TBNilles/LEANN
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Format code style in leann_multi_vector.py for better readability

Browse Source 
🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
This commit is contained in:
yichuan-w
2025-12-17 09:02:48 +00:00
parent 6c39a3427f
commit a878d2459b
1 changed files with 17 additions and 13 deletions
Download Patch File Download Diff File Expand all files Collapse all files

30
apps/multimodal/vision-based-pdf-multi-vector/leann_multi_vector.py
View File

@@ -97,21 +97,23 @@ def _natural_sort_key(name: str) -> int:
return int(m.group()) if m else 0 return int(m.group()) if m else 0
def _load_images_from_dir(pages_dir: str, recursive: bool = False) -> tuple[list[str], list[Image.Image]]: def _load_images_from_dir(
pages_dir: str, recursive: bool = False
) -> tuple[list[str], list[Image.Image]]:
""" """
Load images from a directory. Load images from a directory.
Args: Args:
pages_dir: Directory path containing images pages_dir: Directory path containing images
recursive: If True, recursively search subdirectories (default: False) recursive: If True, recursively search subdirectories (default: False)
Returns: Returns:
Tuple of (filepaths, images) Tuple of (filepaths, images)
""" """
# Supported image extensions # Supported image extensions
extensions = ("*.png", "*.jpg", "*.jpeg", "*.PNG", "*.JPG", "*.JPEG", "*.webp", "*.WEBP") extensions = ("*.png", "*.jpg", "*.jpeg", "*.PNG", "*.JPG", "*.JPEG", "*.webp", "*.WEBP")
if recursive: if recursive:
# Recursive search # Recursive search
filepaths = [] filepaths = []
@@ -124,31 +126,33 @@ def _load_images_from_dir(pages_dir: str, recursive: bool = False) -> tuple[list
for ext in extensions: for ext in extensions:
pattern = os.path.join(pages_dir, ext) pattern = os.path.join(pages_dir, ext)
filepaths.extend(glob.glob(pattern)) filepaths.extend(glob.glob(pattern))
# Sort files naturally # Sort files naturally
filepaths = sorted(filepaths, key=lambda x: _natural_sort_key(os.path.basename(x))) filepaths = sorted(filepaths, key=lambda x: _natural_sort_key(os.path.basename(x)))
# Load images with error handling # Load images with error handling
images = [] images = []
valid_filepaths = [] valid_filepaths = []
failed_count = 0 failed_count = 0
for filepath in filepaths: for filepath in filepaths:
try: try:
img = Image.open(filepath) img = Image.open(filepath)
# Convert to RGB if necessary (handles RGBA, P, etc.) # Convert to RGB if necessary (handles RGBA, P, etc.)
if img.mode != 'RGB': if img.mode != "RGB":
img = img.convert('RGB') img = img.convert("RGB")
images.append(img) images.append(img)
valid_filepaths.append(filepath) valid_filepaths.append(filepath)
except Exception as e: except Exception as e:
failed_count += 1 failed_count += 1
print(f"Warning: Failed to load image {filepath}: {e}") print(f"Warning: Failed to load image {filepath}: {e}")
continue continue
if failed_count > 0: if failed_count > 0:
print(f"Warning: Failed to load {failed_count} image(s) out of {len(filepaths)} total files") print(
f"Warning: Failed to load {failed_count} image(s) out of {len(filepaths)} total files"
)
return valid_filepaths, images return valid_filepaths, images