#!/usr/bin/env python3 """Build the kmalloc.org static site (plain HTML, no JavaScript, standard library only). Usage: build.py [output-dir] (default: preview) Layout: templates/ page chrome: base.html, header.html, footer.html, ... static/style.css the one stylesheet, inlined into every page content/site.toml values shared by every page (contact, bitcoin address, manifesto version) content//lang.toml the chrome strings of one language content//*.html pages: front matter, then an HTML fragment content//news/*.md news posts: YYYY-MM-DD-slug.md, front matter, then Markdown A page starts with a front matter block: --- id: people pairs the page with its translation path: /people.html where it is published nav: people menu entry highlighted on this page title: ... description: ... --- Placeholders {{name}} are replaced with values from site.toml, lang.toml and the page itself. An unknown placeholder stops the build.""" import html, os, re, shutil, sys, tomllib from email.utils import format_datetime from datetime import datetime, timezone from xml.sax.saxutils import escape as xml_escape HERE = os.path.dirname(os.path.abspath(__file__)) OUT = os.path.abspath(sys.argv[1] if len(sys.argv) > 1 else os.path.join(HERE, "preview")) LANGS = ("en", "it") NAV = ("home", "manifesto", "people", "services", "news") NEWS_ON_HOME = 3 NEWS_IN_FEED = 20 POST_NAME = re.compile(r"^(\d{4}-\d{2}-\d{2})-([a-z0-9]+(?:-[a-z0-9]+)*)$") # Files copied as they are: published name -> source. ASSETS = { "favicon.svg": "logo/svg/favicon-auto.svg", "favicon-auto.svg": "logo/svg/favicon-auto.svg", "logo-auto.svg": "logo/svg/logo-auto.svg", "favicon.ico": "logo/web/favicon.ico", "apple-touch-icon.png": "logo/web/apple-touch-icon.png", "fonts/terminess.woff2": "fonts/terminess.woff2", "badges/w3c-html5.png": "badges/w3c-html5.png", "badges/w3c-css.png": "badges/w3c-css.png", } def fail(msg): raise SystemExit(f"build.py: {msg}") def read(*parts): with open(os.path.join(HERE, *parts), encoding="utf-8") as f: return f.read() def write(rel, text): path = os.path.join(OUT, rel) os.makedirs(os.path.dirname(path), exist_ok=True) with open(path, "w", encoding="utf-8") as f: f.write(text) def toml(*parts): with open(os.path.join(HERE, *parts), "rb") as f: return tomllib.load(f) def parse(text, name): """Split 'front matter, then body' into (dict, body).""" m = re.match(r"---\n(.*?)\n---\n(.*)\Z", text, re.S) if not m: fail(f"{name}: missing front matter") meta = {} for line in m.group(1).splitlines(): key, sep, value = line.partition(": ") if not sep: fail(f"{name}: bad front matter line {line!r}") meta[key.strip()] = value.strip() return meta, m.group(2) def render(text, ctx, name): def sub(m): if m.group(1) not in ctx: fail(f"{name}: unknown placeholder {{{{{m.group(1)}}}}}") return ctx[m.group(1)] return re.sub(r"\{\{(\w+)\}\}", sub, text) # ---------------------------------------------------------------- Markdown def no_em_dash(md): """The manifesto sources use em dashes; the site does not: headings 'NN — Title' become 'NN. Title', 'Word — Title' becomes 'Word: Title', attribution lines lose the dash, parenthetical dashes become commas.""" md = re.sub(r"^(#+ )(\d\d) — ", r"\1\2. ", md, flags=re.M) md = re.sub(r"^(#+ [^\d\n][^—\n]*?) — ", r"\1: ", md, flags=re.M) md = re.sub(r"^> — ", "> ", md, flags=re.M) for a, b in ((" — where technically feasible — ", ", where technically feasible, "), (" — dove tecnicamente possibile — ", ", dove tecnicamente possibile, "), (" — and continues", ", and continues"), (" — e continua", ", e continua")): md = md.replace(a, b) md = md.replace("kmalloc.org — ", "kmalloc.org: ") assert "—" not in md, [l for l in md.splitlines() if "—" in l] return md def inline(s): s = html.escape(s, quote=False) s = re.sub(r"`([^`]+)`", r"\1", s) s = re.sub(r"\[([^\]]+)\]\(([^)\s]+)\)", r'\1', s) s = re.sub(r"\*\*(.+?)\*\*", r"\1", s) s = re.sub(r"(?\1", s) return s def slug(heading): m = re.match(r"(\d\d)\b", heading) if m: return m.group(1) h = heading.lower() if h.startswith(("preamble", "preambolo")): return "preamble" if h.startswith(("our commitments", "i nostri impegni")): return "commitments" if h.startswith(("epilogue", "epilogo")): return "epilogue" return re.sub(r"[^a-z0-9]+", "-", h).strip("-") def md_to_html(md): """Headings (##), paragraphs, quotes, bullet lists and inline markup. Returns (title, toc, html).""" out, toc = [], [] para, quote, items = [], [], [] def flush(): nonlocal para, quote, items if para: out.append("

\n" + "\n".join(para) + "\n

\n") para = [] if quote: out.append("
\n

\n" + "\n".join(quote) + "\n

\n
\n") quote = [] if items: out.append("
    \n" + "\n".join(f"
  • {i}
  • " for i in items) + "\n
\n") items = [] title = "" for raw in md.splitlines(): line = raw.rstrip("\n") br = line.endswith(" ") s = line.strip() if not s: flush() continue if s.startswith("# "): flush() title = s[2:] continue if s.startswith("## "): flush() h = s[3:] sl = slug(h) toc.append((sl, h)) out.append(f'

{inline(h)}

\n') continue if s.startswith("> "): if para or items: flush() quote.append(inline(s[2:]) + ("
" if br else "")) continue if s.startswith("- "): if para or quote: flush() items.append(inline(s[2:])) continue if quote or items: flush() para.append(inline(s)) flush() return title, toc, "".join(out) # ---------------------------------------------------------------- site model def load_pages(lang): """Pages of one language: static pages from content//*.html and posts from content//news/*.md.""" pages = [] folder = os.path.join(HERE, "content", lang) for fn in sorted(os.listdir(folder)): if fn.endswith(".html"): meta, body = parse(read("content", lang, fn), f"content/{lang}/{fn}") for key in ("id", "path", "nav", "title", "description"): if key not in meta: fail(f"content/{lang}/{fn}: front matter lacks '{key}'") if meta["nav"] not in NAV: fail(f"content/{lang}/{fn}: unknown nav '{meta['nav']}'") pages.append(dict(meta, lang=lang, kind="page", body=body, src=f"content/{lang}/{fn}")) news_dir = os.path.join(folder, "news") if os.path.isdir(news_dir): prefix = "" if lang == "en" else "/it" for fn in sorted(os.listdir(news_dir)): if not fn.endswith(".md"): continue m = POST_NAME.match(fn[:-3]) if not m: fail(f"content/{lang}/news/{fn}: name must be YYYY-MM-DD-slug.md") src = f"content/{lang}/news/{fn}" meta, body = parse(read("content", lang, "news", fn), src) if "title" not in meta: fail(f"{src}: front matter lacks 'title'") if "—" in body or "—" in meta["title"]: fail(f"{src}: no em dashes, please") pages.append(dict(meta, lang=lang, kind="post", id="news/" + fn[:-3], date=m.group(1), nav="news", path=f"{prefix}/news/{m.group(2)}.html", body=body, src=src, description=meta.get("summary", meta["title"]))) return pages def rss(lang, st, site, lang_pages): """RSS 2.0 feed of the news of one language: the latest posts, full text, absolute links.""" base = site["base_url"] posts = sorted((p for p in lang_pages if p["kind"] == "post"), key=lambda p: p["date"], reverse=True)[:NEWS_IN_FEED] def stamp(d): return format_datetime(datetime.strptime(d, "%Y-%m-%d").replace(tzinfo=timezone.utc)) items = [] for p in posts: text = md_to_html(p["body"])[2].replace('href="/', f'href="{base}/') items.append(f""" {xml_escape(p["title"])} {base}{p["path"]} {base}{p["path"]} {stamp(p["date"])} {xml_escape(text)} """) built = f"{stamp(posts[0]['date'])}\n" if posts else "" url = f"{base}{st['url_news']}" return f""" kmalloc.org: {xml_escape(st["nav_news"])} {url} {xml_escape(st["news_p"])} {lang} {built} {"".join(items)} """ def build(): css = read("static", "style.css").strip("\n") tpl = {n[:-5]: read("templates", n) for n in os.listdir(os.path.join(HERE, "templates")) if n.endswith(".html")} site = toml("content", "site.toml") site["contact_shown"] = site["contact"].replace("@", " (at) ") # no mailto link, against address harvesters strings = {lang: toml("content", lang, "lang.toml") for lang in LANGS} pages = {lang: load_pages(lang) for lang in LANGS} by_id = {(p["id"], p["lang"]): p for lang in LANGS for p in pages[lang]} # Manifesto: the user's own Markdown files, rendered once per language. manifesto = {} for lang in LANGS: title, toc, body = md_to_html(no_em_dash(read(f"MANIFESTO.{lang}.md"))) manifesto[lang] = dict( manifesto_title=html.escape(title), manifesto_toc="\n".join(f'
  • {inline(h)}
  • ' for sl, h in toc), manifesto_body=body) def news_lists(lang): posts = sorted((p for p in pages[lang] if p["kind"] == "post"), key=lambda p: p["date"], reverse=True) def li(p): line = f'
  • {html.escape(p["title"], quote=False)}' if p.get("summary"): line += ": " + html.escape(p["summary"], quote=False) return line + "
  • " items = "\n".join(li(p) for p in posts) full = f"
      \n{items}\n
    " if posts else f"

    \n{strings[lang]['news_empty']}\n

    " return full, items, posts[:NEWS_ON_HOME] news = {lang: news_lists(lang) for lang in LANGS} for lang in LANGS: other = "it" if lang == "en" else "en" st = strings[lang] for p in pages[lang]: name = p["src"] alt = by_id.get((p["id"], other)) ctx = dict(site, **st, **manifesto[lang]) page_title = p["title"] if p["kind"] == "page" else "kmalloc.org: " + p["title"] ctx.update(css=css, title=p["title"], description=p["description"], path=p["path"], title_html=html.escape(page_title), post_title=html.escape(p["title"]), description_attr=html.escape(p["description"], quote=True), doc_url=site["base_url"] + p["path"], date=p.get("date", "")) for k in NAV: ctx["cur_" + k] = ' aria-current="page"' if p["nav"] == k else "" ctx["alt_url"] = alt["path"] if alt else "" ctx["alt_switch"] = p.get("switch") or (alt["path"] if alt else strings[other]["url_news"]) ctx["alt_links"] = (f'\n' f'') if alt else "" full, items, latest = news[lang] ctx["news_list"] = full ctx["latest_news"] = render(tpl["latest-news"], dict(ctx, news_items="\n".join( f'
  • {html.escape(x["title"], quote=False)}
  • ' for x in latest)), "templates/latest-news.html") if latest else "" if p["kind"] == "post": _, _, post_html = md_to_html(p["body"]) ctx["post_body"] = post_html ctx["content"] = render(tpl["news-post"], ctx, "templates/news-post.html") lead = "news" else: ctx["content"] = render(p["body"].strip("\n") + "\n", ctx, name) lead = p.get("lead", "back") ctx["lead"] = render(tpl["lead-" + lead], ctx, f"templates/lead-{lead}.html") ctx["badges"] = render(tpl["badges"], dict(ctx, doc_q=ctx["doc_url"].replace(":", "%3A").replace("/", "%2F")), "templates/badges.html") ctx["header"] = render(tpl["header"], ctx, "templates/header.html") ctx["footer"] = render(tpl["footer"], ctx, "templates/footer.html") rel = p["path"].lstrip("/") write(rel + "index.html" if rel == "" or rel.endswith("/") else rel, render(tpl["base"], ctx, "templates/base.html")) for lang in LANGS: write(strings[lang]["url_news"].lstrip("/") + "feed.xml", rss(lang, strings[lang], site, pages[lang])) for dest, src in ASSETS.items(): target = os.path.join(OUT, dest) os.makedirs(os.path.dirname(target), exist_ok=True) shutil.copy(os.path.join(HERE, src), target) write("robots.txt", "User-agent: *\nDisallow: /\n") count = sum(len(v) for v in pages.values()) print(f"built {count} pages into {OUT}") build()