40 lines
1.2 KiB
Python
40 lines
1.2 KiB
Python
import json
|
|
import sys
|
|
from pathlib import Path
|
|
|
|
sys.stdout.reconfigure(encoding='utf-8')
|
|
|
|
# Load 1st audit results
|
|
with open('resources/knowledge/technical_info/01_임도/05_원가정보/품셈_원문md_마스터데이터_대조_검증보고서.json', 'r', encoding='utf-8') as f:
|
|
audit_v1 = json.load(f)
|
|
|
|
refs_27 = audit_v1['item5_references']
|
|
|
|
# Load enriched tables to get original MD lines
|
|
sys.path.insert(0, '.')
|
|
from scratch.enrich_tables import tables
|
|
|
|
tables_by_id = {t['table_id']: t for t in tables}
|
|
|
|
print(f"Analyzing all {len(refs_27)} suspect reference tables from original MD:\n")
|
|
|
|
for r in refs_27:
|
|
tid = r['table_id']
|
|
t = tables_by_id.get(tid)
|
|
if not t:
|
|
continue
|
|
|
|
print(f"==================== [{tid} | Line {t['start_line']}] {t['chapter']} | {t['section']} ====================")
|
|
# Print pre-context (last 5 lines before table)
|
|
print("--- Pre-context ---")
|
|
for pl in t['pre_context'][-5:]:
|
|
print(f" {pl}")
|
|
print("--- Table Header & First 2 Rows ---")
|
|
for tl in t['lines'][:4]:
|
|
print(f" {tl.strip()}")
|
|
if t['notes']:
|
|
print("--- Notes ---")
|
|
for n in t['notes'][:3]:
|
|
print(f" {n}")
|
|
print()
|