import sys, os, re
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()

def count_tables_in_text(text):
    lines = text.split('\n')
    tables = 0
    curr = []
    for l in lines:
        ls = l.strip()
        if ls.startswith('|') and ls.endswith('|'):
            curr.append(ls)
        else:
            if curr:
                if any(re.match(r'^\|(?:\s*:?-+:?\s*\|)+$', r) for r in curr):
                    tables += 1
                curr = []
    if curr and any(re.match(r'^\|(?:\s*:?-+:?\s*\|)+$', r) for r in curr):
        tables += 1
    return tables

# Split merged_text by divisions:
# 공통부문, 토목부문, 건축부문, 기계설비부문, 유지관리부문
div_names = ["공통부문", "토목부문", "건축부문", "기계설비부문", "유지관리부문"]
# find division boundaries in merged
div_spans = []
for dn in div_names:
    # search for '# ' or main header of division
    m = re.search(rf'(?:^|\n)#+\s*{dn}', merged_text)
    if not m:
        m = re.search(rf'(?:^|\n){dn}', merged_text)
    pos = m.start() if m else -1
    div_spans.append((dn, pos))

print(f"Division spans in merged: {div_spans}")

# Count per division folder in split
all_split_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_split_files.append(os.path.join(root, f))

split_by_div = {}
for p in all_split_files:
    rel = os.path.relpath(p, base_dir)
    div = rel.split(os.sep)[0]
    with open(p, 'r', encoding='utf-8') as f:
        txt = f.read()
    c = count_tables_in_text(txt)
    split_by_div[div] = split_by_div.get(div, 0) + c

for d, cnt in sorted(split_by_div.items()):
    print(f"Split {d}: {cnt} tables")