Mypyc/cython changes

pyproject.toml

@@ -102,7 +102,12 @@ keywords = [
 codegen = "codegen.cli.cli:main"
 
 [project.optional-dependencies]
-types = ["types-networkx>=3.2.1.20240918", "types-tabulate>=0.9.0.20240106"]
+types = [
+  "types-networkx>=3.2.1.20240918",
+  "types-tabulate>=0.9.0.20240106",
+  "types-requests>=2.32.0.20241016",
+  "types-toml>=0.10.8.20240310",
+]
 [tool.uv]
 cache-keys = [{ git = { commit = true, tags = true } }]
 dev-dependencies = [
@@ -199,6 +204,7 @@ tmp_path_retention_policy = "failed"
 requires = ["hatchling>=1.26.3", "hatch-vcs>=0.4.0", "setuptools-scm>=8.0.0"]
 build-backend = "hatchling.build"
 
+
 [tool.deptry]
 extend_exclude = [".*/eval/test_files/.*.py", ".*conftest.py"]
 pep621_dev_dependency_groups = ["types"]

src/codegen/sdk/codebase/flagging/code_flag.py

@@ -1,14 +1,14 @@
 from dataclasses import dataclass
-from typing import TYPE_CHECKING
+from typing import Generic, TypeVar
 
 from codegen.sdk.codebase.flagging.enums import MessageType
+from codegen.sdk.core.interfaces.editable import Editable
 
-if TYPE_CHECKING:
-    from codegen.sdk.core.interfaces.editable import Editable
+Symbol = TypeVar("Symbol", bound=Editable | None)
 
 
 @dataclass
-class CodeFlag[Symbol: Editable | None]:
+class CodeFlag(Generic[Symbol]):
     symbol: Symbol
     message: str | None = None  # a short desc of the code flag/violation. ex: enums should be ordered alphabetically
     message_type: MessageType = MessageType.GITHUB | MessageType.CODEGEN  # where to send the message (either Github or Slack)

src/codegen/sdk/codebase/flagging/flags.py

@@ -1,21 +1,24 @@
 from dataclasses import dataclass, field
+from typing import TypeVar
 
 from codegen.sdk.codebase.flagging.code_flag import CodeFlag
 from codegen.sdk.codebase.flagging.enums import MessageType
 from codegen.sdk.codebase.flagging.group import Group
 from codegen.sdk.core.interfaces.editable import Editable
 from codegen.shared.decorators.docs import noapidoc
 
+Symbol = TypeVar("Symbol", bound=Editable)
+
 
 @dataclass
 class Flags:
     _flags: list[CodeFlag] = field(default_factory=list)
     _find_mode: bool = False
     _active_group: list[CodeFlag] | None = None
 
-    def flag_instance[Symbol: Editable | None](
+    def flag_instance(
         self,
-        symbol: Symbol = None,
+        symbol: Symbol | None = None,
         message: str | None = None,
         message_type: MessageType = MessageType.GITHUB | MessageType.CODEGEN,
         message_recipient: str | None = None,

src/codegen/sdk/codebase/multigraph.py

@@ -1,15 +1,18 @@
 from collections import defaultdict
 from dataclasses import dataclass, field
+from typing import Generic, TypeVar
 
 from codegen.sdk import TYPE_CHECKING
 from codegen.sdk.core.detached_symbols.function_call import FunctionCall
 
 if TYPE_CHECKING:
     from codegen.sdk.core.function import Function
 
+TFunction = TypeVar("TFunction", bound=Function)
+
 
 @dataclass
-class MultiGraph[TFunction: Function]:
+class MultiGraph(Generic[TFunction]):
     """Mapping of API endpoints to their definitions and usages across languages."""
 
     api_definitions: dict[str, TFunction] = field(default_factory=dict)

src/codegen/sdk/core/file.py

@@ -587,7 +587,7 @@ def invalidate(self):
 
     @classmethod
     @noapidoc
-    def from_content(cls, filepath: str, content: str, G: CodebaseGraph, sync: bool = True, verify_syntax: bool = True) -> Self | None:
+    def from_content(cls, filepath: str | PathLike | Path, content: str, G: CodebaseGraph, sync: bool = True, verify_syntax: bool = True) -> Self | None:
         """Creates a new file from content and adds it to the graph."""
         path = G.to_absolute(filepath)
         ts_node = parse_file(path, content)
@@ -605,7 +605,7 @@ def from_content(cls, filepath: str, content: str, G: CodebaseGraph, sync: bool
             G.add_single_file(path)
             return G.get_file(filepath)
         else:
-            return cls(ts_node, filepath, G)
+            return cls(ts_node, Path(filepath), G)
 
     @classmethod
     @noapidoc

src/codegen/sdk/core/interfaces/editable.py

@@ -22,7 +22,7 @@
 from codegen.shared.decorators.docs import apidoc, noapidoc
 
 if TYPE_CHECKING:
-    from collections.abc import Callable, Generator, Iterable
+    from collections.abc import Callable, Generator, Iterable, Sequence
 
     import rich.repr
     from rich.console import Console, ConsoleOptions, RenderResult
@@ -157,7 +157,7 @@
     def __rich_repr__(self) -> rich.repr.Result:
         yield escape(self.filepath)
 
-    __rich_repr__.angular = ANGULAR_STYLE
+    __rich_repr__.angular = ANGULAR_STYLE  # type: ignore
 
     def __rich_console__(self, console: Console, options: ConsoleOptions) -> RenderResult:
         yield Pretty(self, max_string=MAX_STRING_LENGTH)
@@ -315,14 +315,14 @@
     @property
     @reader
     @noapidoc
-    def children(self) -> list[Editable]:
+    def children(self) -> list[Editable[Self]]:
         """List of Editable instances that are children of this node."""
         return [self._parse_expression(child) for child in self.ts_node.named_children]
 
     @property
     @reader
     @noapidoc
-    def _anonymous_children(self) -> list[Editable]:
+    def _anonymous_children(self) -> list[Editable[Self]]:
         """All anonymous children of an editable."""
         return [self._parse_expression(child) for child in self.ts_node.children if not child.is_named]
 
@@ -343,28 +343,28 @@
     @property
     @reader
     @noapidoc
-    def next_named_sibling(self) -> Editable | None:
+    def next_named_sibling(self) -> Editable[Parent] | None:
         if self.ts_node is None:
             return None
 
         next_named_sibling_node = self.ts_node.next_named_sibling
         if next_named_sibling_node is None:
             return None
 
-        return self._parse_expression(next_named_sibling_node)
+        return self.parent._parse_expression(next_named_sibling_node)
 
     @property
     @reader
     @noapidoc
-    def previous_named_sibling(self) -> Editable | None:
+    def previous_named_sibling(self) -> Editable[Parent] | None:
         if self.ts_node is None:
             return None
 
         previous_named_sibling_node = self.ts_node.prev_named_sibling
         if previous_named_sibling_node is None:
             return None
 
-        return self._parse_expression(previous_named_sibling_node)
+        return self.parent._parse_expression(previous_named_sibling_node)
 
     @property
     def file(self) -> SourceFile:
@@ -377,7 +377,7 @@
         """
         if self._file is None:
             self._file = self.G.get_node(self.file_node_id)
-        return self._file
+        return self._file  # type: ignore
 
     @property
     def filepath(self) -> str:
@@ -391,7 +391,7 @@
         return self.file.file_path
 
     @reader
-    def find_string_literals(self, strings_to_match: list[str], fuzzy_match: bool = False) -> list[Editable]:
+    def find_string_literals(self, strings_to_match: list[str], fuzzy_match: bool = False) -> list[Editable[Self]]:
         """Returns a list of string literals within this node's source that match any of the given
         strings.
 
@@ -400,19 +400,20 @@
             fuzzy_match (bool): If True, matches substrings within string literals. If False, only matches exact strings. Defaults to False.
 
         Returns:
-            list[Editable]: A list of Editable objects representing the matching string literals.
+            list[Editable[Self]]: A list of Editable objects representing the matching string literals.
         """
-        matches = []
+        matches: list[Editable[Self]] = []
         for node in self.extended_nodes:
             matches.extend(node._find_string_literals(strings_to_match, fuzzy_match))
         return matches
 
     @noapidoc
     @reader
-    def _find_string_literals(self, strings_to_match: list[str], fuzzy_match: bool = False) -> list[Editable]:
+    def _find_string_literals(self, strings_to_match: list[str], fuzzy_match: bool = False) -> Sequence[Editable[Self]]:
         all_string_nodes = find_all_descendants(self.ts_node, type_names={"string"})
         editables = []
         for string_node in all_string_nodes:
+            assert string_node.text is not None
             full_string = string_node.text.strip(b'"').strip(b"'")
             if fuzzy_match:
                 if not any([str_to_match.encode("utf-8") in full_string for str_to_match in strings_to_match]):
@@ -461,7 +462,7 @@
         if not is_regex:
             old = re.escape(old)
 
-        for match in re.finditer(old.encode("utf-8"), self.ts_node.text):
+        for match in re.finditer(old.encode("utf-8"), self.ts_node.text):  # type: ignore
             start_byte = self.ts_node.start_byte + match.start()
             end_byte = self.ts_node.start_byte + match.end()
             t = EditTransaction(
@@ -506,7 +507,7 @@
         # Use search to find string
         search_results = itertools.chain.from_iterable(map(self._search, map(re.escape, strings_to_match)))
         if exact:
             search_results = filter(lambda result: result.source in strings_to_match, search_results)
 
         # Combine and deduplicate results
         return list(search_results)
@@ -538,7 +539,7 @@
 
         pattern = re.compile(regex_pattern.encode("utf-8"))
         start_byte_offset = self.ts_node.byte_range[0]
-        for match in pattern.finditer(string):
+        for match in pattern.finditer(string):  # type: ignore
             matching_byte_ranges.append((match.start() + start_byte_offset, match.end() + start_byte_offset))
 
         matches: list[Editable] = []
@@ -738,7 +739,7 @@
         # Delete the node
         t = RemoveTransaction(removed_start_byte, removed_end_byte, self.file, priority=priority, exec_func=exec_func)
         if self.transaction_manager.add_transaction(t, dedupe=dedupe):
-            if exec_func:
+            if exec_func is not None:
                 self.parent._removed_child()
 
         # If there are sibling nodes, delete the surrounding whitespace & formatting (commas)
@@ -873,11 +874,13 @@
                 Editable corresponds to a TreeSitter node instance where the variable
                 is referenced.
         """
-        usages = []
+        usages: Sequence[Editable[Self]] = []
         identifiers = get_all_identifiers(self.ts_node)
         for identifier in identifiers:
             # Excludes function names
             parent = identifier.parent
+            if parent is None:
+                continue
             if parent.type in ["call", "call_expression"]:
                 continue
             # Excludes local import statements
@@ -894,12 +897,12 @@
             if arguments is not None and any(identifier == arg.child_by_field_name("left") for arg in arguments.named_children):
                 continue
 
             usages.append(self._parse_expression(identifier))
 
         return usages
 
     @reader
-    def get_variable_usages(self, var_name: str, fuzzy_match: bool = False) -> list[Editable]:
+    def get_variable_usages(self, var_name: str, fuzzy_match: bool = False) -> Sequence[Editable[Self]]:
         """Returns Editables for all TreeSitter nodes corresponding to instances of variable usage
         that matches the given variable name.
 
@@ -917,6 +920,12 @@
         else:
             return [usage for usage in self.variable_usages if var_name == usage.source]
 
+    @overload
+    def _parse_expression(self, node: TSNode, **kwargs) -> Expression[Self]: ...
+
+    @overload
+    def _parse_expression(self, node: TSNode | None, **kwargs) -> Expression[Self] | None: ...
+
     def _parse_expression(self, node: TSNode | None, **kwargs) -> Expression[Self] | None:
         return self.G.parser.parse_expression(node, self.file_node_id, self.G, self, **kwargs)
 
@@ -945,14 +954,14 @@
 
     @commiter
     @noapidoc
     def _add_symbol_usages(self: HasName, identifiers: list[TSNode], usage_type: UsageKind, dest: HasName | None = None) -> None:
         from codegen.sdk.core.expressions import Name
         from codegen.sdk.core.interfaces.resolvable import Resolvable
 
         if dest is None:
             dest = self
         for x in identifiers:
             if dep := self._parse_expression(x, default=Name):
                 assert isinstance(dep, Resolvable)
                 dep._compute_dependencies(usage_type, dest)
 
@@ -962,7 +971,7 @@
         id_types = self.G.node_classes.resolvables
         # Skip identifiers that are part of a property
         identifiers = find_all_descendants(self.ts_node, id_types, nested=False)
         return self._add_symbol_usages(identifiers, usage_type, dest)
 
     @commiter
     @noapidoc
@@ -971,7 +980,7 @@
         id_types = self.G.node_classes.resolvables
         # Skip identifiers that are part of a property
         identifiers = find_all_descendants(child, id_types, nested=False)
         return self._add_symbol_usages(identifiers, usage_type, dest)
 
     @noapidoc
     def _log_parse(self, msg: str, *args, **kwargs):
@@ -997,7 +1006,7 @@
 
     @cached_property
     @noapidoc
     def github_url(self) -> str | None:
         if self.file.github_url:
             return self.file.github_url + f"#L{self.start_point[0] + 1}-L{self.end_point[0] + 1}"
 
@@ -1058,7 +1067,7 @@
         dest = self
         while dest and not isinstance(dest, Importable):
             dest = dest.parent
         return dest
 
     @cached_property
     @noapidoc

src/codegen/sdk/types.py

@@ -1 +1,3 @@
-type JSON = dict[str, JSON] | list[JSON] | str | int | float | bool | None
+from typing import TypeAlias
+
+JSON: TypeAlias = dict[str, "JSON"] | list["JSON"] | str | int | float | bool | None

src/codegen/sdk/typescript/statements/switch_case.py

@@ -10,7 +10,7 @@
 
 if TYPE_CHECKING:
     from codegen.sdk.codebase.codebase_graph import CodebaseGraph
-    from src.codegen.sdk.typescript.statements.switch_statement import TSSwitchStatement
+    from codegen.sdk.typescript.statements.switch_statement import TSSwitchStatement
 
 
 @ts_apidoc

tests/integration/codemod/conftest.py

@@ -2,6 +2,7 @@
 import shutil
 from collections.abc import Generator
 from pathlib import Path
+from typing import TYPE_CHECKING
 from unittest.mock import MagicMock
 
 import filelock
@@ -13,12 +14,14 @@
 from codegen.git.repo_operator.repo_operator import RepoOperator
 from codegen.sdk.codebase.config import CodebaseConfig, GSFeatureFlags, ProjectConfig
 from codegen.sdk.core.codebase import Codebase
-from codemods.codemod import Codemod
 from tests.shared.codemod.constants import DIFF_FILEPATH
 from tests.shared.codemod.models import BASE_PATH, BASE_TMP_DIR, VERIFIED_CODEMOD_DIFFS, CodemodMetadata, Repo, Size
 from tests.shared.codemod.test_discovery import find_codemod_test_cases, find_repos, find_verified_codemod_cases
 from tests.shared.utils.recursion import set_recursion_limit
 
+if TYPE_CHECKING:
+    from codemods.codemod import Codemod
+
 logger = logging.getLogger(__name__)
 
 ONLY_STORE_CHANGED_DIFFS = True
@@ -201,7 +204,7 @@ def codemod(raw_codemod: type["Codemod"]):
 
 
 @pytest.fixture
-def verified_codemod(codemod_metadata: CodemodMetadata, expected: Path) -> YieldFixture[Codemod]:
+def verified_codemod(codemod_metadata: CodemodMetadata, expected: Path) -> YieldFixture["Codemod"]:
     # write the diff to the file
     diff_path = expected
     diff_path.parent.mkdir(parents=True, exist_ok=True)