Merge pull request #902 from myhloli/fix-line-over-512

refactor(pdf_parse): adjust line count threshold for layoutreader
opendatalab · Nov 8, 2024 · 5f79453 · 5f79453
2 parents 2600d32 + 5936684
commit 5f79453
Showing 1 changed file with 1 addition and 1 deletion.
diff --git a/magic_pdf/pdf_parse_union_core_v2.py b/magic_pdf/pdf_parse_union_core_v2.py
@@ -298,7 +298,7 @@ def sort_lines_by_model(fix_blocks, page_w, page_h, line_height):
                 block['lines'].append({'bbox': line, 'spans': []})
             page_line_list.extend(lines)
 
-    if len(page_line_list) > 512:  # layoutreader最高支持512line
+    if len(page_line_list) > 200:  # layoutreader最高支持512line
         return None
 
     # 使用layoutreader排序