Files
Wusunjiance/ocr_work/export_md.py
T

54 lines
1.8 KiB
Python

import json
import sys
import os
def extract_text_from_content(content_list):
"""Recursively extract text from content structure"""
result = []
for item in content_list:
if isinstance(item, dict):
if item.get('type') == 'text' and 'content' in item:
text = item['content']
if isinstance(text, str):
result.append(text)
elif 'content' in item:
result.extend(extract_text_from_content(item['content']))
return result
def json_to_markdown(json_path):
with open(json_path, 'r', encoding='utf-8') as f:
data = json.load(f)
md_lines = []
for page in data['pages']:
page_idx = page['page_idx']
md_lines.append(f"\n<!-- Page {page_idx + 1} -->\n")
for block in page['blocks']:
btype = block.get('type', 'text')
texts = extract_text_from_content(block.get('content', []))
full_text = ''.join(texts).strip()
if not full_text:
continue
if btype == 'doc_title':
level = block.get('level', 1)
md_lines.append(f"{'#' * level} {full_text}\n")
elif btype == 'text':
md_lines.append(full_text + "\n")
elif btype == 'header':
md_lines.append(f"<!-- header: {full_text} -->\n")
elif btype == 'footer':
md_lines.append(f"<!-- footer: {full_text} -->\n")
elif btype == 'page_number':
md_lines.append(f"<!-- page: {full_text} -->\n")
else:
md_lines.append(full_text + "\n")
return '\n'.join(md_lines)
if __name__ == '__main__':
json_path = sys.argv[1]
md = json_to_markdown(json_path)
print(md)