#!/usr/bin/env python3
"""Build the blog: every topics/**/*.md becomes a section in one index.html.

Each post may start with a small header:

    ---
    title: My first post
    date: 2026-10-03
    ---

Without it, the title comes from the first "# Heading" (or the file name)
and the date from a YYYY-MM-DD file-name prefix, if any. A post in a
subfolder (topics/python/foo.md) is tagged with that folder's name.

Usage: python3 build.py [source_dir] [output_dir]
"""
import html
import re
import shutil
import sys
from datetime import date
from pathlib import Path

import markdown

SRC = Path(sys.argv[1] if len(sys.argv) > 1 else ".").resolve()
OUT = Path(sys.argv[2] if len(sys.argv) > 2 else "public").resolve()
TOPICS = SRC / "topics"


def slugify(text):
    return re.sub(r"[^a-z0-9]+", "-", text.lower()).strip("-") or "post"


def parse(path):
    text = path.read_text(encoding="utf-8")
    meta = {}
    m = re.match(r"^---\s*\n(.*?)\n---\s*\n", text, re.S)
    if m:
        for line in m.group(1).splitlines():
            key, sep, value = line.partition(":")
            if sep:
                meta[key.strip().lower()] = value.strip().strip("\"'")
        text = text[m.end():]

    title = meta.get("title")
    if not title:
        h1 = re.match(r"^\s*#\s+(.+)\n", text)
        if h1:
            title = h1.group(1).strip()
            text = text[h1.end():]  # don't render the title twice
        else:
            title = re.sub(r"^\d{4}-\d{2}-\d{2}-", "", path.stem).replace("-", " ").capitalize()

    raw_date = meta.get("date") or (re.match(r"^\d{4}-\d{2}-\d{2}", path.stem) or [None])[0]
    try:
        when = date.fromisoformat(raw_date) if raw_date else None
    except ValueError:
        sys.exit(f"{path.relative_to(SRC)}: date '{raw_date}' is not YYYY-MM-DD")

    rel = path.relative_to(TOPICS)
    return {
        "title": title,
        "date": when,
        "topic": rel.parts[0] if len(rel.parts) > 1 else meta.get("topic"),
        "slug": slugify(meta.get("slug") or re.sub(r"^\d{4}-\d{2}-\d{2}-", "", path.stem)),
        "body": markdown.markdown(text, extensions=["fenced_code", "tables", "footnotes"]),
    }


def main():
    posts = [parse(p) for p in sorted(TOPICS.rglob("*.md"))]
    # Newest first; undated posts go last.
    posts.sort(key=lambda p: (p["date"] is not None, p["date"] or date.min), reverse=True)

    seen = {}
    for p in posts:  # keep anchors unique
        n = seen.get(p["slug"], 0)
        seen[p["slug"]] = n + 1
        if n:
            p["slug"] += f"-{n + 1}"

    toc, sections = [], []
    for p in posts:
        title = html.escape(p["title"])
        when = f'<time datetime="{p["date"].isoformat()}">{p["date"]:%d %B %Y}</time>' if p["date"] else ""
        tag = f'<span class="tag">{html.escape(p["topic"])}</span>' if p["topic"] else ""
        toc.append(f'<li><a href="#{p["slug"]}">{title}</a>{when}</li>')
        sections.append(
            f'<article id="{p["slug"]}">\n'
            f'<header><h2><a href="#{p["slug"]}">{title}</a></h2><p class="meta">{when}{tag}</p></header>\n'
            f'{p["body"]}\n</article>'
        )

    page = (SRC / "template.html").read_text(encoding="utf-8")
    page = page.replace("{{toc}}", "\n".join(toc) or "<li>No posts yet.</li>")
    page = page.replace("{{posts}}", "\n\n".join(sections))
    page = page.replace("{{year}}", str(date.today().year))

    if OUT.exists():
        shutil.rmtree(OUT)
    OUT.mkdir(parents=True)
    (OUT / "index.html").write_text(page, encoding="utf-8")
    if (SRC / "static").is_dir():  # images etc.: topics can link to static/foo.png
        shutil.copytree(SRC / "static", OUT / "static")
    print(f"Built {len(posts)} post(s) into {OUT}")


if __name__ == "__main__":
    main()