import subprocess, sys, re

sys.stdout.reconfigure(encoding='utf-8')

res = subprocess.run(['git', 'diff', '-U3'], capture_output=True, text=True, encoding='utf-8')
diff_text = res.stdout

file_diffs = diff_text.split('diff --git ')
print(f'Total file diffs: {len(file_diffs)-1}')

merged_path = 'resources/knowledge/original/원가계산/건설공사_표준품셈/2026년_건설공사_표준품셈.md'
with open(merged_path, 'r', encoding='utf-8') as f:
    text_merged = f.read()

initial_len = len(text_merged)
total_hunks = 0
applied_hunks = 0

for fd in file_diffs[1:]:
    lines = fd.split('\n')
    header = lines[0]
    filename = header.split(' ')[0][2:]
    
    raw_hunks = re.split(r'\n@@\s+[^@]+\s+@@\n', fd)[1:]
    total_hunks += len(raw_hunks)
    
    for h_idx, h in enumerate(raw_hunks):
        h_lines = h.split('\n')
        old_chunk_lines = []
        new_chunk_lines = []
        for l in h_lines:
            if l.startswith('---') or l.startswith('+++') or l.startswith('\\ No newline'):
                continue
            if l.startswith('-'):
                old_chunk_lines.append(l[1:])
            elif l.startswith('+'):
                new_chunk_lines.append(l[1:])
            elif l.startswith(' '):
                old_chunk_lines.append(l[1:])
                new_chunk_lines.append(l[1:])
        
        old_chunk = '\n'.join(old_chunk_lines)
        new_chunk = '\n'.join(new_chunk_lines)
        
        if text_merged.count(old_chunk) == 1:
            text_merged = text_merged.replace(old_chunk, new_chunk, 1)
            applied_hunks += 1
        else:
            print(f"Warning: could not uniquely match hunk {h_idx} of {filename}")

print(f"Total hunks: {total_hunks}, Applied to merged: {applied_hunks}")

if applied_hunks == total_hunks:
    with open(merged_path, 'w', encoding='utf-8') as f:
        f.write(text_merged)
    print(f"SUCCESS: Merged file perfectly updated! Length: {initial_len} -> {len(text_merged)}")
else:
    print("FAILED: Not all hunks were applied. Merged file was not modified.")
    sys.exit(1)