Russell Ballestrini
cbed6d9d9a
- Use lxml parser (2-5x faster than html.parser), fallback to html.parser - Combine 20+ noise selectors into single query (1 DOM traversal vs 20) - Add hint parameter to skip content detection for known types - Added lxml to requirements.txt
676 lines
24 KiB
Python
676 lines
24 KiB
Python
#!/usr/bin/env python3
|
|
"""
|
|
Smart HTML to Markdown converter.
|
|
|
|
Detects content structure (forums, blogs, Q&A, e-commerce) and generates
|
|
clean markdown that preserves semantic meaning.
|
|
"""
|
|
|
|
from bs4 import BeautifulSoup, Tag
|
|
from typing import Optional, List, Dict, Tuple
|
|
from urllib.parse import urljoin
|
|
import re
|
|
|
|
|
|
class SmartMarkdownConverter:
|
|
"""Convert HTML to markdown with semantic structure preservation."""
|
|
|
|
def __init__(self, base_url: str = ''):
|
|
self.base_url = base_url
|
|
|
|
def convert(self, html: str, hint: str = None) -> str:
|
|
"""Convert HTML to markdown, auto-detecting content type.
|
|
|
|
Args:
|
|
html: Raw HTML content
|
|
hint: Optional content type hint ('forum', 'blog', 'qa', 'ecommerce') to skip detection
|
|
"""
|
|
# lxml is 2-5x faster than html.parser, fall back if not installed
|
|
try:
|
|
soup = BeautifulSoup(html, 'lxml')
|
|
except Exception:
|
|
soup = BeautifulSoup(html, 'html.parser')
|
|
|
|
# Remove noise
|
|
self._remove_noise(soup)
|
|
|
|
# Fast path: if hint provided, skip detection
|
|
if hint == 'forum':
|
|
posts = soup.select('.post_container, .topic-post, .post-stream > article')
|
|
if posts:
|
|
return self._render_forum(self._extract_forum_posts(soup, posts))
|
|
|
|
# Detect content type and extract accordingly
|
|
content_type, content = self._detect_and_extract(soup)
|
|
|
|
if content_type == 'forum':
|
|
return self._render_forum(content)
|
|
elif content_type == 'blog':
|
|
return self._render_blog(content)
|
|
elif content_type == 'qa':
|
|
return self._render_qa(content)
|
|
elif content_type == 'ecommerce':
|
|
return self._render_ecommerce(content)
|
|
else:
|
|
# Fallback: clean conversion of main content
|
|
return self._render_generic(soup)
|
|
|
|
def _remove_noise(self, soup: BeautifulSoup):
|
|
"""Remove navigation, scripts, styles, and other noise."""
|
|
# Remove script, style, and other non-content elements
|
|
for tag in soup(['script', 'style', 'noscript', 'iframe', 'svg']):
|
|
tag.decompose()
|
|
|
|
# Remove common navigation/chrome elements - single combined selector
|
|
noise_selector = ', '.join([
|
|
'nav', 'header', 'footer', 'aside',
|
|
'.sidebar', '.nav', '.navigation', '.menu', '.breadcrumb',
|
|
'.header', '.footer', '.logo', '.site-logo',
|
|
'#header', '#footer', '#nav', '#sidebar',
|
|
'.d-header', '.d-footer',
|
|
'.pagination', '.pager',
|
|
'[role="banner"]', '[role="navigation"]', '[role="contentinfo"]',
|
|
'.cookie-banner', '.modal', '.popup', '.overlay',
|
|
'.ad', '.ads', '.advertisement', '.sponsored',
|
|
'.social-share', '.share-buttons',
|
|
])
|
|
for tag in soup.select(noise_selector):
|
|
tag.decompose()
|
|
|
|
def _detect_and_extract(self, soup: BeautifulSoup) -> Tuple[str, any]:
|
|
"""Detect content type and extract structured content."""
|
|
|
|
# Forum detection (Discourse, phpBB, vBulletin, etc.)
|
|
forum_indicators = [
|
|
'.post_container', # Discourse archived (has avatar_container + post)
|
|
'.topic-post', # Discourse live
|
|
'.post-stream > article', # Discourse alt
|
|
'.message', # XenForo
|
|
'.postcontainer', # vBulletin
|
|
'article.boxed', # Generic
|
|
]
|
|
for post_sel in forum_indicators:
|
|
posts = soup.select(post_sel)
|
|
if len(posts) >= 2: # At least 2 posts to be a forum thread
|
|
return 'forum', self._extract_forum_posts(soup, posts)
|
|
|
|
# Q&A detection (Stack Exchange, etc.)
|
|
qa_indicators = [
|
|
('.question', '.answer'),
|
|
('#question', '.answer'),
|
|
('.question-page', '.answercell'),
|
|
]
|
|
for q_sel, a_sel in qa_indicators:
|
|
questions = soup.select(q_sel)
|
|
answers = soup.select(a_sel)
|
|
if questions or answers:
|
|
return 'qa', self._extract_qa(soup, questions, answers)
|
|
|
|
# Blog detection
|
|
blog_indicators = [
|
|
'article', '.post', '.entry', '.blog-post',
|
|
'.hentry', '[itemtype*="BlogPosting"]',
|
|
]
|
|
for selector in blog_indicators:
|
|
articles = soup.select(selector)
|
|
if articles and len(articles) <= 10: # Not a listing page
|
|
return 'blog', self._extract_blog(soup, articles)
|
|
|
|
# E-commerce detection
|
|
ecom_indicators = [
|
|
'.product', '.product-detail', '[itemtype*="Product"]',
|
|
'.product-info', '.pdp-main',
|
|
]
|
|
for selector in ecom_indicators:
|
|
products = soup.select(selector)
|
|
if products:
|
|
return 'ecommerce', self._extract_ecommerce(soup, products)
|
|
|
|
return 'generic', soup
|
|
|
|
def _extract_forum_posts(self, soup: BeautifulSoup, posts: List[Tag]) -> List[Dict]:
|
|
"""Extract forum posts with avatar, username, content, timestamp."""
|
|
extracted = []
|
|
|
|
for post in posts:
|
|
post_data = {
|
|
'avatar': None,
|
|
'username': None,
|
|
'timestamp': None,
|
|
'content': None,
|
|
'quotes': [],
|
|
}
|
|
|
|
# Avatar - look for common patterns (may be sibling or child)
|
|
avatar = (
|
|
post.select_one('.avatar_container img.avatar') or # Discourse archived
|
|
post.select_one('.avatar_container img') or
|
|
post.select_one('.avatar-container img') or
|
|
post.select_one('img.avatar') or
|
|
post.select_one('.avatar img') or
|
|
post.select_one('.avatar-flair img') or
|
|
post.select_one('.user-avatar img') or
|
|
post.select_one('.postprofile img') or
|
|
post.select_one('.author img')
|
|
)
|
|
if avatar:
|
|
src = avatar.get('src', '')
|
|
# Skip placeholder avatars
|
|
if src and '{size}' not in src:
|
|
post_data['avatar'] = self._resolve_url(src)
|
|
|
|
# Username - Discourse archived uses .user_name, live uses .username
|
|
username_el = (
|
|
post.select_one('.user_name') or # Discourse archived
|
|
post.select_one('.username') or
|
|
post.select_one('.author') or
|
|
post.select_one('.creator a') or
|
|
post.select_one('.user-card-name') or
|
|
post.select_one('a[data-user-card]') or
|
|
post.select_one('.names .name') or
|
|
post.select_one('.postprofile dt') or
|
|
post.select_one('strong.username')
|
|
)
|
|
if username_el:
|
|
post_data['username'] = username_el.get_text(strip=True)
|
|
|
|
# Timestamp
|
|
time_el = (
|
|
post.select_one('time') or
|
|
post.select_one('.post-date') or
|
|
post.select_one('.timestamp') or
|
|
post.select_one('.date') or
|
|
post.select_one('.relative-date')
|
|
)
|
|
if time_el:
|
|
post_data['timestamp'] = time_el.get('title') or time_el.get_text(strip=True)
|
|
|
|
# Content - the main post body
|
|
content_el = (
|
|
post.select_one('.post_content') or # Discourse archived
|
|
post.select_one('.cooked') or # Discourse live
|
|
post.select_one('.post-content') or
|
|
post.select_one('.message-body') or
|
|
post.select_one('.postcontent') or
|
|
post.select_one('.post_body') or
|
|
post.select_one('.entry-content') or
|
|
post.select_one('.content') or
|
|
post.select_one('.post') # Fallback to inner .post div
|
|
)
|
|
if content_el:
|
|
post_data['content'] = self._element_to_markdown(content_el)
|
|
else:
|
|
# Fallback: use whole post but try to exclude metadata
|
|
clone = BeautifulSoup(str(post), 'html.parser')
|
|
for sel in ['.avatar', '.avatar_container', '.user_name', '.username', '.author', '.date', '.post-actions', '.post-menu']:
|
|
for el in clone.select(sel):
|
|
el.decompose()
|
|
post_data['content'] = self._element_to_markdown(clone)
|
|
|
|
if post_data['content']:
|
|
extracted.append(post_data)
|
|
|
|
return extracted
|
|
|
|
def _extract_qa(self, soup: BeautifulSoup, questions: List[Tag], answers: List[Tag]) -> Dict:
|
|
"""Extract Q&A content with votes, user info."""
|
|
qa_data = {'question': None, 'answers': []}
|
|
|
|
if questions:
|
|
q = questions[0]
|
|
qa_data['question'] = {
|
|
'title': self._get_text(q, '.question-title, h1'),
|
|
'votes': self._get_text(q, '.vote-count, .js-vote-count'),
|
|
'content': self._element_to_markdown(q.select_one('.post-text, .s-prose, .question-body')),
|
|
'author': self._get_text(q, '.user-info .user-details a, .author'),
|
|
}
|
|
|
|
for a in answers:
|
|
answer_data = {
|
|
'votes': self._get_text(a, '.vote-count, .js-vote-count'),
|
|
'content': self._element_to_markdown(a.select_one('.post-text, .s-prose, .answer-body')),
|
|
'author': self._get_text(a, '.user-info .user-details a, .author'),
|
|
'accepted': bool(a.select_one('.accepted-answer, .is-accepted')),
|
|
}
|
|
if answer_data['content']:
|
|
qa_data['answers'].append(answer_data)
|
|
|
|
return qa_data
|
|
|
|
def _extract_blog(self, soup: BeautifulSoup, articles: List[Tag]) -> List[Dict]:
|
|
"""Extract blog articles with metadata."""
|
|
extracted = []
|
|
|
|
for article in articles:
|
|
article_data = {
|
|
'title': self._get_text(article, 'h1, h2, .entry-title, .post-title'),
|
|
'author': self._get_text(article, '.author, .byline, [rel="author"]'),
|
|
'date': self._get_text(article, 'time, .date, .published, .post-date'),
|
|
'content': None,
|
|
}
|
|
|
|
content_el = (
|
|
article.select_one('.entry-content') or
|
|
article.select_one('.post-content') or
|
|
article.select_one('.article-body') or
|
|
article.select_one('.content') or
|
|
article
|
|
)
|
|
article_data['content'] = self._element_to_markdown(content_el)
|
|
|
|
if article_data['content']:
|
|
extracted.append(article_data)
|
|
|
|
return extracted
|
|
|
|
def _extract_ecommerce(self, soup: BeautifulSoup, products: List[Tag]) -> List[Dict]:
|
|
"""Extract product information."""
|
|
extracted = []
|
|
|
|
for product in products:
|
|
product_data = {
|
|
'name': self._get_text(product, '.product-name, .product-title, h1, h2'),
|
|
'price': self._get_text(product, '.price, .product-price, [itemprop="price"]'),
|
|
'description': self._element_to_markdown(
|
|
product.select_one('.description, .product-description, [itemprop="description"]')
|
|
),
|
|
'image': None,
|
|
'rating': self._get_text(product, '.rating, .stars, [itemprop="ratingValue"]'),
|
|
}
|
|
|
|
img = product.select_one('.product-image img, .gallery img, [itemprop="image"]')
|
|
if img:
|
|
product_data['image'] = self._resolve_url(img.get('src', ''))
|
|
|
|
if product_data['name']:
|
|
extracted.append(product_data)
|
|
|
|
return extracted
|
|
|
|
def _element_to_markdown(self, el: Optional[Tag]) -> str:
|
|
"""Convert a single element to markdown."""
|
|
if not el:
|
|
return ''
|
|
|
|
lines = []
|
|
self._walk_element(el, lines)
|
|
return '\n'.join(lines).strip()
|
|
|
|
def _walk_element(self, el: Tag, lines: List[str], depth: int = 0):
|
|
"""Recursively walk element and build markdown lines."""
|
|
if isinstance(el, str):
|
|
text = el.strip()
|
|
if text:
|
|
lines.append(text)
|
|
return
|
|
|
|
if not isinstance(el, Tag):
|
|
return
|
|
|
|
tag = el.name.lower() if el.name else ''
|
|
|
|
# Block elements
|
|
if tag in ('p', 'div'):
|
|
# Check if contains block-level elements (pre, ul, ol, etc.)
|
|
block_tags = {'pre', 'ul', 'ol', 'blockquote', 'h1', 'h2', 'h3', 'h4', 'h5', 'h6', 'table'}
|
|
has_blocks = any(
|
|
isinstance(c, Tag) and c.name and c.name.lower() in block_tags
|
|
for c in el.children
|
|
)
|
|
if has_blocks:
|
|
# Recursively process children to preserve block structure
|
|
for child in el.children:
|
|
self._walk_element(child, lines, depth)
|
|
else:
|
|
content = self._inline_content(el)
|
|
if content:
|
|
lines.append(content)
|
|
lines.append('')
|
|
|
|
elif tag in ('h1', 'h2', 'h3', 'h4', 'h5', 'h6'):
|
|
level = int(tag[1])
|
|
content = self._inline_content(el)
|
|
if content:
|
|
lines.append(f"{'#' * level} {content}")
|
|
lines.append('')
|
|
|
|
elif tag == 'blockquote':
|
|
quote_lines = []
|
|
for child in el.children:
|
|
self._walk_element(child, quote_lines, depth + 1)
|
|
for line in quote_lines:
|
|
if line:
|
|
lines.append(f"> {line}")
|
|
else:
|
|
lines.append('>')
|
|
lines.append('')
|
|
|
|
elif tag in ('ul', 'ol'):
|
|
for i, li in enumerate(el.find_all('li', recursive=False)):
|
|
prefix = f"{i+1}." if tag == 'ol' else "-"
|
|
content = self._inline_content(li)
|
|
lines.append(f"{prefix} {content}")
|
|
lines.append('')
|
|
|
|
elif tag == 'pre':
|
|
code = el.get_text()
|
|
# Try to detect language from class
|
|
lang = ''
|
|
code_el = el.select_one('code')
|
|
if code_el:
|
|
classes = code_el.get('class', [])
|
|
for cls in classes:
|
|
if cls.startswith('language-') or cls.startswith('lang-'):
|
|
lang = cls.split('-', 1)[1]
|
|
break
|
|
lines.append(f"```{lang}")
|
|
lines.append(code.strip())
|
|
lines.append("```")
|
|
lines.append('')
|
|
|
|
elif tag == 'code' and el.parent and el.parent.name != 'pre':
|
|
# Inline code handled in _inline_content
|
|
pass
|
|
|
|
elif tag == 'hr':
|
|
lines.append('---')
|
|
lines.append('')
|
|
|
|
elif tag == 'br':
|
|
lines.append('')
|
|
|
|
elif tag == 'img':
|
|
src = self._resolve_url(el.get('src', ''))
|
|
alt = el.get('alt', '')
|
|
if src:
|
|
lines.append(f"")
|
|
lines.append('')
|
|
|
|
elif tag == 'a':
|
|
# Links are handled inline
|
|
pass
|
|
|
|
elif tag == 'table':
|
|
lines.extend(self._table_to_markdown(el))
|
|
lines.append('')
|
|
|
|
else:
|
|
# Generic container - recurse
|
|
for child in el.children:
|
|
self._walk_element(child, lines, depth)
|
|
|
|
def _inline_content(self, el: Tag) -> str:
|
|
"""Extract inline content from an element, handling formatting."""
|
|
parts = []
|
|
|
|
for child in el.children:
|
|
if isinstance(child, str):
|
|
text = child.strip()
|
|
if text:
|
|
# Collapse whitespace
|
|
text = re.sub(r'\s+', ' ', text)
|
|
parts.append(text)
|
|
elif isinstance(child, Tag):
|
|
tag = child.name.lower() if child.name else ''
|
|
|
|
if tag in ('strong', 'b'):
|
|
inner = self._inline_content(child)
|
|
if inner:
|
|
parts.append(f"**{inner}**")
|
|
|
|
elif tag in ('em', 'i'):
|
|
inner = self._inline_content(child)
|
|
if inner:
|
|
parts.append(f"*{inner}*")
|
|
|
|
elif tag == 'code':
|
|
code = child.get_text()
|
|
if code:
|
|
parts.append(f"`{code}`")
|
|
|
|
elif tag == 'a':
|
|
href = self._resolve_url(child.get('href', ''))
|
|
text = self._inline_content(child) or href
|
|
if href:
|
|
parts.append(f"[{text}]({href})")
|
|
else:
|
|
parts.append(text)
|
|
|
|
elif tag == 'img':
|
|
src = self._resolve_url(child.get('src', ''))
|
|
alt = child.get('alt', '')
|
|
if src:
|
|
# Emoji images (alt like :smile:) stay inline
|
|
if alt.startswith(':') and alt.endswith(':'):
|
|
parts.append(f"")
|
|
else:
|
|
# Regular images are block-level
|
|
parts.append(f"\n\n\n\n")
|
|
|
|
elif tag == 'br':
|
|
parts.append('\n')
|
|
|
|
elif tag in ('span', 'small', 'mark'):
|
|
# Pass through
|
|
inner = self._inline_content(child)
|
|
if inner:
|
|
parts.append(inner)
|
|
|
|
else:
|
|
# Unknown inline - just get text
|
|
inner = self._inline_content(child)
|
|
if inner:
|
|
parts.append(inner)
|
|
|
|
return ' '.join(parts).strip()
|
|
|
|
def _table_to_markdown(self, table: Tag) -> List[str]:
|
|
"""Convert HTML table to markdown."""
|
|
lines = []
|
|
rows = table.find_all('tr')
|
|
if not rows:
|
|
return lines
|
|
|
|
# Extract headers
|
|
header_row = rows[0]
|
|
headers = [self._inline_content(th) for th in header_row.find_all(['th', 'td'])]
|
|
if headers:
|
|
lines.append('| ' + ' | '.join(headers) + ' |')
|
|
lines.append('| ' + ' | '.join(['---'] * len(headers)) + ' |')
|
|
|
|
# Extract body rows
|
|
for row in rows[1:]:
|
|
cells = [self._inline_content(td) for td in row.find_all(['td', 'th'])]
|
|
if cells:
|
|
# Pad cells if needed
|
|
while len(cells) < len(headers):
|
|
cells.append('')
|
|
lines.append('| ' + ' | '.join(cells) + ' |')
|
|
|
|
return lines
|
|
|
|
def _get_text(self, el: Tag, selectors: str) -> str:
|
|
"""Get text from first matching selector."""
|
|
for selector in selectors.split(','):
|
|
found = el.select_one(selector.strip())
|
|
if found:
|
|
return found.get_text(strip=True)
|
|
return ''
|
|
|
|
def _resolve_url(self, url: str) -> str:
|
|
"""Resolve relative URL to absolute."""
|
|
if not url or url.startswith('data:'):
|
|
return url
|
|
if self.base_url and not url.startswith(('http://', 'https://', '//')):
|
|
return urljoin(self.base_url, url)
|
|
return url
|
|
|
|
# Renderers for different content types
|
|
|
|
def _render_forum(self, posts: List[Dict]) -> str:
|
|
"""Render forum posts to markdown."""
|
|
lines = []
|
|
|
|
for i, post in enumerate(posts):
|
|
if i > 0:
|
|
lines.append('')
|
|
lines.append('---')
|
|
lines.append('')
|
|
|
|
# Post header with avatar and username
|
|
# Always include avatar (placeholder if none) to maintain grid layout
|
|
if post['avatar']:
|
|
avatar_md = f""
|
|
else:
|
|
# Use # as placeholder - JS will detect and replace with initial
|
|
avatar_md = ""
|
|
|
|
header_parts = [avatar_md]
|
|
if post['username']:
|
|
header_parts.append(f"**{post['username']}**")
|
|
if post['timestamp']:
|
|
header_parts.append(f"*{post['timestamp']}*")
|
|
|
|
lines.append(' '.join(header_parts))
|
|
lines.append('')
|
|
|
|
# Post content
|
|
if post['content']:
|
|
lines.append(post['content'])
|
|
|
|
return '\n'.join(lines)
|
|
|
|
def _render_qa(self, qa: Dict) -> str:
|
|
"""Render Q&A content to markdown."""
|
|
lines = []
|
|
|
|
if qa['question']:
|
|
q = qa['question']
|
|
if q['title']:
|
|
lines.append(f"# {q['title']}")
|
|
lines.append('')
|
|
if q['votes']:
|
|
lines.append(f"**{q['votes']} votes**")
|
|
if q['author']:
|
|
lines.append(f"*Asked by {q['author']}*")
|
|
lines.append('')
|
|
if q['content']:
|
|
lines.append(q['content'])
|
|
lines.append('')
|
|
|
|
if qa['answers']:
|
|
lines.append('---')
|
|
lines.append('')
|
|
lines.append(f"## {len(qa['answers'])} Answers")
|
|
lines.append('')
|
|
|
|
for answer in qa['answers']:
|
|
if answer['accepted']:
|
|
lines.append('### ✓ Accepted Answer')
|
|
else:
|
|
lines.append('### Answer')
|
|
if answer['votes']:
|
|
lines.append(f"**{answer['votes']} votes**")
|
|
if answer['author']:
|
|
lines.append(f"*By {answer['author']}*")
|
|
lines.append('')
|
|
if answer['content']:
|
|
lines.append(answer['content'])
|
|
lines.append('')
|
|
lines.append('---')
|
|
lines.append('')
|
|
|
|
return '\n'.join(lines)
|
|
|
|
def _render_blog(self, articles: List[Dict]) -> str:
|
|
"""Render blog articles to markdown."""
|
|
lines = []
|
|
|
|
for article in articles:
|
|
if article['title']:
|
|
lines.append(f"# {article['title']}")
|
|
lines.append('')
|
|
|
|
meta = []
|
|
if article['author']:
|
|
meta.append(f"By {article['author']}")
|
|
if article['date']:
|
|
meta.append(article['date'])
|
|
if meta:
|
|
lines.append(f"*{' | '.join(meta)}*")
|
|
lines.append('')
|
|
|
|
if article['content']:
|
|
lines.append(article['content'])
|
|
|
|
return '\n'.join(lines)
|
|
|
|
def _render_ecommerce(self, products: List[Dict]) -> str:
|
|
"""Render product info to markdown."""
|
|
lines = []
|
|
|
|
for product in products:
|
|
if product['name']:
|
|
lines.append(f"# {product['name']}")
|
|
lines.append('')
|
|
|
|
if product['image']:
|
|
lines.append(f"![{product['name']}]({product['image']})")
|
|
lines.append('')
|
|
|
|
if product['price']:
|
|
lines.append(f"**Price:** {product['price']}")
|
|
lines.append('')
|
|
|
|
if product['rating']:
|
|
lines.append(f"**Rating:** {product['rating']}")
|
|
lines.append('')
|
|
|
|
if product['description']:
|
|
lines.append(product['description'])
|
|
|
|
return '\n'.join(lines)
|
|
|
|
def _render_generic(self, soup: BeautifulSoup) -> str:
|
|
"""Fallback generic rendering."""
|
|
# Find main content area
|
|
main = (
|
|
soup.select_one('main') or
|
|
soup.select_one('article') or
|
|
soup.select_one('.content') or
|
|
soup.select_one('#content') or
|
|
soup.select_one('.main') or
|
|
soup.body or
|
|
soup
|
|
)
|
|
|
|
return self._element_to_markdown(main)
|
|
|
|
|
|
def html_to_markdown(html: str, base_url: str = '', hint: str = None) -> str:
|
|
"""Convert HTML to markdown with smart structure detection.
|
|
|
|
Args:
|
|
html: Raw HTML content
|
|
base_url: Base URL for resolving relative links
|
|
hint: Content type hint ('forum', 'blog', 'qa', 'ecommerce') to skip auto-detection
|
|
|
|
Returns:
|
|
Clean markdown with preserved semantic structure
|
|
"""
|
|
converter = SmartMarkdownConverter(base_url)
|
|
return converter.convert(html, hint=hint)
|
|
|
|
|
|
if __name__ == '__main__':
|
|
import sys
|
|
|
|
if len(sys.argv) < 2:
|
|
print("Usage: python html2md.py <file.html> [base_url]")
|
|
sys.exit(1)
|
|
|
|
with open(sys.argv[1]) as f:
|
|
html = f.read()
|
|
|
|
base_url = sys.argv[2] if len(sys.argv) > 2 else ''
|
|
print(html_to_markdown(html, base_url))
|