has verbatim citation but no citation group"
     if has_accessible_original_citation(nam):
+        citation = nam.original_citation
+        assert citation is not None
         if (
             nam.page_described is None
             and not (
@@ -4108,7 +4110,6 @@ def check_required_fields(nam: Name, cfg: LintConfig) -> Iterable[LintResult]:
             yield "has original citation but no original_rank"
         if nam.author_tags is None:
             message = "has original citation but no author_tags"
-            citation = nam.original_citation
             if citation.issupplement() and citation.parent is not None:
                 authors = citation.parent.author_tags
             else:
@@ -4119,7 +4120,7 @@ def check_required_fields(nam: Name, cfg: LintConfig) -> Iterable[LintResult]:
                 yield field_issue(message, nam, "author_tags", authors)
         if nam.year is None:
             message = "has original citation but no year"
-            yield field_issue(message, nam, "year", nam.original_citation.year)
+            yield field_issue(message, nam, "year", citation.year)
         if (
             nam.name_complex is None
             and nam.group is Group.genus
@@ -5508,7 +5509,6 @@ def check_structured_verbatim_citation_fields(
                         yield replace_tag_issue(
                             msg, nam, tag, new_tag, field="type_tags"
                         )
-                        tag = new_tag
                         end_page = candidate
                 except ValueError:
                     pass

taxonomy/db/models/name/test_lint.py

@@ -8,6 +8,7 @@
 from taxonomy.db import coordinate_lint, models
 from taxonomy.db.constants import (
     AgeClass,
+    ArticleType,
     Group,
     NamingConvention,
     NomenclatureStatus,
@@ -20,6 +21,7 @@
 )
 from taxonomy.db.models.base import LintConfig
 from taxonomy.db.models.location import Location
+from taxonomy.db.models.person import AuthorTag
 
 from .lint import (
     _create_name_variant_issue,
@@ -62,8 +64,13 @@ def test_redirect_name_issue_changes_only_redirect_fields() -> None:
 
 
 def test_take_over_name_issue_uses_explicit_fields_and_tag_removal() -> None:
-    citation = SimpleNamespace(
-        parent=None, author_tags=("Author",), year="1900", issupplement=lambda: False
+    author = AuthorTag.Author(person=models.Person.virtual(family_name="Author"))
+    citation = models.Article.virtual(
+        name="citation",
+        parent=None,
+        author_tags=(author,),
+        year="1900",
+        type=ArticleType.JOURNAL,
     )
     ce = SimpleNamespace(article=citation, page="12", name="Original name")
     page_link = TypeTag.AuthorityPageLink(
@@ -91,7 +98,7 @@ def test_take_over_name_issue_uses_explicit_fields_and_tag_removal() -> None:
     assert name.original_citation is citation
     assert name.page_described == "12"
     assert name.original_name == "Original name"
-    assert name.author_tags == ("Author",)
+    assert name.author_tags == (author,)
     assert name.year == "1900"
     assert name.type_tags == ()
 

taxonomy/db/models/test_base.py

@@ -1,5 +1,5 @@
 import sqlite3
-from collections.abc import Iterator
+from collections.abc import Generator
 from contextlib import contextmanager
 from types import SimpleNamespace
 
@@ -59,7 +59,7 @@ def test_lint_all_uses_read_only_context_without_autofix(
     events: list[str] = []
 
     @contextmanager
-    def readonly() -> Iterator[None]:
+    def readonly() -> Generator[None]:
         events.append("enter")
         try:
             yield

taxonomy/db/models/test_lint_registry.py

@@ -412,7 +412,7 @@ def test_duplicate_finder_can_return_structured_fix() -> None:
     def duplicate_key(_obj: FakeObject) -> str:
         return "same"
 
-    issues = list(duplicate_key.linter(second, LintConfig(autofix=False)))
+    issues = list(duplicate_key.linter(cast(Any, second), LintConfig(autofix=False)))
 
     assert len(issues) == 1
     assert isinstance(issues[0], LintIssue)
@@ -475,7 +475,7 @@ def test_legacy_autofix_branches_are_explicitly_allowlisted() -> None:
                 parent, (ast.FunctionDef, ast.AsyncFunctionDef)
             ):
                 parent = parents.get(parent)
-            assert parent is not None
+            assert isinstance(parent, (ast.FunctionDef, ast.AsyncFunctionDef))
             branch = (str(path.relative_to(models_dir)), parent.name)
             if branch in {
                 ("base.py", "_process_lint_results"),

taxonomy/db/nomenclature_book.py

@@ -394,7 +394,7 @@ def get_type_specimen_text(name: Name) -> tuple[str, list[str]]:
 
 
 def get_row(taxon: Taxon, name: Name, taxon_to_ces: TaxonToCEs) -> Row:
-    todos = []
+    todos: list[str] = []
     order = taxon.get_derived_field("order")
     family = taxon.get_derived_field("family")
     interpreted_tl = name.get_type_tag(TypeTag.InterpretedTypeLocality)
@@ -411,7 +411,10 @@ def get_row(taxon: Taxon, name: Name, taxon_to_ces: TaxonToCEs) -> Row:
             todos.append(
                 f"Base name is not valid (status: {name.nomenclature_status.name})"
             )
-        todos += check_full_expected_base_name.linter(taxon, LintConfig())
+        todos.extend(
+            str(issue)
+            for issue in check_full_expected_base_name.linter(taxon, LintConfig())
+        )
 
     nomenclature_text = ""
     if nomen_novum_for := name.get_tag_target(NameTag.NomenNovumFor):