mirror of
https://github.com/lllin000/PaperForge.git
synced 2026-07-22 06:50:53 +00:00
23 lines
1 KiB
Python
23 lines
1 KiB
Python
import json, sys
|
|
sys.path.insert(0, '.')
|
|
from paperforge.worker.ocr_figures import build_figure_inventory
|
|
|
|
for paper in ['6QNRHRKX', 'KIX7SKXQ']:
|
|
with open(f'D:/L/OB/Literature-hub/System/PaperForge/ocr/{paper}/structure/blocks.structured.jsonl','r',encoding='utf-8') as f:
|
|
blocks = [json.loads(l) for l in f.read().strip().split('\n') if l.strip()]
|
|
result = build_figure_inventory(blocks)
|
|
mt = len(result.get('matched_figures', []))
|
|
amb = len(result.get('ambiguous_figures', []))
|
|
print(f'{paper}: matched={mt} ambiguous={amb}')
|
|
for a in result.get('ambiguous_figures',[]):
|
|
fn = a.get('figure_number','?')
|
|
hr = a.get('hold_reason','?')
|
|
ca = len(a.get('candidates',[]))
|
|
pg = a.get('page','?')
|
|
print(f' Fig {fn} p{pg}: {ca}cand {hr}')
|
|
for m in result.get('matched_figures',[]):
|
|
fn = m.get('figure_number','?')
|
|
pg = m.get('page')
|
|
n = len(m.get('matched_assets',[]))
|
|
fl = m.get('flags',[])
|
|
print(f' MATCHED Fig {fn} p{pg}: {n} assets {fl}')
|