Commit `195e1d0ff1`

195e1d0ff130ff53d7915832e8403f123a3ed629

parent: 5d60a70d6a

Unsigned

cmc <hello@cleberg.net> · 2026-08-12 00:16 UTC

Build with orgo

The generator changes; the steps around it that orgo does not do stay. Four of
them go, because it does:

- the blog index groups itself by year (a template, using page.year)
- the tags page is a collection
- the recent-posts list on the home page comes from the blog collection
- the sitemap is written once site.base_url is set

What remains is minifying CSS, copying the org sources for readers who want
them, and rewriting img.cleberg.net to /img/ so the onion serves images from its
own origin. That last one is why this script still exists.

Verified against the live site before deploying: 195 pages, 186 identical in
content, and every URL in the live sitemap still produced. The nine that differ
are the ones we chose — the reply footer is opt-in now, /tags/ has a page per
tag, and footnotes carry aria labels.

publish.el and theme/templates/ are left alone. They are weblorg's, and there is
no reason to delete a fallback on the same day the replacement goes live.

Layout: unified · split

build.py +18 −216

		@@ -32,7 +32,6 @@ import sys
32	32	from datetime import datetime
33	33	from html import escape
34	34	from pathlib import Path
35		from urllib.parse import quote
36	35
37	36	SITE_TEMPLATE_VARS = {
38	37	"site_name": "cleberg.net",
		@@ -51,66 +50,6 @@ def run_ruff():
51	50	sys.exit(1)
52	51
53	52
54		def update_marked_section(
55		html_snippet,
56		template_path="./.build/index.html",
57		begin_marker="<!-- BEGIN_POSTS -->",
58		end_marker="<!-- END_POSTS -->",
59		):
60		"""
61		Read the file at `template_path`, replace everything between `begin_marker`
62		and `end_marker` with the provided html_snippet, and write the updated
63		content back to the same file.
64		"""
65		with open(template_path, "r", encoding="utf-8") as f:
66		content = f.read()
67
68		# Find the indices of the markers
69		begin_index = content.find(begin_marker)
70		end_index = content.find(end_marker)
71
72		if begin_index == -1 or end_index == -1:
73		raise ValueError(f"Markers not found in {template_path}")
74
75		# Compute insertion points: after the end of begin_marker line, before end_marker
76		# Include the newline after BEGIN_POSTS
77		insert_start = begin_index + len(begin_marker)
78		# Ensure we capture the newline character if present
79		if content[insert_start : insert_start + 1] == "\n":
80		insert_start += 1
81
82		# If there is a newline before END_POSTS, trim trailing whitespace from snippet block
83		# We will preserve indentation of BEGIN_POSTS line
84		indent = ""
85		# Determine the indentation by looking at characters after the newline that follows begin_marker
86		lines_after_begin = content[begin_index:].splitlines(True)
87		if len(lines_after_begin) > 1:
88		# The second line starts with the indentation to preserve
89		second_line = lines_after_begin[1]
90		indent = ""
91		for ch in second_line:
92		if ch.isspace():
93		indent += ch
94		else:
95		break
96
97		# Prepare the replacement block: indent each line of html_snippet
98		snippet_lines = html_snippet.splitlines()
99		indented_snippet = "\n".join(indent + line for line in snippet_lines) + "\n"
100
101		# Compute the position just before end_marker (excluding any preceding whitespace/newline)
102		end_line_start = content.rfind("\n", 0, end_index)
103		if end_line_start == -1:
104		end_line_start = end_index
105
106		# Construct the new content
107		new_content = content[:insert_start] + indented_snippet + content[end_line_start:]
108
109		# Write back to index.html
110		with open(template_path, "w", encoding="utf-8") as f:
111		f.write(new_content)
112
113
114	53	def render_base_template(main_html, subtitle="", title=None):
115	54	"""
116	55	Render a small subset of the site's shared templates for Python-generated
		@@ -257,24 +196,6 @@ def get_blog_posts(content_dir="./content/blog"):
257	196	return posts
258	197
259	198
260		def get_recent_posts_html(content_dir="./content/blog", num_posts=3):
261		"""
262		Return an HTML snippet for the `num_posts` most recent blog posts.
263		"""
264		recent = get_blog_posts(content_dir)[:num_posts]
265
266		lines = []
267		for post in recent:
268		lines.append('\t<li class="post-list-item">')
269		lines.append(
270		f'\t\t<time datetime="{post["date_str"]}">{post["date_full"]}</time>'
271		)
272		lines.append(f'\t\t<a href="/blog/{post["slug"]}.html">{post["title"]}</a>')
273		lines.append("\t</li>")
274
275		return "\n".join(lines)
276
277
278	199	def prompt(prompt_text):
279	200	try:
280	201	return input(prompt_text).strip()
		@@ -303,20 +224,6 @@ def minify_css(src_css, dest_css):
303	224	sys.exit(1)
304	225
305	226
306		def minify_html(src_html, dest_html):
307		print(f"Minifying HTML: {src_html} → {dest_html}")
308		result = subprocess.run(
309		["minify", "-o", str(dest_html), str(src_html)],
310		capture_output=True,
311		text=True,
312		check=False,
313		)
314		if result.returncode != 0:
315		print("Error during HTML minification:")
316		print(result.stderr, file=sys.stderr)
317		sys.exit(1)
318
319
320	227	def rewrite_img_urls(build_dir=".build"):
321	228	"""
322	229	Rewrite absolute img.cleberg.net URLs to root-relative /img/ paths so the
		@@ -342,31 +249,32 @@ def rewrite_img_urls(build_dir=".build"):
342	249	print(f"Rewrote {count} img.cleberg.net references to /img/")
343	250
344	251
345		def run_emacs_publish(dev_mode=True):
346		mode = "development" if dev_mode else "production"
347		print(f"Running Emacs publish script ({mode})...")
	252	def run_orgo_build(dev_mode=True):
	253	"""
	254	Build the site with orgo.
	255
	256	Replaces the weblorg/Emacs publish. orgo reads content/orgo.toml, so the routes,
	257	templates and collections that used to live in publish.el live there now. Four of the
	258	steps this script used to perform afterwards are gone with it: the blog index groups
	259	itself by year, the tags page is a collection, the recent-posts list is generated from
	260	the blog collection, and the sitemap is written by orgo once base_url is set.
348	261
	262	What is left around it is what orgo does not do: minifying CSS, copying the org
	263	sources for readers who want them, and rewriting image URLs for the onion.
	264	"""
	265	print("Building with orgo...")
349	266	result = subprocess.run(
350		["emacs", "--script", "publish.el"],
	267	["orgo", "build", "content", "-o", ".build", "--strict"],
351	268	stdout=subprocess.PIPE,
352	269	stderr=subprocess.STDOUT,
353	270	text=True,
354	271	check=False,
355	272	)
356
	273	print(result.stdout, end="")
357	274	if result.returncode != 0:
358		print("Error running publish.el output:")
359		print(result.stdout)
	275	print("orgo build failed", file=sys.stderr)
360	276	sys.exit(1)
361	277
362		annoying_file = Path(".build/cleberg-net.html")
363		if annoying_file.exists():
364		os.remove(annoying_file)
365		else:
366		print(
367		"Warning: .build/cleberg-net.html not found, but Emacs exited successfully."
368		)
369
370	278
371	279	def copy_org_sources(content_dir="./content", build_dir="./.build/org"):
372	280	print(f"Copying org sources: {content_dir} → {build_dir}")
		@@ -395,90 +303,6 @@ def copy_org_sources(content_dir="./content", build_dir="./.build/org"):
395	303	print(f" {src_path.name} → {dest_name}")
396	304
397	305
398		def generate_sitemap(build_dir=".build", base_url="https://cleberg.net"):
399		"""
400		Generates a sitemap.xml based on contents of the .build directory.
401		Only includes .html files (except 404.html).
402		"""
403		sitemap_entries = []
404		for root, dirs, files in os.walk(build_dir):
405		for filename in files:
406		if filename.endswith(".html") and filename != "404.html":
407		full_path = os.path.join(root, filename)
408		rel_path = os.path.relpath(full_path, build_dir)
409		url_path = "/" + quote(rel_path.replace(os.sep, "/"))
410		# Remove index.html for cleaner URLs
411		if url_path.endswith("/index.html"):
412		url_path = url_path[:-10] or "/"
413		elif url_path == "/index.html":
414		url_path = "/"
415		loc = f"{base_url}{url_path}"
416
417		# Last modified time
418		lastmod = datetime.fromtimestamp(os.path.getmtime(full_path)).strftime(
419		"%Y-%m-%d"
420		)
421
422		sitemap_entries.append(f""" <url>
423		<loc>{loc}</loc>
424		<lastmod>{lastmod}</lastmod>
425		</url>""")
426
427		sitemap_xml = f"""<?xml version="1.0" encoding="UTF-8"?>
428		<urlset xmlns="http://www.sitemaps.org/schemas/sitemap/0.9">
429		{os.linesep.join(sitemap_entries)}
430		</urlset>
431		"""
432		# Write to .build/sitemap.xml
433		sitemap_path = os.path.join(build_dir, "sitemap.xml")
434		with open(sitemap_path, "w", encoding="utf-8") as f:
435		f.write(sitemap_xml)
436		print(f"Sitemap generated at {sitemap_path} with {len(sitemap_entries)} entries.")
437
438
439		def inject_blog_year_separators(blog_index_path="./.build/blog/index.html"):
440		"""
441		Post-processes the rendered blog index to inject year separator <li> elements
442		between groups of posts. Weblorg/templatel doesn't support mutable loop state,
443		so this runs after the HTML is generated.
444
445		Finds each <li class="post-list-item"> that contains a <time datetime="YYYY-MM-DD">,
446		and inserts <li class="post-list-year">YYYY</li> before the first post of each year.
447		"""
448		path = Path(blog_index_path)
449		if not path.exists():
450		print(f"Warning: {blog_index_path} not found, skipping year separators.")
451		return
452
453		content = path.read_text(encoding="utf-8")
454
455		# Match each post list item, capturing the date and the full element
456		item_pattern = re.compile(
457		r'(<li class="post-list-item">.*?</li>)',
458		re.DOTALL,
459		)
460		date_pattern = re.compile(r"datetime=['\"]?(\d{4})-\d{2}-\d{2}['\"]?")
461
462		current_year = None
463
464		def replace_item(m):
465		nonlocal current_year
466		item_html = m.group(1)
467		date_match = date_pattern.search(item_html)
468		if not date_match:
469		return item_html
470		year = date_match.group(1)
471		if year != current_year:
472		current_year = year
473		separator = f'<li class="post-list-year">{year}</li>'
474		return f"{separator}\n{item_html}"
475		return item_html
476
477		new_content = item_pattern.sub(replace_item, content)
478		path.write_text(new_content, encoding="utf-8")
479		print(f"Blog year separators injected into {blog_index_path}")
480
481
482	306	def get_tags_html(content_dir="./content/blog"):
483	307	"""
484	308	Build the tag index HTML snippet for the rendered tags template.
		@@ -541,17 +365,6 @@ def get_tags_html(content_dir="./content/blog"):
541	365	)
542	366
543	367
544		def generate_tags_page(content_dir="./content/blog", build_dir="./.build"):
545		"""
546		Render the tags page using the shared template structure.
547		"""
548		tags_html = get_tags_html(content_dir)
549		out_path = Path(build_dir) / "tags" / "index.html"
550		out_path.parent.mkdir(parents=True, exist_ok=True)
551		out_path.write_text(render_tags_page_html(tags_html), encoding="utf-8")
552		print(f"Tags page written to {out_path}")
553
554
555	368	def deploy_to_server(build_dir, server):
556	369	remote_path = f"{server}:/var/www/cleberg.net/"
557	370	print(f"Deploying .build/ → {remote_path}")
		@@ -584,7 +397,6 @@ def main():
584	397	env = os.environ.get("ENV", "").casefold()
585	398	if env != "prod":
586	399	run_ruff()
587		html_snippet = get_recent_posts_html("./content/blog", num_posts=3)
588	400
589	401	build_dir = Path(".build")
590	402	theme_dir = Path("theme/static")
		@@ -599,14 +411,9 @@ def main():
599	411	if build:
600	412	remove_build_directory(build_dir)
601	413	minify_css(css_src, css_min)
602		run_emacs_publish(dev_mode=False)
	414	run_orgo_build(dev_mode=False)
603	415	copy_org_sources()
604		update_marked_section(html_snippet)
605		inject_blog_year_separators()
606		generate_tags_page()
607	416	rewrite_img_urls(build_dir)
608		# minify_html("./.build/index.html", "./.build/index.html")
609		generate_sitemap()
610	417	if deploy:
611	418	print("Deploying to production...")
612	419	deploy_to_server(build_dir, "homelab")
		@@ -616,13 +423,8 @@ def main():
616	423	if build:
617	424	remove_build_directory(build_dir)
618	425	minify_css(css_src, css_min)
619		run_emacs_publish(dev_mode=True)
	426	run_orgo_build(dev_mode=True)
620	427	copy_org_sources()
621		update_marked_section(html_snippet)
622		inject_blog_year_separators()
623		generate_tags_page()
624		minify_html("./.build/index.html", "./.build/index.html")
625		generate_sitemap()
626	428	if deploy:
627	429	start_dev_server(build_dir)
628	430

Commit 195e1d0ff1

Commit `195e1d0ff1`