compile: static site generator

1 files changed, 225 insertions, 0 deletions

diff --git a/bin/compile b/bin/compile
new file mode 100755
index 0000000..6c6d067
--- /dev/null
+++ b/bin/compile
@@ -0,0 +1,225 @@
+#!/usr/bin/env python3
+"""Build the site from flat Obsidian Markdown notes."""
+
+from __future__ import annotations
+
+import datetime as dt
+import html
+import re
+import shutil
+import sys
+from dataclasses import dataclass
+from pathlib import Path
+from urllib.parse import quote
+import markdown as md
+
+
+ROOT = Path(__file__).resolve().parent.parent
+POSTS = ROOT / "posts"
+PAGES = ROOT / "pages"
+BUILD = ROOT / "bld"
+
+
+@dataclass
+class Note:
+    path: Path
+    slug: str
+    title: str
+    display: str
+    order: int
+    link: str | None
+    date: dt.datetime | None
+    tags: list[str]
+    body: str
+
+
+def clean_tag(tag: str) -> str:
+    return tag.strip().lstrip("#").replace("/", "-").replace(" ", "-").lower()
+
+
+def parse_front_matter(text: str) -> tuple[dict[str, object], str]:
+    if not text.startswith("---\n"):
+        return {}, text
+    end = re.search(r"^---\s*$", text[4:], re.M)
+    if not end:
+        return {}, text
+    raw = text[4 : 4 + end.start()]
+    data: dict[str, object] = {}
+    for line in raw.splitlines():
+        if not line or line.lstrip().startswith("#") or ":" not in line:
+            continue
+        key, value = line.split(":", 1)
+        value = value.strip().strip("\"'")
+        if value.startswith("[") and value.endswith("]"):
+            data[key.strip().lower()] = [v.strip().strip("\"'") for v in value[1:-1].split(",") if v.strip()]
+        elif value:
+            data[key.strip().lower()] = value
+    return data, text[4 + end.end() :].lstrip("\n")
+
+
+def parse_date(value: object) -> dt.datetime | None:
+    if not value:
+        return None
+    text = str(value).strip().replace("Z", "+00:00")
+    for pattern in ("%Y-%m-%d %H:%M", "%Y-%m-%d", "%Y/%m/%d %H:%M", "%Y/%m/%d"):
+        try:
+            return dt.datetime.strptime(text, pattern)
+        except ValueError:
+            pass
+    try:
+        return dt.datetime.fromisoformat(text)
+    except ValueError:
+        raise ValueError(f"Invalid date: {value!r}. Use YYYY-MM-DD or YYYY-MM-DD HH:MM.")
+
+
+def parse_note(path: Path, post: bool = False) -> Note:
+    meta, body = parse_front_matter(path.read_text(encoding="utf-8"))
+    heading = re.search(r"^#\s+(.+?)\s*$", body, re.M)
+    title = str(meta.get("title") or (heading.group(1) if heading else path.stem.replace("-", " ")))
+    display = str(meta.get("display") or title)
+    order = int(meta.get("order") or 0)
+    link = str(meta.get("link")) if meta.get("link") else None
+    if heading and not meta.get("title"):
+        body = body[: heading.start()] + body[heading.end() :].lstrip("\n")
+    raw_tags = meta.get("tags", [])
+    if isinstance(raw_tags, str):
+        raw_tags = [piece.strip() for piece in raw_tags.split(",")]
+    tags = [clean_tag(str(tag)) for tag in raw_tags]
+    if post:
+        tags.extend(clean_tag(match) for match in re.findall(r"(?<![\w/])#([\w][\w/-]*)", body))
+    tags = list(dict.fromkeys(tag for tag in tags if tag))
+    return Note(path, path.stem, title, display, order, link, parse_date(meta.get("date")), tags, body)
+
+
+def inline(text: str) -> str:
+    text = html.escape(text, quote=False)
+    text = re.sub(r"!\[([^]]*)\]\(([^ )]+)(?:\s+[^)]*)?\)", r'<img src="\2" alt="\1">', text)
+    text = re.sub(r"\[([^]]+)\]\(([^ )]+)(?:\s+[^)]*)?\)", r'<a href="\2">\1</a>', text)
+    text = re.sub(r"\[\[([^]|]+)(?:\|([^]]+))?\]\]", lambda m: f'<a href="{quote(m.group(1).replace(" ", "-") + ".html")}">{m.group(2) or m.group(1)}</a>', text)
+    text = re.sub(r"`([^`]+)`", r"<code>\1</code>", text)
+    text = re.sub(r"\*\*([^*]+)\*\*|__([^_]+)__", lambda m: f"<strong>{m.group(1) or m.group(2)}</strong>", text)
+    text = re.sub(r"(?<!\*)\*([^*]+)\*(?!\*)|(?<!_)_([^_]+)_(?!_)", lambda m: f"<em>{m.group(1) or m.group(2)}</em>", text)
+    return text
+
+
+def markdown(text: str) -> str:
+    return md.markdown(text)
+    lines, output, paragraph, in_code, list_type = text.splitlines(), [], [], False, None
+
+    def flush_paragraph() -> None:
+        nonlocal paragraph
+        if paragraph:
+            output.append(f"<p>{inline(' '.join(part.strip() for part in paragraph))}</p>")
+            paragraph = []
+
+    def close_list() -> None:
+        nonlocal list_type
+        if list_type:
+            output.append(f"</{list_type}>")
+            list_type = None
+
+    for line in lines:
+        if line.startswith("```"):
+            flush_paragraph(); close_list()
+            output.append("</code></pre>" if in_code else "<pre><code>")
+            in_code = not in_code
+        elif in_code:
+            output.append(html.escape(line))
+        elif not line.strip():
+            flush_paragraph(); close_list()
+        elif match := re.match(r"^(#{1,6})\s+(.+)$", line):
+            flush_paragraph(); close_list()
+            level = len(match.group(1)); output.append(f"<h{level}>{inline(match.group(2))}</h{level}>")
+        elif line.startswith("> "):
+            flush_paragraph(); close_list(); output.append(f"<blockquote><p>{inline(line[2:])}</p></blockquote>")
+        elif line.strip() in ("---", "***", "___"):
+            flush_paragraph(); close_list(); output.append("<hr>")
+        elif match := re.match(r"^\s*([-+*]|\d+\.)\s+(.+)$", line):
+            flush_paragraph()
+            kind = "ol" if match.group(1)[0].isdigit() else "ul"
+            if list_type != kind:
+                close_list(); output.append(f"<{kind}>"); list_type = kind
+            output.append(f"<li>{inline(match.group(2))}</li>")
+        else:
+            paragraph.append(line)
+    flush_paragraph(); close_list()
+    return "\n".join(output)
+
+
+def tag_links(tags: list[str], prefix: str = "") -> str:
+    return " ".join(f'<a class="tag" href="{prefix}tags/{quote(tag)}.html">#{html.escape(tag)}</a>' for tag in tags)
+
+
+def page(title: str, content: str, nav: str = "") -> str:
+    return f"""<!doctype html>
+<html lang="en"><head><meta charset="utf-8"><meta name="viewport" content="width=device-width, initial-scale=1">
+<title>{html.escape(title)}</title><link rel="stylesheet" href="{nav}style.css"></head>
+<body><main>{content}</main></body></html>\n"""
+
+
+def post_html(note: Note) -> str:
+    date = note.date.strftime("%Y/%m/%d %H:%M") if note.date else "Undated"
+    return page(note.title, f'<a class="back" href="index.html">← index</a><article><header><h1>{html.escape(note.display)}</h1><p class="metadata">{date} &nbsp; {tag_links(note.tags)}</p></header>{markdown(note.body)}</article>')
+
+
+def list_html(posts: list[Note], heading: str, static_pages: list[Note], tag: str | None = None) -> str:
+    rows = []
+    for note in posts:
+        if tag is None or tag in note.tags:
+            date = note.date.strftime("%Y/%m/%d %H:%M") if note.date else "Undated"
+            prefix = "../" if tag else ""
+            tags = tag_links(note.tags, prefix)
+            rows.append(f'<li><time>{date}</time><a class="post-link" href="{prefix}{quote(note.slug)}.html">{html.escape(note.title)}</a><span class="tags">{tags}</span></li>')
+    extras = ""
+    if static_pages:
+        page_prefix = "../" if tag else ""
+        links = " · ".join(f'<a href="{note.link or f"{page_prefix}{quote(note.slug)}.html"}">{html.escape(note.title)}</a>' for note in static_pages)
+        extras = f'<nav class="pages">{links}</nav>'
+    back = '<a class="back" href="../index.html">← index</a>' if tag else ""
+    empty = "<p class=\"empty\">No posts yet.</p>" if not rows else f"<ol class=\"post-list\">{''.join(rows)}</ol>"
+    return page(heading, f'{back}<header class="list-header"><h1>{html.escape(heading)}</h1>{extras}</header>{empty}', "../" if tag else "")
+
+
+def copy_assets() -> None:
+    for source in POSTS.iterdir():
+        if not source.is_file() or source.suffix.lower() == ".md":
+            continue
+        shutil.copy2(source, BUILD / source.name)
+    for source in PAGES.iterdir():
+            if not source.is_file() or source.suffix.lower() == ".md":
+                continue
+            shutil.copy2(source, BUILD / source.name)
+
+
+def main() -> None:
+    for directory in (POSTS, PAGES):
+        directory.mkdir(exist_ok=True)
+    if BUILD.exists():
+        shutil.rmtree(BUILD)
+    (BUILD / "tags").mkdir(parents=True)
+    posts = [parse_note(path, True) for path in POSTS.glob("*.md")]
+    for note in posts:
+        if note.date is None:
+            raise ValueError(f"{note.path.relative_to(ROOT)} needs a date in its front matter")
+    posts.sort(key=lambda note: note.date or dt.datetime.min, reverse=True)
+    static_pages = [parse_note(path) for path in PAGES.glob("*.md")]
+    static_pages.sort(key=lambda note: note.order or 0)
+    for note in posts:
+        (BUILD / f"{note.slug}.html").write_text(post_html(note), encoding="utf-8")
+    for note in static_pages:
+        body = f'<a class="back" href="index.html">← index</a><article><h1>{html.escape(note.display)}</h1>{markdown(note.body)}</article>'
+        (BUILD / f"{note.slug}.html").write_text(page(note.title, body), encoding="utf-8")
+    (BUILD / "index.html").write_text(list_html(posts, "Alejandro's blog", static_pages), encoding="utf-8")
+    tags = sorted({tag for note in posts for tag in note.tags})
+    for tag in tags:
+        (BUILD / "tags" / f"{tag}.html").write_text(list_html(posts, f"#{tag}", static_pages, tag), encoding="utf-8")
+    shutil.copy2(ROOT / "style.css", BUILD / "style.css")
+    copy_assets()
+    print(f"Built {len(posts)} post(s), {len(tags)} tag page(s), and {len(static_pages)} page(s) in {BUILD.relative_to(ROOT)}/")
+
+
+if __name__ == "__main__":
+    try:
+        main()
+    except (OSError, ValueError) as error:
+        sys.exit(f"compile: {error}")