feat(docs): generate Starlight content from the manual tree

Add build_manual.build_site(), which walks docs/manual and emits the
Astro Starlight content collection: plain pages copied with pipeline-only
frontmatter (man/site/manTitle/helpKeywords) stripped, and *-functions/
category files exploded into one page per function plus a CardGrid/LinkCard
overview. Writes src/sidebar.json for astro.config.mjs to import. Wires
--site alongside the existing --concat flag.

Fixes two latent defects found while building the real site against the
strict content.config.ts schema:
- _split_entries now tracks fenced code blocks (like
  manualtools.shift_headings does) so a `## ` inside a fence can't be
  mistaken for an entry boundary.
- LinkCard title/description are escaped for JSX attribute context, since
  shell synopses routinely contain `<placeholder>` angle brackets that
  would otherwise open unterminated MDX/JSX parsing.

Also fixes the generated sidebar shape for the functions category: Starlight
0.39+ dropped support for a bare `autogenerate` sibling of `label` on a
top-level group, so the autogenerate config now nests inside `items`.

Verified with a full `astro build` (temporarily pointing astro.config.mjs
at the generated sidebar.json, then reverted since replacing that config
is a later task's deliverable): 120 pages built cleanly, no
content-collection/frontmatter/MDX errors.
This commit is contained in:
2026-07-25 22:44:28 -04:00
parent c096433b5d
commit a4054de4f6
2 changed files with 189 additions and 8 deletions
+168 -8
View File
@@ -8,6 +8,9 @@
"""
import argparse
import json
import re
import shutil
import sys
from pathlib import Path
@@ -45,21 +48,178 @@ def build_concat(root: Path) -> str:
return "\n\n".join(chunks) + "\n"
SENTENCE_RE = re.compile(r"^(.+?[.!?])(\s|$)", re.S)
PIPELINE_KEYS = ("man", "site", "manTitle", "helpKeywords")
JSX_ATTR_ESCAPES = (
("&", "&amp;"),
('"', "&quot;"),
("<", "&lt;"),
("{", "&#123;"),
)
def _jsx_attr_escape(value: str) -> str:
"""Escape a string for safe use inside a quoted JSX attribute value.
`&` must go first so escaping later characters doesn't double-escape
the ampersands it introduces. `"` closes the attribute early; `<` and
`{` are otherwise-live MDX/JSX syntax that must not be interpreted.
"""
for char, escape in JSX_ATTR_ESCAPES:
value = value.replace(char, escape)
return value
def _first_sentence(body: str) -> str:
"""Extract a one-line description from the start of an entry body."""
for line in body.split("\n"):
line = line.strip()
if not line or line.startswith(("#", "```", "|", "-", "*", ">")):
continue
m = SENTENCE_RE.match(line)
return (m.group(1) if m else line)[:160]
return ""
def _page_fm(fm: dict) -> dict:
"""Strip pipeline-only keys from frontmatter destined for the site."""
return {k: v for k, v in fm.items() if k not in PIPELINE_KEYS}
def _split_entries(body: str) -> tuple[str, list[tuple[str, str]]]:
"""Split a category body into (intro, [(entry title, entry body)]).
Fence-aware: an H2-looking line (`## ...`) inside a fenced code block
(tracked the same way as `manualtools.shift_headings`) is treated as
ordinary body text, not an entry boundary.
"""
lines = body.split("\n")
heading_re = re.compile(r"^## (.+)$")
boundaries: list[tuple[int, str]] = []
in_fence = False
for i, line in enumerate(lines):
if mt.FENCE_RE.match(line):
in_fence = not in_fence
continue
if not in_fence:
m = heading_re.match(line)
if m:
boundaries.append((i, m.group(1)))
if not boundaries:
return body.strip(), []
intro = "\n".join(lines[: boundaries[0][0]]).strip()
entries = []
for idx, (line_no, title) in enumerate(boundaries):
start = line_no + 1
end = boundaries[idx + 1][0] if idx + 1 < len(boundaries) else len(lines)
entry_body = "\n".join(lines[start:end]).strip()
entries.append((title.strip(), entry_body))
return intro, entries
def build_site(root: Path, out: Path) -> list[dict]:
"""Write the Starlight content tree. Returns the sidebar structure."""
if out.exists():
shutil.rmtree(out)
out.mkdir(parents=True)
sidebar: list[dict] = []
for path, _depth in mt.walk(root):
fm, body = mt.parse(path)
if not fm.get("site", True):
continue
rel = path.relative_to(root)
is_function_dir = rel.parts and rel.parts[0].endswith("-functions")
if not is_function_dir:
target = out / rel
target.parent.mkdir(parents=True, exist_ok=True)
target.write_text(mt.serialize(_page_fm(fm), body))
if rel.name != "index.md":
sidebar.append({"label": fm["title"], "link": "/" + rel.stem + "/"})
continue
# Section 5: category index page keeps its slot; entries explode.
slug_dir = "functions"
if rel.name == "index.md":
target = out / slug_dir / "index.md"
target.parent.mkdir(parents=True, exist_ok=True)
target.write_text(mt.serialize(_page_fm(fm), body))
sidebar.append(
{
"label": fm["title"],
"collapsed": True,
# Starlight >=0.39 rejects a bare `autogenerate` sibling
# of `label` on a top-level group (removed in v0.39.0);
# the autogenerate config must be nested inside `items`.
"items": [{"autogenerate": {"directory": slug_dir}}],
}
)
continue
category = re.sub(r"^\d+-", "", rel.stem)
cat_dir = out / slug_dir / category
cat_dir.mkdir(parents=True, exist_ok=True)
intro, entries = _split_entries(body)
cards = []
for title, entry_body in entries:
entry_slug = re.sub(r"[^\w-]+", "-", title.strip().lower()).strip("-")
desc = _first_sentence(entry_body)
entry_fm = {"title": title}
if desc:
entry_fm["description"] = desc
(cat_dir / f"{entry_slug}.md").write_text(
mt.serialize(entry_fm, entry_body)
)
href = f"/{slug_dir}/{category}/{entry_slug}/"
safe_title = _jsx_attr_escape(title)
safe_desc = _jsx_attr_escape(desc)
cards.append(
f' <LinkCard title="{safe_title}" href="{href}"'
+ (f' description="{safe_desc}"' if desc else "")
+ " />"
)
overview = (
"import { CardGrid, LinkCard } from '@astrojs/starlight/components';\n\n"
+ (f"{intro}\n\n" if intro else "")
+ "<CardGrid>\n"
+ "\n".join(cards)
+ "\n</CardGrid>\n"
)
(cat_dir / "index.mdx").write_text(mt.serialize(_page_fm(fm), overview))
return sidebar
def main() -> int:
ap = argparse.ArgumentParser(description=__doc__)
ap.add_argument("--concat", action="store_true", help="emit the pandoc document")
ap.add_argument("--site", action="store_true", help="emit the Starlight content tree")
ap.add_argument("-o", "--output", type=Path, help="write to PATH instead of stdout")
args = ap.parse_args()
if not args.concat:
ap.error("nothing to do: pass --concat")
if not (args.concat or args.site):
ap.error("nothing to do: pass --concat and/or --site")
text = build_concat(MANUAL)
if args.output:
args.output.write_text(text)
print(f"wrote {args.output}")
else:
sys.stdout.write(text)
if args.site:
src = DOCS / "site" / "src"
out = src / "content" / "docs"
sidebar = build_site(MANUAL, out)
(src / "sidebar.json").write_text(json.dumps(sidebar, indent=2) + "\n")
print(f"wrote site content to {out} ({len(sidebar)} sidebar entries)")
if args.concat:
text = build_concat(MANUAL)
if args.output:
args.output.write_text(text)
print(f"wrote {args.output}")
else:
sys.stdout.write(text)
return 0
+21
View File
@@ -192,6 +192,27 @@ def test_every_index_keyword_resolves():
assert not missing, "unresolvable index keywords:\n " + "\n ".join(missing)
def test_site_build_produces_function_pages():
import tempfile
import build_manual
docs = Path(__file__).parent
with tempfile.TemporaryDirectory() as d:
out = Path(d)
sidebar = build_manual.build_site(docs / "manual", out)
pages = list(out.rglob("*.md*"))
assert (out / "index.md").exists(), "landing page missing"
fn_pages = [p for p in pages if "functions" in p.parts and p.name != "index.mdx"]
assert len(fn_pages) > 80, f"expected >80 function pages, got {len(fn_pages)}"
assert not (out / "00-name.md").exists(), "site:false page was published"
assert sidebar, "sidebar structure is empty"
for page in pages:
fm, _ = mt.parse(page)
assert "manTitle" not in fm, f"{page.name} leaked manTitle into site output"
assert "title" in fm, f"{page.name} has no title"
TESTS = [v for k, v in sorted(globals().items()) if k.startswith("test_")]