Aislo/scratch/wiki_linter.py

import os
import re
from pathlib import Path

# 위키 루트 경로 설정
wiki_root = Path(r"D:\02_Software_Prog\임도설계 및 견적자동화 프로그램 개발\docs\wiki")
raw_root = Path(r"D:\02_Software_Prog\임도설계 및 견적자동화 프로그램 개발\docs\raw")

def load_frontmatter(file_path):
    content = file_path.read_text(encoding="utf-8")
    # Frontmatter regex
    match = re.match(r"^---\s*\n(.*?)\n---\s*\n", content, re.DOTALL)
    if match:
        fm_text = match.group(1)
        fm = {}
        # Simple YAML key-value parser for basic string/list properties
        for line in fm_text.splitlines():
            line = line.strip()
            if not line or line.startswith("#"):
                continue
            if ":" in line:
                key, val = line.split(":", 1)
                key = key.strip()
                val = val.strip().strip("'\"")
                fm[key] = val
        return fm, content[match.end():]
    return None, content

def extract_wiki_links(text):
    # Matches [[link]] or [[link|alias]] or [[link#section]] or [[link#section|alias]]
    links = re.findall(r"\[\[(.*?)\]\]", text)
    cleaned_links = []
    for l in links:
        # Split alias
        if "|" in l:
            l = l.split("|")[0]
        # Split section
        if "#" in l:
            l = l.split("#")[0]
        l = l.strip()
        if l:
            cleaned_links.append(l)
    return cleaned_links

def run_lint():
    print("=== Start Wiki Linting (No PyYAML, Filtered) ===")
    
    all_files = list(wiki_root.glob("**/*.md"))
    index_file = wiki_root / "index.md"
    log_file = wiki_root / "log.md"
    
    # 1. 파일 목록화 및 Frontmatter 체크
    pages = {}
    concepts = {}
    errors = []
    warnings = []
    
    for f in all_files:
        rel_path = f.relative_to(wiki_root).as_posix()
        # Skip output files generated by graphify in graphify-out
        if rel_path.startswith("graphify-out/") or rel_path in ["index.md", "log.md", "AGENTS.md", "CLAUDE.md"]:
            continue
            
        fm, body = load_frontmatter(f)
        
        # 기본 규칙 검사
        if not fm:
            errors.append(f"Missing Frontmatter: {rel_path}")
            continue
            
        # status 값 검사
        status = fm.get("status")
        if status not in ["draft", "stable", "stale"]:
            errors.append(f"Invalid status '{status}' in {rel_path}. Must be draft, stable, or stale.")
            
        # 100줄 제한 검사 (10% 즉 110줄 허용)
        lines_count = len(f.read_text(encoding="utf-8").splitlines())
        if lines_count > 110:
            warnings.append(f"File exceeds 110 lines ({lines_count} lines): {rel_path}")
            
        # 페이지와 컨셉 분류
        if "pages/" in rel_path:
            pages[rel_path] = {
                "fm": fm,
                "body": body,
                "path": f,
                "links": extract_wiki_links(body)
            }
            # 파일명 형식 규칙 5 검사: {page_id}_{기능명}.md
            filename = f.name
            page_id = fm.get("page_id")
            if not page_id:
                errors.append(f"Missing page_id in page: {rel_path}")
            else:
                expected_prefix = page_id.split("_")[0] # e.g. A01
                if not filename.startswith(expected_prefix):
                    warnings.append(f"Filename does not match page_id format: {rel_path} (page_id: {page_id})")
        else:
            concepts[rel_path] = {
                "fm": fm,
                "body": body,
                "path": f,
                "links": extract_wiki_links(body)
            }
            
    # 2. 위키 링크 정합성 검증 (Broken Links)
    # 개념/페이지 맵 구성
    available_links = {}
    for p_rel in pages:
        name_no_ext = Path(p_rel).with_suffix("").as_posix()
        available_links[name_no_ext] = p_rel
        # Also map short form if distinct
        short_name = Path(p_rel).name[:-3]
        available_links[short_name] = p_rel
        
    for c_rel in concepts:
        name_no_ext = Path(c_rel).with_suffix("").as_posix()
        available_links[name_no_ext] = c_rel
        # Short form
        short_name = Path(c_rel).name[:-3]
        available_links[short_name] = c_rel
        # Subdirectories for concepts like db_schema/*
        if "concepts/" in name_no_ext:
            available_links[name_no_ext.replace("concepts/", "")] = c_rel

    # 링크 검사
    for rel_path, info in {**pages, **concepts}.items():
        for link in info["links"]:
            if link.startswith("http://") or link.startswith("https://") or link.startswith("file:///"):
                continue
            normalized_link = link.replace("\\", "/")
            if normalized_link not in available_links and f"concepts/{normalized_link}" not in available_links and f"pages/{normalized_link}" not in available_links:
                warnings.append(f"Broken Link in {rel_path}: [[{link}]]")

    # 3. index.md 등록 상태 확인
    index_content = index_file.read_text(encoding="utf-8") if index_file.exists() else ""
    for rel_path in pages:
        short_name = Path(rel_path).name[:-3]
        dir_name = Path(rel_path).parent.name
        expected_ref_dir = f"{dir_name}/{short_name}"
        if expected_ref_dir not in index_content and f"[[{short_name}]]" not in index_content and short_name not in index_content:
            warnings.append(f"Page not indexed in index.md: {rel_path} (Expected link to [[{expected_ref_dir}]] or [[{short_name}]])")

    # 4. 결과 출력
    print(f"\nScanning completed: {len(all_files)} total markdown files.")
    print(f"Detected {len(pages)} pages and {len(concepts)} concept documents.")
    
    print(f"\n--- Errors ({len(errors)}) ---")
    for e in errors:
        print(f"[ERROR] {e}")
        
    print(f"\n--- Warnings ({len(warnings)}) ---")
    for w in warnings:
        print(f"[WARN] {w}")

if __name__ == "__main__":
    run_lint()