"""문서 파서: HTML/Markdown을 구조화."""
from __future__ import annotations
import re
from dataclasses import dataclass
from typing import Any, Dict, List, Optional
from bs4 import BeautifulSoup # type: ignore
import markdown as md # type: ignore
@dataclass(frozen=True, slots=True)
class ParsedSection:
title: str
content: str
code_blocks: List[str]
anchors: List[str]
class DocsParser:
"""단순한 HTML/Markdown 파서 (기본 구현)."""
def parse_html(self, html: str) -> List[ParsedSection]:
soup = BeautifulSoup(html, "html.parser")
sections: List[ParsedSection] = []
for heading in soup.find_all(["h1", "h2", "h3"]):
title = heading.get_text(strip=True)
content_parts: List[str] = []
code_blocks: List[str] = []
anchors: List[str] = []
# siblings until next heading
for sibling in heading.next_siblings:
if getattr(sibling, "name", None) in ["h1", "h2", "h3"]:
break
text = ""
if hasattr(sibling, "get_text"):
text = sibling.get_text(" ", strip=True)
elif isinstance(sibling, str):
text = sibling.strip()
if text:
content_parts.append(text)
if getattr(sibling, "name", None) == "pre":
code_text = sibling.get_text("\n", strip=True)
if code_text:
code_blocks.append(code_text)
if title or content_parts:
sections.append(
ParsedSection(
title=title or "Untitled",
content="\n".join(content_parts),
code_blocks=code_blocks,
anchors=anchors,
)
)
return sections
def parse_markdown(self, markdown_text: str) -> List[ParsedSection]:
# Convert markdown to HTML then reuse html parser
html = md.markdown(markdown_text)
return self.parse_html(html)
__all__ = ["DocsParser", "ParsedSection"]
