scripts/fix-articles.py - master - ArtKAtMan/arbat

scripts/fix-articles.py
155 строк5 KB
Mavis (M3)
fix: article markdown — bold/FAQ rendering, wider container, related grid 4-col
10 июл 2026, 18:41
10 июл 2026, 18:41a416f50
#!/usr/bin/env python3
"""
One-shot fixer for arbat-seo articles.

Problems to fix:
1. First # heading duplicates the h1 in hero. Remove if it matches the title.
2. ### H3. prefix on H3 subheadings — render artifact, remove it.
3. **bold** glued to surrounding text without spaces. Add space inside markers.
4. FAQ: each Q/A currently collapsed into one giant H2 line. Split on ** markers
   and add blank lines so markdown renders them as paragraphs/headings properly.

Usage:  python fix_articles.py
"""
import re
from pathlib import Path

CONTENT_DIR = Path('src/content/articles')

# Step 1: split frontmatter and body
FM_RE = re.compile(r'^---\s*\n(.*?)\n---\s*\n(.*)$', re.DOTALL)

# Step 2: H3 prefix to remove
H3_PREFIX = re.compile(r'^(#{3,6})\s*H\d+\.\s+', re.MULTILINE)

# Step 3: bold that needs spaces inside
# `**Word**` (no spaces) -> `** Word **`
BOLD_NO_SPACES = re.compile(r'\*\*(\S[^*]{0,200}?)\*\*')

# Step 4: a single H2 that has a big block of `**Q?** A. **Q2?** A2. ...`
# We replace the H2 with proper structure: H2, then each Q+A on its own block.
# Pattern: `## FAQ` followed by `**Q?** ... **Q2?** ...` all in one chunk.
FAQ_RE = re.compile(
    r'^(##\s*FAQ\s*)(\*\*[^*]+\?\*\*[\s\S]+?)(?=^##\s|\Z)',
    re.MULTILINE,
)


def split_frontmatter(text: str):
    m = FM_RE.match(text)
    if not m:
        return text, ''
    return m.group(1), m.group(2)


def join_frontmatter(fm: str, body: str) -> str:
    if not fm:
        return body
    return f'---\n{fm}\n---\n\n{body}'


def remove_first_h1(body: str, title: str) -> str:
    """Remove the first # heading if it matches the article title."""
    lines = body.split('\n')
    out = []
    removed = False
    title_norm = title.strip().lower()
    for line in lines:
        if not removed and line.startswith('# ') and not line.startswith('## '):
            heading = line[2:].strip().lower()
            if heading == title_norm or heading.startswith(title_norm[:40]):
                removed = True
                continue
        out.append(line)
    return '\n'.join(out)


def fix_h3_prefix(body: str) -> str:
    return H3_PREFIX.sub(r'\1 ', body)


def fix_bold_spaces(body: str) -> str:
    # **слово** (no surrounding spaces) -> ** слово **
    def repl(m: re.Match) -> str:
        inner = m.group(1)
        # already has surrounding spaces? leave alone
        if inner.startswith(' ') or inner.endswith(' '):
            return m.group(0)
        return f'** {inner} **'

    # Run multiple times in case of nested patterns
    for _ in range(2):
        new = BOLD_NO_SPACES.sub(repl, body)
        if new == body:
            break
        body = new
    return body


def fix_faq_block(body: str) -> str:
    """Replace the giant one-line H2-FAQ with a proper H2 + per-question block."""
    def repl(m: re.Match) -> str:
        h2 = m.group(1).rstrip()  # `## FAQ` or `## FAQ `
        content = m.group(2).strip()

        # Split by `**...?**` markers
        # Each Q ends with `?` and the answer is the text up to the next **
        parts = re.split(r'(\*\*[^*]+\?\*\*)', content)
        # parts: [pre, Q1, A1, Q2, A2, ...]
        out = [h2, '']
        i = 1
        while i < len(parts):
            q = parts[i].strip()
            a = parts[i + 1].strip() if i + 1 < len(parts) else ''
            out.append(q)
            if a:
                out.append('')
                out.append(a)
            out.append('')
            i += 2
        return '\n'.join(out).rstrip() + '\n\n'

    return FAQ_RE.sub(repl, body)


def process_file(path: Path) -> bool:
    raw = path.read_text(encoding='utf-8')
    fm, body = split_frontmatter(raw)
    if not fm:
        return False

    # extract title
    title_match = re.search(r'^title:\s*["\']?(.+?)["\']?\s*$', fm, re.MULTILINE)
    title = title_match.group(1) if title_match else ''

    original = body
    body = remove_first_h1(body, title)
    body = fix_h3_prefix(body)
    body = fix_bold_spaces(body)
    body = fix_faq_block(body)

    if body == original:
        return False

    new_text = join_frontmatter(fm, body)
    path.write_text(new_text, encoding='utf-8')
    return True


def main():
    files = sorted(CONTENT_DIR.glob('*.md'))
    changed = 0
    for f in files:
        try:
            if process_file(f):
                changed += 1
                print(f'fixed: {f.name}')
            else:
                print(f'unchanged: {f.name}')
        except Exception as e:
            print(f'ERROR {f.name}: {e}')
    print(f'\nTotal changed: {changed} / {len(files)}')


if __name__ == '__main__':
    main()