import sys, os, re, json
sys.stdout.reconfigure(encoding='utf-8')

base_dir = "resources/knowledge/original/원가계산/건설공사_표준품셈"
merged_file = os.path.join(base_dir, "2026년_건설공사_표준품셈.md")

with open(merged_file, 'r', encoding='utf-8') as f:
    merged_text = f.read()

all_md_files = []
for root, dirs, files in os.walk(base_dir):
    for f in files:
        if f.endswith('.md') and not f.startswith('_') and '개정사항' not in f and '2026년_건설공사_표준품셈.md' not in f:
            all_md_files.append(os.path.join(root, f))

# 1. Check divider column mismatch in all MD files (including merged)
files_to_check = all_md_files + [merged_file]

divider_mismatches = []
for fpath in files_to_check:
    rel = os.path.relpath(fpath, base_dir)
    with open(fpath, 'r', encoding='utf-8') as f:
        lines = f.readlines()
        
    for i in range(len(lines) - 1):
        l1 = lines[i].strip()
        l2 = lines[i+1].strip()
        
        # Check if l2 is divider
        if l1.startswith('|') and l1.endswith('|') and l2.startswith('|') and l2.endswith('|'):
            # check if l2 consists of --- and |
            if re.match(r'^\|(?:\s*:?-+:?\s*\|)+$', l2):
                c1 = len(l1.split('|')) - 2
                c2 = len(l2.split('|')) - 2
                if c1 != c2:
                    divider_mismatches.append({
                        "file": rel,
                        "line": i + 2,
                        "header_cols": c1,
                        "divider_cols": c2,
                        "header": l1[:60],
                        "divider": l2[:40]
                    })

print(f"=== Divider Column Mismatches across all files: {len(divider_mismatches)} ===")
for dm in divider_mismatches[:20]:
    print(f" - {dm['file']} (L{dm['line']}): header={dm['header_cols']} vs divider={dm['divider_cols']} | {dm['header']}")

# 2. Check 3-line spec displacement
# Tables where header has 형식 and 출력(kW) or similar
spec_displacements = []
for fpath in files_to_check:
    rel = os.path.relpath(fpath, base_dir)
    with open(fpath, 'r', encoding='utf-8') as f:
        lines = f.readlines()
        
    for i, l in enumerate(lines):
        if '출력' in l and '㎾' in l and '|' in l:
            # check subsequent rows
            # if next rows have misplaced cell values or shifted empty cells
            pass

with open('scratch/divider_mismatches.json', 'w', encoding='utf-8') as f:
    json.dump(divider_mismatches, f, ensure_ascii=False, indent=2)