Wechat Article Publisher

#!/usr/bin/env python3
"""
Parse Markdown for WeChat Official Account article publishing.

Extracts:
- Title (from first H1/H2 or first line)
- Cover image (first image)
- Content images with block index for precise positioning
- HTML content (images stripped)

Usage:
    python parse_markdown.py <markdown_file> [--output json|html]

Output (JSON):
{
    "title": "Article Title",
    "cover_image": "/path/to/cover.jpg",
    "content_images": [
        {"path": "/path/to/img.jpg", "block_index": 3, "after_text": "context..."},
        ...
    ],
    "html": "<p>Content...</p><h2>Section</h2>...",
    "total_blocks": 25
}

The block_index indicates which block element (0-indexed) the image should follow.
This allows precise positioning without relying on text matching.
"""

import argparse
import json
import os
import re
import sys
from pathlib import Path


def split_into_blocks(markdown: str) -> list[str]:
    """Split markdown into logical blocks (paragraphs, headers, quotes, code blocks, etc.)."""
    blocks = []
    current_block = []
    in_code_block = False
    code_block_lines = []

    lines = markdown.split('\n')

    for line in lines:
        stripped = line.strip()

        # Handle code block boundaries
        if stripped.startswith('```'):
            if in_code_block:
                # End of code block
                in_code_block = False
                if code_block_lines:
                    # Mark as code block with special prefix for later processing
                    # Use ___CODE_BLOCK_START___ and ___CODE_BLOCK_END___ to preserve content
                    blocks.append('___CODE_BLOCK_START___' + '\n'.join(code_block_lines) + '___CODE_BLOCK_END___')
                code_block_lines = []
            else:
                # Start of code block
                if current_block:
                    blocks.append('\n'.join(current_block))
                    current_block = []
                in_code_block = True
            continue

        # If inside code block, collect ALL lines (including empty lines)
        if in_code_block:
            code_block_lines.append(line)
            continue

        # Empty line signals end of block
        if not stripped:
            if current_block:
                blocks.append('\n'.join(current_block))
                current_block = []
            continue

        # Headers, blockquotes are their own blocks
        if stripped.startswith(('#', '>')):
            if current_block:
                blocks.append('\n'.join(current_block))
                current_block = []
            blocks.append(stripped)
            continue

        # Image on its own line is its own block
        if re.match(r'^!\[.*\]\(.*\)$', stripped):
            if current_block:
                blocks.append('\n'.join(current_block))
                current_block = []
            blocks.append(stripped)
            continue

        current_block.append(line)

    if current_block:
        blocks.append('\n'.join(current_block))

    # Handle unclosed code block
    if code_block_lines:
        blocks.append('___CODE_BLOCK_START___' + '\n'.join(code_block_lines) + '___CODE_BLOCK_END___')

    return blocks


def extract_images_with_block_index(markdown: str, base_path: Path) -> tuple[list[dict], str, int]:
    """Extract images with their block index position.

    Returns:
        (image_list, markdown_without_images, total_blocks)
    """
    blocks = split_into_blocks(markdown)
    images = []
    clean_blocks = []

    img_pattern = re.compile(r'^!\[([^\]]*)\]\(([^)]+)\)$')

    for i, block in enumerate(blocks):
        match = img_pattern.match(block.strip())
        if match:
            alt_text = match.group(1)
            img_path = match.group(2)

            # Resolve relative paths
            if not os.path.isabs(img_path):
                full_path = str(base_path / img_path)
            else:
                full_path = img_path

            # block_index is the index