mobea/generator/patch_head_meta.py

#!/usr/bin/env python3
"""Inject favicon and social meta tags into existing static HTML pages."""

from __future__ import annotations

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

from generate import build_archiv_entries, log, render_template, write_text_atomic


def extract_meta(html: str, prop: str) -> str | None:
    match = re.search(rf'property="{re.escape(prop)}" content="([^"]*)"', html)
    return match.group(1) if match else None


def extract_name_meta(html: str, name: str) -> str | None:
    match = re.search(rf'name="{re.escape(name)}" content="([^"]*)"', html)
    return match.group(1) if match else None


def extract_hero_image(html: str) -> str | None:
    match = re.search(r'<img src="/img/(\d{4}-\d{2}-\d{2})(?:-\d+w)?\.jpg"', html)
    return f"/img/{match.group(1)}.jpg" if match else None


def render_head_block(
    *,
    site_url: str,
    canonical_path: str,
    og_title: str,
    og_description: str,
    og_image: str | None,
) -> str:
    return render_template(
        "_head_common.html.j2",
        {
            "site_url": site_url,
            "canonical_path": canonical_path,
            "og_title": og_title,
            "og_description": og_description,
            "og_image": og_image,
        },
    ).rstrip()


def inject_head_block(html: str, block: str) -> str:
    if 'rel="icon"' in html:
        return html
    marker = '<meta property="og:description"'
    idx = html.find(marker)
    if idx == -1:
        marker = '<meta name="description"'
        idx = html.find(marker)
    if idx == -1:
        raise ValueError("could not find description meta tag")
    line_end = html.find("\n", idx)
    if line_end == -1:
        raise ValueError("unexpected HTML structure")
    return html[: line_end + 1] + block + "\n" + html[line_end + 1 :]


def patch_startup_page(path: Path, canonical_path: str, site_url: str) -> bool:
    html = path.read_text(encoding="utf-8")
    og_title = extract_meta(html, "og:title") or extract_name_meta(html, "twitter:title")
    og_description = extract_meta(html, "og:description") or extract_name_meta(html, "description")
    if not og_title or not og_description:
        log(f"warning: skipping {path}, missing title/description")
        return False
    block = render_head_block(
        site_url=site_url,
        canonical_path=canonical_path,
        og_title=og_title,
        og_description=og_description,
        og_image=extract_hero_image(html),
    )
    updated = inject_head_block(html, block)
    if updated != html:
        write_text_atomic(path, updated)
        log(f"patched {path}")
        return True
    return False


def patch_archiv_index(path: Path, site_url: str, og_image: str) -> bool:
    html = path.read_text(encoding="utf-8")
    og_title = "MOBEA Archiv – Unsere bisherigen Pivots"
    og_description = "Jeden Tag ein neues Startup. Hier ruhen die bisherigen."
    if 'property="og:title"' not in html:
        html = html.replace(
            '<meta name="description" content="Jeden Tag ein neues Startup. Hier ruhen die bisherigen.">',
            '<meta name="description" content="Jeden Tag ein neues Startup. Hier ruhen die bisherigen.">\n'
            f'  <meta property="og:title" content="{og_title}">\n'
            f'  <meta property="og:description" content="{og_description}">',
            1,
        )
    block = render_head_block(
        site_url=site_url,
        canonical_path="/archiv/",
        og_title=og_title,
        og_description=og_description,
        og_image=og_image,
    )
    updated = inject_head_block(html, block)
    if updated != html:
        write_text_atomic(path, updated)
        log(f"patched {path}")
        return True
    return False


def main() -> int:
    repo_root = Path(__file__).resolve().parent.parent
    site_dir = Path(os.environ.get("SITE_DIR", repo_root / "public"))
    data_dir = Path(os.environ.get("DATA_DIR", repo_root / "data"))
    site_url = os.environ.get("SITE_URL", "https://mobea.de")

    archive = json.loads((data_dir / "archive.json").read_text(encoding="utf-8"))
    latest_image = next(
        (
            f"/img/{entry['date']}.jpg"
            for entry in sorted(archive, key=lambda e: e["date"], reverse=True)
            if entry.get("has_image")
        ),
        "/og-default.jpg",
    )

    patched = 0
    index_path = site_dir / "index.html"
    if index_path.exists():
        patched += patch_startup_page(index_path, "/", site_url)

    archiv_dir = site_dir / "archiv"
    for path in sorted(archiv_dir.glob("2026-*.html")):
        patched += patch_startup_page(path, f"/archiv/{path.name}", site_url)

    archiv_index = archiv_dir / "index.html"
    if archiv_index.exists():
        patched += patch_archiv_index(archiv_index, site_url, latest_image)

    log(f"done, patched {patched} file(s)")
    return 0


if __name__ == "__main__":
    sys.exit(main())