import os
import unicodedata
import pandas as pd
import re

dir1 = '/Users/hogyujhang/Library/CloudStorage/Dropbox/Business/Eroom AI Partners/Legal POC portfolio/민사에이전트-사해행위취소및대여금청구/POC - Law-aid Agent/프롬프트/prompt_updates_sequential/요건사실론/사건종류_요건사실론문서자료_done'
dir2 = '/Users/hogyujhang/Library/CloudStorage/Dropbox/Business/Eroom AI Partners/Legal POC portfolio/민사에이전트-사해행위취소및대여금청구/POC - Law-aid Agent/프롬프트/prompt_updates_sequential/요건사실론/사건종류_요건사실론문서'
output_excel = '/Users/hogyujhang/Library/CloudStorage/Dropbox/Business/Eroom AI Partners/Legal POC portfolio/민사에이전트-사해행위취소및대여금청구/POC - Law-aid Agent/프롬프트/prompt_updates_sequential/요건사실론/요건사실론_추가작업대상리스트.xlsx'

# Extract ### from dir1
# File format: 요건사실문서자료_###_1.md
set1 = set()
for f in os.listdir(dir1):
    f_norm = unicodedata.normalize('NFC', f)
    if f_norm.endswith('.md'):
        # Extract between first '_' and last '_'
        parts = f_norm.split('_')
        if len(parts) >= 3:
            # Reconstruct in case ### has '_' inside
            name = '_'.join(parts[1:-1])
            set1.add(name)

# Extract xxx from dir2
# File format: 요건사실론_xxx_v1.md
set2 = set()
for f in os.listdir(dir2):
    f_norm = unicodedata.normalize('NFC', f)
    if f_norm.endswith('.md'):
        parts = f_norm.split('_')
        if len(parts) >= 3:
            name = '_'.join(parts[1:-1])
            set2.add(name)

diff = sorted(list(set1 - set2))

df = pd.DataFrame(diff, columns=['추가작업대상_사건종류'])
df.to_excel(output_excel, index=False)

print(f"Excel file created with {len(diff)} items.")