omdev 0.0.0.dev348__tar.gz → 0.0.0.dev350__tar.gz

{omdev-0.0.0.dev348/omdev.egg-info → omdev-0.0.0.dev350}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: omdev
-Version: 0.0.0.dev348
+Version: 0.0.0.dev350
 Summary: omdev
 Author: wrmsr
 License: BSD-3-Clause
@@ -12,7 +12,7 @@ Classifier: Operating System :: OS Independent
 Classifier: Operating System :: POSIX
 Requires-Python: >=3.12
 License-File: LICENSE
-Requires-Dist: omlish==0.0.0.dev348
+Requires-Dist: omlish==0.0.0.dev350
 Provides-Extra: all
 Requires-Dist: black~=25.1; extra == "all"
 Requires-Dist: pycparser~=2.22; extra == "all"

{omdev-0.0.0.dev348 → omdev-0.0.0.dev350}/omdev/.manifests.json

@@ -306,7 +306,7 @@
     "module": ".tools.cloc",
     "attr": "_CLI_MODULE",
     "file": "omdev/tools/cloc.py",
-    "line": 202,
+    "line": 248,
     "value": {
       "$.cli.types.CliModule": {
         "cmd_name": "cloc",

{omdev-0.0.0.dev348 → omdev-0.0.0.dev350}/omdev/tools/cloc.py

@@ -8,46 +8,83 @@ import os
 import re
 import typing as ta
 
+from omlish import collections as col
+
 from ..cli import CliModule
 
 
 ##
 
 
-SUPPORTED_EXTENSIONS: ta.Mapping[str, str] = {
-    '.c': 'c',
-    '.cpp': 'c',
-    '.h': 'c',
-    '.hpp': 'c',
-    '.py': 'python',
-    '.js': 'javascript',
-}
+@dc.dataclass(frozen=True)
+class Lang:
+    name: str
+    extensions: ta.AbstractSet[str]
+
+    _: dc.KW_ONLY
+
+    comment_pats: tuple[str, str | None, str | None]
+    is_test_file_fn: ta.Callable[[str], bool] | None = None
+
+
+_LANG = [
+    Lang(
+        'python',
+        {'.py'},
+        comment_pats=(r'#', None, None),
+        is_test_file_fn=lambda fp: 'tests' in os.path.normpath(fp).split(os.sep),
+    ),
 
+    Lang(
+        'c',
+        {'.c', '.cc', '.cpp', '.h', '.hh', '.hpp'},
+        comment_pats=(r'//', r'/\*', r'\*/'),
+    ),
 
-COMMENT_PATTERNS: ta.Mapping[str, tuple[str, str | None, str | None]] = {
-    'c': (r'//', r'/\*', r'\*/'),
-    'python': (r'#', None, None),
-    'javascript': (r'//', r'/\*', r'\*/'),
-}
+    Lang(
+        'javascript',
+        {'.js'},
+        comment_pats=(r'//', r'/\*', r'\*/'),
+    ),
+]
+
+
+LANGS = col.make_map_by(lambda l: l.name, _LANG, strict=True)
+LANGS_BY_EXTENSION = col.make_map([(e, l) for l in _LANG for e in l.extensions], strict=True)
+
+
+##
 
 
 @dc.dataclass(frozen=True)
-class FileLineCount:
+class FileStats:
     loc: int
     blanks: int
     comments: int
 
 
-def count_lines(file_path: str, language: str) -> FileLineCount:
-    single_line_comment, block_comment_start, block_comment_end = COMMENT_PATTERNS[language]
+@dc.dataclass(frozen=True)
+class File:
+    path: str
+    lang: Lang
+    stats: FileStats
+    is_test: bool
+
+
+def examine_file(path: str, lang: Lang | None = None) -> File | None:
+    if lang is None:
+        ext = os.path.splitext(os.path.basename(path))[1]
+        if (lang := LANGS_BY_EXTENSION.get(ext)) is None:
+            return None
 
+    single_line_comment, block_comment_start, block_comment_end = lang.comment_pats
     in_block_comment = False
 
     loc = 0
     blank_lines = 0
     comment_lines = 0
 
-    with open(file_path, encoding='utf-8') as file:
+    with open(path, encoding='utf-8') as file:
         for line in file:
             stripped = line.strip()
             if not stripped:
@@ -72,26 +109,30 @@ def count_lines(file_path: str, language: str) -> FileLineCount:
 
             loc += 1
 
-    return FileLineCount(
-        loc=loc,
-        blanks=blank_lines,
-        comments=comment_lines,
+    return File(
+        path=path,
+        lang=lang,
+        stats=FileStats(
+            loc=loc,
+            blanks=blank_lines,
+            comments=comment_lines,
+        ),
+        is_test=lang.is_test_file_fn(path) if lang.is_test_file_fn is not None else False,
     )
 
 
+##
+
+
 def count_lines_in_directory(
         directory: str,
         *,
         include: list[re.Pattern[str]] | None = None,
         exclude: list[re.Pattern[str]] | None = None,
-) -> ta.Mapping[str, FileLineCount]:
-    results: dict[str, FileLineCount] = {}
+) -> ta.Mapping[str, FileStats]:
+    results: dict[str, FileStats] = {}
     for root, _, files in os.walk(directory):
         for file in files:
-            ext = os.path.splitext(file)[1]
-            if ext not in SUPPORTED_EXTENSIONS:
-                continue
-
             file_path = os.path.join(root, file)
 
             if include and not any(p.fullmatch(file_path) for p in include):
@@ -99,13 +140,15 @@ def count_lines_in_directory(
             if exclude and any(p.fullmatch(file_path) for p in exclude):
                 continue
 
-            language = SUPPORTED_EXTENSIONS[ext]
-            results[file_path] = count_lines(file_path, language)
+            if (ex := examine_file(file_path)) is None:
+                continue
+
+            results[file_path] = ex.stats
 
     return results
 
 
-def display_results(results: ta.Mapping[str, FileLineCount]) -> None:
+def display_results(results: ta.Mapping[str, FileStats]) -> None:
     total_loc = total_blanks = total_comments = 0
     file_width = max(map(len, results))
     dash_width = 41 + file_width
@@ -150,6 +193,9 @@ def display_results(results: ta.Mapping[str, FileLineCount]) -> None:
     )
 
 
+##
+
+
 def _main() -> None:
     import argparse
 
@@ -174,7 +220,7 @@ def _main() -> None:
 
     #
 
-    results_by_directory: dict[str, FileLineCount] = {}
+    results_by_directory: dict[str, FileStats] = {}
     for directory in args.directory:
         results = count_lines_in_directory(
             directory,
@@ -188,7 +234,7 @@ def _main() -> None:
         display_results(results)
         print()
 
-        results_by_directory[directory] = FileLineCount(
+        results_by_directory[directory] = FileStats(
             loc=sum(flc.loc for flc in results.values()),
             blanks=sum(flc.blanks for flc in results.values()),
             comments=sum(flc.comments for flc in results.values()),

{omdev-0.0.0.dev348 → omdev-0.0.0.dev350/omdev.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: omdev
-Version: 0.0.0.dev348
+Version: 0.0.0.dev350
 Summary: omdev
 Author: wrmsr
 License: BSD-3-Clause
@@ -12,7 +12,7 @@ Classifier: Operating System :: OS Independent
 Classifier: Operating System :: POSIX
 Requires-Python: >=3.12
 License-File: LICENSE
-Requires-Dist: omlish==0.0.0.dev348
+Requires-Dist: omlish==0.0.0.dev350
 Provides-Extra: all
 Requires-Dist: black~=25.1; extra == "all"
 Requires-Dist: pycparser~=2.22; extra == "all"

{omdev-0.0.0.dev348 → omdev-0.0.0.dev350}/omdev.egg-info/requires.txt

@@ -1,4 +1,4 @@
-omlish==0.0.0.dev348
+omlish==0.0.0.dev350
 
 [all]
 black~=25.1

{omdev-0.0.0.dev348 → omdev-0.0.0.dev350}/pyproject.toml

@@ -12,7 +12,7 @@ authors = [
 urls = {source = 'https://github.com/wrmsr/omlish'}
 license = {text = 'BSD-3-Clause'}
 requires-python = '>=3.12'
-version = '0.0.0.dev348'
+version = '0.0.0.dev350'
 classifiers = [
     'License :: OSI Approved :: BSD License',
     'Development Status :: 2 - Pre-Alpha',
@@ -22,7 +22,7 @@ classifiers = [
 ]
 description = 'omdev'
 dependencies = [
-    'omlish == 0.0.0.dev348',
+    'omlish == 0.0.0.dev350',
 ]
 
 [project.optional-dependencies]