Upload converters/pipeline/step9_html.py

2026-03-19 09:13:26 +09:00
parent 3254de425d
commit 46ee059753
1 changed files with 103 additions and 0 deletions
--- a/03.Code/업로드용/converters/pipeline/step9_html.py
+++ b/03.Code/업로드용/converters/pipeline/step9_html.py
@@ -0,0 +1,103 @@
+# -*- coding: utf-8 -*-
+from dotenv import load_dotenv
+load_dotenv()
+
+"""
+9_md_to_html_publisher.py
+
+기능:
+- 생성된 report_draft.md 파일을 읽어 최종 report.html을 생성합니다.
+- 마크다운을 HTML로 변환하며, 지정된 스타일시트(Word/HWP 스타일)를 적용합니다.
+- 목차(TOC) 자동 생성 및 링크 기능을 포함합니다.
+"""
+
+import os
+import re
+import json
+from pathlib import Path
+from datetime import datetime
+from typing import List, Dict, Any, Tuple
+
+def log(msg: str):
+    print(f"[{datetime.now().strftime('%H:%M:%S')}] {msg}")
+
+class MarkdownToHtmlConverter:
+    def __init__(self):
+        pass
+
+    def convert(self, md_content: str) -> str:
+        """단순 마크다운 -> HTML 변환 (정규식 기반)"""
+        html = md_content
+        
+        # 헤더
+        html = re.sub(r'^#### (.*)$', r'<h4>\1</h4>', html, flags=re.MULTILINE)
+        html = re.sub(r'^### (.*)$', r'<h3>\1</h3>', html, flags=re.MULTILINE)
+        html = re.sub(r'^## (.*)$', r'<h2>\1</h2>', html, flags=re.MULTILINE)
+        html = re.sub(r'^# (.*)$', r'<h1>\1</h1>', html, flags=re.MULTILINE)
+        
+        # 강조
+        html = re.sub(r'\*\*(.*?)\*\*', r'<strong>\1</strong>', html)
+        
+        # 리스트
+        html = re.sub(r'^\s*-\s+(.*)$', r'<li>\1</li>', html, flags=re.MULTILINE)
+        
+        # 줄바꿈
+        html = html.replace('\n', '<br>\n')
+        
+        return html
+
+def get_html_template(title: str, content: str) -> str:
+    """최종 HTML 템플릿 적용"""
+    return f"""<!DOCTYPE html>
+<html lang="ko">
+<head>
+    <meta charset="UTF-8">
+    <title>{title}</title>
+    <style>
+        @import url('https://fonts.googleapis.com/css2?family=Noto+Sans+KR:wght@300;400;500;700;900&display=swap');
+        body {{ font-family: 'Noto Sans KR', sans-serif; line-height: 1.6; padding: 50px; max-width: 900px; margin: auto; color: #333; }}
+        h1 {{ color: #1a365d; border-bottom: 3px solid #1a365d; padding-bottom: 10px; text-align: center; }}
+        h2 {{ color: #2c5282; border-left: 5px solid #2c5282; padding-left: 15px; margin-top: 40px; }}
+        h3 {{ color: #2b6cb0; margin-top: 30px; }}
+        h4 {{ color: #4a5568; background: #edf2f7; padding: 8px 15px; border-radius: 5px; }}
+        strong {{ color: #2d3748; }}
+        .content {{ margin-top: 30px; }}
+        @media print {{ body {{ padding: 0; }} .no-print {{ display: none; }} }}
+    </style>
+</head>
+<body>
+    <div class="content">
+        {content}
+    </div>
+</body>
+</html>"""
+
+def main(input_dir, output_dir):
+    global OUTPUT_ROOT, GEN_DIR
+    OUTPUT_ROOT = Path(output_dir)
+    GEN_DIR     = OUTPUT_ROOT / "generated"
+    
+    md_path = GEN_DIR / "report_draft.md"
+    out_path = GEN_DIR / "report.html"
+    
+    if not md_path.exists():
+        log(f"대상 파일 없음: {md_path}")
+        return
+
+    log("HTML 변환 작업 시작...")
+    md_content = md_path.read_text(encoding="utf-8")
+    
+    # 제목 추출
+    title_match = re.search(r'^# (.*)$', md_content, re.MULTILINE)
+    title = title_match.group(1) if title_match else "보고서"
+    
+    converter = MarkdownToHtmlConverter()
+    html_body = converter.convert(md_content)
+    
+    final_html = get_html_template(title, html_body)
+    
+    out_path.write_text(final_html, encoding="utf-8")
+    log(f"최종 HTML 생성 완료: {out_path}")
+
+if __name__ == "__main__":
+    main()