Support catalogue

Review the proposed questions in SUPPORT_QUESTIONS.md. The corresponding support-questions.json contains stable IDs, categories, source guides, suggested batches and English answer drafts. Read the first batch in SUPPORT_ANSWERS.md. The first 32 answers have been approved and added to the published catalogue; 10 later questions remain drafts. These files are excluded from the site build and are not read by the support-page generator.

Published content lives in script/support.json. Add approved questions there after writing and checking their answers. Entries can be placed under any section and subsection; the support browser reads all of them into one searchable list. If a draft has a replaces field, update that existing entry while retaining its public ID, rather than adding a duplicate answer. review_id, source, answer_notes, review_notes, code_sources, batch and replaces are editorial metadata and do not need to be copied into the published catalogue. The catalogue’s required_locales requires Japanese, Simplified Chinese, Traditional Chinese and Korean questions and answers before publication. Generation rejects missing translations or English text copied into a translation field. Keep new entries as drafts until their translations are complete. Updated entries retain their public IDs.

Each entry needs:

Optional fields:

Category labels use label and label_locales. Filters are generated from categories with published entries. Search matches all query words across the question, aliases, answer and visible metadata. Question matches rank first; ties preserve catalogue order. Category counts respect the search and answer type, while showing the number available in each category.

The shared include _includes/support-browser.html serves all five language pages. Interface text lives in _data/support_ui.json. Without JavaScript, answers remain accessible as native expandable details; search controls stay hidden.

Generate support pages only:

python3 script/generate_features.py --support

script/build_site.sh regenerates features and support before building Jekyll. To check the support data and browser:

python3 -m unittest discover -s script -p 'test_support_data.py'
# Build/serve the site first; Playwright must be available to Node.
SUPPORT_BASE_URL=http://localhost:4000 node script/test_support_browser.cjs