fix: the manual's figures carry a description, and the check that would have caught them

Three pages opened in nothing in Yelp -- overview (twice), math, trigger -- while every check
passed. They are exactly the pages whose <figure> had a sentence for a <title> and no <desc>:
a Mallard <figure> is title?, desc?, media, and the page is refused without the description. The
idiomatic shape, as the shipped sets on this machine write it, moves the sentence into <desc> and
leaves the <title> a name. Note trigger-capture.png is referenced from both overview and trigger,
which is why those two pages are the ones that were noticed.

The gap was in the check, not in its reach: tools/check_help.sh ran links/media/orphans/ids and an
HTML render, and all four are silent about a page that is well linked, well parsed and
unrenderable. `yelp-check validate` is the one that would have caught it, and it cannot run here:
it generates the RelaxNG schema with xsltproc, but the schema includes the base grammar over the
network and this machine's libxml cannot fetch. So the structural rules this project has actually
needed are checked with xmllint's XPath, offline, each written after a page needed it.

One trap, paid for while writing it: a Mallard page declares
xmlns="http://projectmallard.org/1.0/", XPath 1.0 has no default namespace, so `//figure` matches
nothing and every rule passes -- the first version of this check reported "ok" for the very defect
it was written for. Every expression uses local-name() for that reason, and the script says so
where the next reader will be.

Verified: on the committed pages the rules exit 1 and name the three figures, which is the
failing-before case kept as the regression test; on the fixed pages all five rules pass and
yelp-build renders all 15; make -C build test-help goes from red to green.
This commit is contained in:
Martino Ferrari committed 2026-10-14 17:15:31 +02:00
1 parent 6fe1b22524
commit e140642ed0
4 files changed
+54 -14

No files matched your search

+3 -2
View File
@@ -89,8 +89,9 @@
are rebuilt whenever it changes.</p> are rebuilt whenever it changes.</p>
<figure> <figure>
<title>The math channel editor: nodes as boxes on a canvas, the orange dot marking the <title>The math channel editor</title>
output, and the selected node's parameters beside it.</title> <desc>Nodes as boxes on a canvas, the orange dot marking the output, and the selected
node's parameters beside it.</desc>
<media type="image" mime="image/png" src="figures/math-editor.png" width="938" height="420"> <media type="image" mime="image/png" src="figures/math-editor.png" width="938" height="420">
<p>A sine source wired through a gain into a filter, with the output on the last <p>A sine source wired through a gain into a filter, with the output on the last
node.</p> node.</p>
+6 -5
View File
@@ -33,8 +33,9 @@
is still there when you need it — an export writes it as a second column.</p> is still there when you need it — an export writes it as a second column.</p>
<figure> <figure>
<title>The udpscope window on a live source: the header bar above the plots, the side <title>The udpscope window</title>
panel on the right, and the status bar along the bottom.</title> <desc>The header bar above the plots, the side panel on the right, and the status bar
along the bottom.</desc>
<media type="image" mime="image/png" src="figures/main-window.png" width="1280" height="800"> <media type="image" mime="image/png" src="figures/main-window.png" width="1280" height="800">
<p>One figure of three rows with four signals enabled, the Measure tab open in the <p>One figure of three rows with four signals enabled, the Measure tab open in the
side panel, and the status bar reading <gui>live</gui>.</p> side panel, and the status bar reading <gui>live</gui>.</p>
@@ -95,9 +96,9 @@
</terms> </terms>
<figure> <figure>
<title>A held capture: the trigger has latched, the state reads <title>A held capture</title>
<gui>event held</gui>, and the axis is pinned to the event rather than to the live <desc>The trigger has latched, the state reads <gui>event held</gui>, and the axis is
edge.</title> pinned to the event rather than to the live edge.</desc>
<media type="image" mime="image/png" src="figures/trigger-capture.png" width="1280" height="800"> <media type="image" mime="image/png" src="figures/trigger-capture.png" width="1280" height="800">
<p>The trigger marker sits at x = 0 and the pre-trigger share of the window is drawn <p>The trigger marker sits at x = 0 and the pre-trigger share of the window is drawn
to its left.</p> to its left.</p>
+3 -2
View File
@@ -41,8 +41,9 @@
</steps> </steps>
<figure> <figure>
<title>A held capture: <gui>event held</gui> in the status bar, the axis pinned to the <title>A held capture</title>
event, and the trigger marker at x = 0.</title> <desc><gui>event held</gui> in the status bar, the axis pinned to the event, and the
trigger marker at x = 0.</desc>
<media type="image" mime="image/png" src="figures/trigger-capture.png" width="1280" height="800"> <media type="image" mime="image/png" src="figures/trigger-capture.png" width="1280" height="800">
<p>The pre-trigger share is drawn to the left of the marker.</p> <p>The pre-trigger share is drawn to the left of the marker.</p>
</media> </media>
+42 -5
View File
@@ -3,12 +3,26 @@
# SPDX-FileCopyrightText: 2026 Martino Ferrari # SPDX-FileCopyrightText: 2026 Martino Ferrari
# #
# The manual's check, the way tools/check_ui.py is the .ui files' check: it runs the four # The manual's check, the way tools/check_ui.py is the .ui files' check: it runs the four
# yelp-check subcommands that need no network, then renders the whole set to HTML with # yelp-check subcommands that need no network, checks the structural rules below, then renders
# yelp-build so a page that only fails at the transform step is caught too. # the whole set to HTML with yelp-build so a page that only fails at the transform step is
# caught too.
# #
# `yelp-check validate` is deliberately absent. It fetches the Mallard RelaxNG schema over # `yelp-check validate` is deliberately absent, and it is worth being exact about why: it
# the network and fails on an offline machine for that reason rather than for a fault in the # builds the RelaxNG schema locally with `xsltproc` and `mal-rng.xsl`, but the schema it
# pages -- links/media/orphans/ids are the offline checks. # generates includes the base grammar over the network, so on a machine where fetching is
# blocked it fails for that reason rather than for a fault in the pages.
#
# What that leaves unverified is *structure*, and that is not a small gap: three pages
# carrying a `<figure>` whose `<title>` was a sentence and which had no `<desc>` opened in
# nothing, while links, media, orphans, ids and the HTML render all passed. So the rules this
# project has actually needed are checked here, with xmllint's XPath and no schema at all.
# Add a rule when a page *teaches* one -- and do not add one you have not seen broken: a rule
# that fails a legal page is worse than no rule.
#
# Every expression uses local-name() on purpose. A Mallard page declares
# `xmlns="http://projectmallard.org/1.0/"`, and XPath 1.0 has no default namespace, so
# `//figure` matches *nothing* and every rule below silently passes. Written naively, this
# check reported "ok" for the very defect it was written for.
# #
# Exit status is nonzero on the first finding. # Exit status is nonzero on the first finding.
set -eu set -eu
@@ -27,6 +41,29 @@ for sub in links media orphans ids; do
yelp-check "$sub" "$dir" >/dev/null yelp-check "$sub" "$dir" >/dev/null
done done
failed=0
for page in "$dir"/*.page; do
while IFS='|' read -r what expr; do
[ -n "$what" ] || continue
n=$(xmllint --xpath "$expr" "$page" 2>/dev/null || echo "?")
if [ "$n" != "0" ]; then
printf ' %s: %s (%s)\n' "$(basename "$page")" "$what" "$n" >&2
failed=1
fi
done <<'RULES'
a figure describes itself|count(//*[local-name()='figure'][not(*[local-name()='desc'])])
a section has a title|count(//*[local-name()='section'][not(*[local-name()='title'])])
info is the page's first child|count(/*[local-name()='page']/*[1][local-name()!='info'])
no text sits directly under page|count(/*[local-name()='page']/text()[normalize-space()])
a media names its src|count(//*[local-name()='media'][not(@src)])
RULES
done
if [ "$failed" != 0 ]; then
echo "check_help: structural findings above" >&2
exit 1
fi
printf 'structure %s: ok\n' "$dir"
mkdir -p "$tmp/html" mkdir -p "$tmp/html"
yelp-build html -o "$tmp/html" "$dir" yelp-build html -o "$tmp/html" "$dir"
pages=$(find "$tmp/html" -maxdepth 1 -name '*.html' | wc -l) pages=$(find "$tmp/html" -maxdepth 1 -name '*.html' | wc -l)