Skip to content

Commit 10aac29

Browse files
ppiegazedocsy
andauthored
docsy(v2): run the icon-name check in CI, and bump infra to provide it (#1460)
* docsy(v2): run the icon-name check in CI, and bump infra to provide it infra#263 added tools/check_icon_names.py and the make targets, but nothing called them, so the guard was manual-only. That does not close DOC-1444, whose whole point is that six dead icon names accumulated across two version lines precisely because nothing was watching. A checker no job runs is the same silence with extra steps. Adds check-icon-names.yml, modelled on check-links.yml (same triggers, same shape, same failure-message convention), and bumps the docs-infra submodule to 2f92666 so `make check-icon-names` exists. Ordering is load-bearing and this is the earliest safe point. The workflow fails unless BOTH the submodule provides the target AND the content is clean, so it could not land before infra#263 (merged) and #1455 (merged, which fixed the four dead v2 names). Verified against this branch: OK. The v1 line needs the same workflow; it ships with the v1 icon fixes in #1456, since v1's CI has drifted from main's before (docs#1281). Co-Authored-By: docsy <docsy@union.ai> Signed-off-by: Peeter Piegaze <1153481+ppiegaze@users.noreply.github.com> * docsy(v2): register check-icon-names in the delegator's TARGETS list CI failed with `make: *** No rule to make target 'check-icon-names'`, which is the wiring working as intended on its first run. The docs root Makefile is a thin delegator that forwards to unionai-docs-infra/Makefile through an explicit TARGETS list, and its own comment says why the list cannot be a pattern rule: "These must be listed explicitly because Make's % pattern rule won't match targets that correspond to existing files/directories." infra#263 added the target to the infra Makefile; nothing registered it here, so the delegator had no rule to forward. Adds check-icon-names and update-icon-names to TARGETS. Verified end to end via `make check-icon-names`, the same entry point CI uses, rather than by invoking the script directly -- which is what hid this: the script ran fine on its own path all along. Co-Authored-By: docsy <docsy@union.ai> Signed-off-by: Peeter Piegaze <1153481+ppiegaze@users.noreply.github.com> --------- Signed-off-by: Peeter Piegaze <1153481+ppiegaze@users.noreply.github.com> Co-authored-by: docsy <docsy@union.ai>
1 parent 6ef1405 commit 10aac29

3 files changed

Lines changed: 35 additions & 1 deletion

File tree

Lines changed: 33 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,33 @@
1+
name: Check Icon Names
2+
3+
on:
4+
pull_request:
5+
push:
6+
branches: [main]
7+
8+
jobs:
9+
check-icon-names:
10+
name: "Check Icon Names"
11+
runs-on: ubuntu-latest
12+
13+
steps:
14+
- name: Checkout repository
15+
uses: actions/checkout@v3
16+
with:
17+
submodules: 'true'
18+
19+
- name: Set up Python
20+
uses: actions/setup-python@v5
21+
with:
22+
python-version: '3.12'
23+
24+
- name: Install uv
25+
uses: astral-sh/setup-uv@v5
26+
27+
- name: Check icon names
28+
run: make check-icon-names
29+
30+
- name: Report status
31+
if: ${{ failure() }}
32+
run: |
33+
echo "::error::An icon name does not exist in the set its shortcode resolves against. Run 'make check-icon-names' locally for details. Bootstrap Icons: https://icons.getbootstrap.com/ — gemoji: https://api.github.com/emojis"

Makefile

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -19,6 +19,7 @@ TARGETS := usage help clean clean-generated base dist variant dev serve \
1919
update-examples init-examples check-jupyter check-images validate-urls \
2020
url-stats llm-docs update-redirects dry-run-redirects deploy-redirects \
2121
check-deleted-pages check-links check-generated-content check-api-docs \
22+
check-icon-names update-icon-names \
2223
check-llm-bundle-notes update-api-docs \
2324
check-helm-docs update-helm-docs generate-helm-docs \
2425
index-search index-search-settings check-search-labels \

0 commit comments

Comments
 (0)