Normalize Windows paths and handle directory paths for include patterns (#217)

Author: filipchristiansen

src/gitingest/ingestion.py

@@ -12,7 +12,7 @@
 from gitingest.utils.path_utils import _is_safe_symlink
 
 try:
- import tomllib
+ import tomllib # type: ignore[import]
 except ImportError:
  import tomli as tomllib
 

src/gitingest/query_parsing.py

@@ -296,6 +296,9 @@ def _parse_patterns(pattern: Union[str, Set[str]]) -> Set[str]:
  # Remove empty string if present
  parsed_patterns = parsed_patterns - {""}
 
+ # Normalize Windows paths to Unix-style paths
+ parsed_patterns = {p.replace("\\", "/") for p in parsed_patterns}
+
  # Validate and normalize each pattern
  for p in parsed_patterns:
  if not _is_valid_pattern(p):

src/gitingest/utils/ingestion_utils.py

@@ -56,6 +56,9 @@ def _should_include(path: Path, base_path: Path, include_patterns: Set[str]) ->
  return False
 
  rel_str = str(rel_path)
+ if path.is_dir():
+ rel_str += "/"
+
  for pattern in include_patterns:
  if fnmatch(rel_str, pattern):
  return True