Merge pull request jxmorris12#126 from mdevolde/types

fix: more explicit MANIFEST.in, added uv.lock and py.typed, fixed type annotations errors and adding mypy to check them

Author: mdevolde

.github/pull_request_template.md

@@ -31,4 +31,4 @@ If applicable, include links to any relevant documentation, articles, or resourc
 - [ ] Updated any relevant tests.
 - [ ] Updated any relevant documentation.
 - [ ] Added comments to your code where necessary.
-- [ ] Formatted your code, run the linters and tests.
+- [ ] Formatted your code, run the linters, checked types and tests.

.github/workflows/test.yml

@@ -72,4 +72,28 @@ jobs:
       - name: Run Ruff Linter
         run: |
           uvx ruff check .
-          uvx ruff format --check .
+          uvx ruff format --check .
+  
+  type_check:
+    name: Type Check with Mypy
+    runs-on: ubuntu-latest
+    steps:
+      - name: Clone repository
+        uses: actions/checkout@v5
+        
+      - name: Install uv
+        uses: astral-sh/setup-uv@v7
+
+      - name: Set up venv
+        run: uv venv .venv
+
+      - name: Install dependencies & build package
+        run: |
+          . .venv/bin/activate
+          uv pip install setuptools wheel build mypy
+          uv build
+          uv pip install dist/*.whl
+
+      - name: Run Mypy Type Checker
+        run: | 
+          uvx mypy

CONTRIBUTING.md

@@ -20,7 +20,7 @@ If you want to contribute, you first need to fork the repo (and preferably creat
 
 To start developing, you can install all the necessary packages in your python environment with this command (optional dependencies will be installed):
 ```shell
-pip install -e .[dev]
+uv sync --group tests --group docs --group types
 ```
 
 When pushing commits, please use the project naming conventions, which are available in [this guide](https://www.conventionalcommits.org/en/v1.0.0/).
@@ -31,10 +31,13 @@ The documentation style used in the project is **ReStructuredText**. Please, if
 Before creating your pull request, when you have made all your commits, you need to run this:
 ```shell
 # Run linters (maybe you will have to fix some issues)
-ruff check language_tool_python tests
+uvx ruff check .
 
 # Format code
-ruff format language_tool_python tests
+uvx ruff format .
+
+# Check types
+uvx mypy
 
 # Tests
 pytest

MANIFEST.in

@@ -1,6 +1,12 @@
-include tests/test_local.bash
-include tests/test_remote.bash
-prune pytest-cache-files-*/
+include pyproject.toml
+include README.md
+include LICENSE
+recursive-include language_tool_python *.py
+include language_tool_python/py.typed
+
+prune .github/
+prune .pytest-cache/
+prune .ruff_cache/
 prune env/
 prune .env
 prune .venv
@@ -9,3 +15,13 @@ prune venv/
 prune ENV/
 prune env.bak/
 prune venv.bak/
+prune docs/
+prune tests/
+
+exclude .gitignore
+exclude build_and_publish.sh
+exclude CONTRIBUTING.md
+exclude extract_long_description.py
+exclude Makefile
+exclude pytest.ini
+exclude uv.lock

language_tool_python/__main__.py

@@ -132,7 +132,7 @@ def __call__(
         namespace: Any,
         values: Any,
         option_string: Optional[str] = None,
-    ):
+    ) -> None:
         """
         This method is called when the action is triggered. It updates the set of rules
         in the namespace with the provided values. The method is invoked automatically

language_tool_python/config_file.py

@@ -5,7 +5,7 @@
 import tempfile
 from dataclasses import dataclass
 from pathlib import Path
-from typing import Any, Callable, Dict, Iterable, Optional, Union
+from typing import Any, Callable, Dict, Iterable, Optional, Tuple, Union
 
 from .exceptions import PathError
 
@@ -22,7 +22,7 @@ class OptionSpec:
     remain constant throughout the application lifecycle.
     """
 
-    py_types: Union[type, tuple[type, ...]]
+    py_types: Union[type, Tuple[type, ...]]
     """The Python type(s) that this option accepts."""
 
     encoder: Callable[[Any], str]
@@ -47,7 +47,7 @@ def _bool_encoder(v: Any) -> str:
     return str(bool(v)).lower()
 
 
-def _comma_list_encoder(v: Any) -> str:
+def _comma_list_encoder(v: Union[str, Iterable[str]]) -> str:
     """
     Encode a value as a comma-separated list string.
 
@@ -57,16 +57,13 @@ def _comma_list_encoder(v: Any) -> str:
     with commas.
 
     :param v: The value to encode. Can be a string or an iterable of values.
-    :type v: Any
+    :type v: Union[str, Iterable[str]]
     :return: A comma-separated string representation of the input value.
     :rtype: str
-    :raises TypeError: If the input is neither a string nor an iterable.
     """
     if isinstance(v, str):
         return v
-    if isinstance(v, Iterable):
-        return ",".join(str(x) for x in v)
-    raise TypeError("expected string or iterable for comma-list option")
+    return ",".join(str(x) for x in v)
 
 
 def _path_encoder(v: Any) -> str:

language_tool_python/download_lt.py

@@ -79,15 +79,15 @@ def parse_java_version(version_text: str) -> Tuple[int, int]:
 
 
 def confirm_java_compatibility(
-    language_tool_version: Optional[str] = LTP_DOWNLOAD_VERSION,
+    language_tool_version: str = LTP_DOWNLOAD_VERSION,
 ) -> None:
     """
     Confirms if the installed Java version is compatible with language-tool-python.
     This function checks if Java is installed and verifies that the major version is at least 8 or 17 (depending on the LanguageTool version).
     It raises an error if Java is not installed or if the version is incompatible.
 
     :param language_tool_version: The version of LanguageTool to check compatibility for.
-    :type language_tool_version: Optional[str]
+    :type language_tool_version: str
     :raises ModuleNotFoundError: If no Java installation is detected.
     :raises SystemError: If the detected Java version is less than the required version.
     """
@@ -228,7 +228,7 @@ def download_zip(url: str, directory: str) -> None:
     logger.info(f"Downloaded {url} to {directory}.")
 
 
-def download_lt(language_tool_version: Optional[str] = LTP_DOWNLOAD_VERSION) -> None:
+def download_lt(language_tool_version: str = LTP_DOWNLOAD_VERSION) -> None:
     """
     Downloads and extracts the specified version of LanguageTool.
     This function checks for Java compatibility, creates the necessary download
@@ -238,7 +238,7 @@ def download_lt(language_tool_version: Optional[str] = LTP_DOWNLOAD_VERSION) ->
     :param language_tool_version: The version of LanguageTool to download. If not
                                   specified, the default version defined by
                                   LTP_DOWNLOAD_VERSION is used.
-    :type language_tool_version: Optional[str]
+    :type language_tool_version: str
     :raises PathError: If the download folder is not a directory.
     :raises ValueError: If the specified version format is invalid.
     """

language_tool_python/language_tag.py

@@ -105,6 +105,9 @@ def _normalize(self, tag: str) -> str:
             return languages[tag.lower().replace("-", "_")]
         except KeyError:
             try:
-                return languages[self._LANGUAGE_RE.match(tag).group(1).lower()]
+                match = self._LANGUAGE_RE.match(tag)
+                if match is None:
+                    raise AttributeError("tag does not match pattern")
+                return languages[match.group(1).lower()]
             except (KeyError, AttributeError) as e:
                 raise ValueError(f"unsupported language: {tag!r}") from e

language_tool_python/match.py

@@ -87,7 +87,7 @@ def four_byte_char_positions(text: str) -> List[int]:
     :return: A list of positions where 4-byte encoded characters are found.
     :rtype: List[int]
     """
-    positions = []
+    positions: List[int] = []
     char_index = 0
     for char in text:
         if len(char.encode("utf-8")) == 4:
@@ -176,10 +176,6 @@ def __init__(self, attrib: Dict[str, Any], text: str) -> None:
         This method adjusts the positions of 4-byte encoded characters in the text
         to ensure the offsets of the matches are correct.
         """
-        if text is None:
-            raise ValueError("The text parameter must not be None")
-        if not isinstance(text, str):
-            raise TypeError("The text parameter must be a string")
 
         # Process rule.
         attrib["category"] = attrib["rule"]["category"]["id"]
@@ -204,7 +200,10 @@ def __init__(self, attrib: Dict[str, Any], text: str) -> None:
             Match.FOUR_BYTES_POSITIONS = four_byte_char_positions(text)
         # Get the positions of 4-byte encoded characters in the text because without
         # carrying out this step, the offsets of the matches could be incorrect.
-        self.offset -= sum(1 for pos in Match.FOUR_BYTES_POSITIONS if pos < self.offset)
+        if Match.FOUR_BYTES_POSITIONS is not None:
+            self.offset -= sum(
+                1 for pos in Match.FOUR_BYTES_POSITIONS if pos < self.offset
+            )
 
     def __repr__(self) -> str:
         """