feat: Add config option to ignore files based on Unix shell-style wil…

…dcard pattern matching

Also updates `skip_files` to treat file paths as OS-agnostic.

Issue pawamoy#21: pawamoy#21
Issue pawamoy#20: pawamoy#20

.gitignore

@@ -5,6 +5,7 @@
 # python
 *.egg-info/
 *.py[cod]
+pdm.lock
 .venv/
 .venvs/
 /build/

README.md

@@ -59,6 +59,10 @@ plugins:
     - credits.md
     - coverage.md
 
+    # skip files matching Unix shell-style wildcards
+    skip_file_globs:
+    - reference/* 
+
     # whether to only check in strict mode
     strict_only: yes
 ```

docs/known_words.txt

@@ -29,4 +29,7 @@ dataclass
 ahrens
 dev
 fr
-uv
+uv
+codespell
+mkdocs_spellcheck
+symspellpy

src/mkdocs_spellcheck/plugin.py

@@ -5,6 +5,7 @@
 
 from __future__ import annotations
 
+import fnmatch
 from pathlib import Path
 from typing import TYPE_CHECKING, Any
 
@@ -66,6 +67,7 @@ class SpellCheckPlugin(BasePlugin):
         ("backends", MkType(list, default=["symspellpy"])),
         ("known_words", MkType((str, list), default=[])),
         ("skip_files", MkType(list, default=[])),
+        ("skip_file_globs", MkType(list, default=[])),
         ("min_length", MkType(int, default=2)),
         ("max_capital", MkType(int, default=1)),
         ("ignore_code", MkType(bool, default=True)),
@@ -89,7 +91,10 @@ def on_config(self, config: MkDocsConfig) -> MkDocsConfig | None:
         """
         self.strict_only = self.config["strict_only"]
         self.backends_config = self.config["backends"]
-        self.skip_files = self.config["skip_files"]
+        # Convert the skip files into a set of Path objects to make it
+        # OS-agnostic and increase lookup efficiency
+        self.skip_files = {Path(x) for x in self.config["skip_files"]}
+        self.skip_file_globs = self.config["skip_file_globs"]
         self.min_length = self.config["min_length"]
         self.max_capital = self.config["max_capital"]
         self.ignore_code = self.config["ignore_code"]
@@ -129,7 +134,10 @@ def on_page_content(self, html: str, page: Page, **kwargs: Any) -> None:  # noqa
             page: The page instance.
             **kwargs: Additional arguments passed by MkDocs.
         """
-        if self.run and page.file.src_path not in self.skip_files:
+        if self.run and not (
+            Path(page.file.src_path) in self.skip_files
+            or any(fnmatch.fnmatch(page.file.src_path, pattern) for pattern in self.skip_file_globs)
+        ):
             words = get_words(
                 html,
                 known_words=self.known_words,