armeta_hackaton / process_all_pdfs.py
bekzhanK1's picture
process 58 pdfs
7fefcdd
#!/usr/bin/env python3
"""
Process all PDF files in the documents folder and generate a single JSON file
in the required format with annotations for signatures and stamps.
"""
import json
import sys
from pathlib import Path
from typing import Dict, Any, List, Tuple
from pipeline import process_pdf_pipeline
def convert_to_annotation_format(
pipeline_result: Dict[str, Any],
annotation_id_start: int = 1
) -> Tuple[Dict[str, Any], int]:
"""
Convert pipeline result to the required annotation format.
Args:
pipeline_result: Result from process_pdf_pipeline
annotation_id_start: Starting annotation ID number
Returns:
Tuple of (formatted_result, next_annotation_id)
"""
pdf_filename = pipeline_result["pdf"]
pages_data = pipeline_result["pages"]
result = {}
annotation_counter = annotation_id_start
for page_data in pages_data:
page_num = page_data.get("page_number", 1)
page_key = f"page_{page_num}"
# Get image dimensions
img_dims = page_data.get("image_dimensions", {})
width = img_dims.get("width", 0)
height = img_dims.get("height", 0)
# Collect all annotations (signatures and stamps only, no QR codes)
annotations = []
# Process signatures
signatures = page_data.get("signatures", [])
for sig in signatures:
bbox = sig.get("bbox", {})
if bbox:
x1 = bbox.get("x1", 0)
y1 = bbox.get("y1", 0)
width_bbox = bbox.get("width", 0)
height_bbox = bbox.get("height", 0)
# Calculate area
area = width_bbox * height_bbox
annotation = {
f"annotation_{annotation_counter}": {
"category": "signature",
"bbox": {
"x": float(x1),
"y": float(y1),
"width": float(width_bbox),
"height": float(height_bbox)
},
"area": float(area)
}
}
annotations.append(annotation)
annotation_counter += 1
# Process stamps
stamps = page_data.get("stamps", [])
for stamp in stamps:
bbox = stamp.get("bbox", {})
if bbox:
x1 = bbox.get("x1", 0)
y1 = bbox.get("y1", 0)
width_bbox = bbox.get("width", 0)
height_bbox = bbox.get("height", 0)
# Calculate area
area = width_bbox * height_bbox
annotation = {
f"annotation_{annotation_counter}": {
"category": "stamp",
"bbox": {
"x": float(x1),
"y": float(y1),
"width": float(width_bbox),
"height": float(height_bbox)
},
"area": float(area)
}
}
annotations.append(annotation)
annotation_counter += 1
# Only include pages that have annotations
if annotations:
result[page_key] = {
"annotations": annotations,
"page_size": {
"width": int(width),
"height": int(height)
}
}
return result, annotation_counter
def process_all_pdfs(
documents_dir: str = "documents",
output_file: str = "all_annotations.json",
stamp_model_path: str = "stamp_detector/stamp_model.pt",
stamp_conf: float = 0.25,
dpi: int = 200
) -> None:
"""
Process all PDF files in the documents folder and generate a single JSON file.
Args:
documents_dir: Directory containing PDF files
output_file: Output JSON file path
stamp_model_path: Path to stamp model
stamp_conf: Confidence threshold for stamp detection
dpi: DPI for PDF to image conversion
"""
documents_path = Path(documents_dir)
if not documents_path.exists():
print(f"Error: Documents directory '{documents_dir}' not found!")
sys.exit(1)
# Find all PDF files
pdf_files = sorted(list(documents_path.glob("*.pdf")))
if not pdf_files:
print(f"No PDF files found in '{documents_dir}' directory!")
sys.exit(1)
print(f"Found {len(pdf_files)} PDF file(s) to process\n")
print("=" * 70)
# Final result dictionary
final_result = {}
annotation_counter = 1
# Process each PDF
for i, pdf_file in enumerate(pdf_files, 1):
print(f"\n[{i}/{len(pdf_files)}] Processing: {pdf_file.name}")
print("-" * 70)
try:
# Process PDF using pipeline
pipeline_result = process_pdf_pipeline(
pdf_path=str(pdf_file),
output_dir="pipeline_outputs",
stamp_model_path=stamp_model_path,
stamp_conf=stamp_conf,
dpi=dpi,
save_intermediate=False
)
# Convert to annotation format
pdf_annotations, annotation_counter = convert_to_annotation_format(
pipeline_result,
annotation_id_start=annotation_counter
)
# Only add to result if there are annotations
if pdf_annotations:
final_result[pdf_file.name] = pdf_annotations
print(f"βœ“ Processed: {len(pdf_annotations)} page(s) with annotations")
else:
print(f"⚠ No annotations found in {pdf_file.name}")
except Exception as e:
print(f"βœ— Error processing {pdf_file.name}: {str(e)}")
import traceback
traceback.print_exc()
continue
# Save to JSON file
output_path = Path(output_file)
with open(output_path, 'w', encoding='utf-8') as f:
json.dump(final_result, f, indent=2, ensure_ascii=False)
print("\n" + "=" * 70)
print("PROCESSING COMPLETE")
print("=" * 70)
print(f"Total PDFs processed: {len(pdf_files)}")
print(f"PDFs with annotations: {len(final_result)}")
print(f"Total annotations: {annotation_counter - 1}")
print(f"Output saved to: {output_path.absolute()}")
print("=" * 70)
if __name__ == "__main__":
import argparse
parser = argparse.ArgumentParser(
description="Process all PDF files in documents folder and generate annotations JSON"
)
parser.add_argument(
"--documents-dir",
default="documents",
help="Directory containing PDF files (default: documents)"
)
parser.add_argument(
"--output",
default="all_annotations.json",
help="Output JSON file path (default: all_annotations.json)"
)
parser.add_argument(
"--stamp-model",
default="stamp_detector/stamp_model.pt",
help="Path to stamp model (default: stamp_detector/stamp_model.pt)"
)
parser.add_argument(
"--stamp-conf",
type=float,
default=0.25,
help="Confidence threshold for stamp detection (default: 0.25)"
)
parser.add_argument(
"--dpi",
type=int,
default=200,
help="DPI for PDF to image conversion (default: 200)"
)
args = parser.parse_args()
process_all_pdfs(
documents_dir=args.documents_dir,
output_file=args.output,
stamp_model_path=args.stamp_model,
stamp_conf=args.stamp_conf,
dpi=args.dpi
)