- variant="default" (listing): cover image conditional (D-03 no fallback),
aspect-[16/9], first tag UBadge, date i18n via Intl.DateTimeFormat,
h2 title, line-clamp-2 description, reading time + extra tags pills,
absolute inset-0 NuxtLink for SEO/a11y (D-02 tags non-clickable)
- variant="compact" (prev/next, D-09/D-10): no image, label row with
UIcon arrow (left/right per direction), h3 title, date mono,
text-right on next / text-left on prev
- Props: article, variant?='default'|'compact', direction?='prev'|'next'
- Slug derived from article.path last segment (locale-agnostic)
- readingMinutes: uses article.minutes (Nitro hook) with useReadingTime
fallback on article.description
- Schema.org BlogPosting markup (headline/description/keywords/url/image/
datePublished) — ready for Phase 7 JSON-LD Article
- a11y aria-label interpolated via t('a11y.blogPrev'|'blogNext', {title})
- Insert { key: 'blog', path: '/blog' } in navLinks computed array
- Position: between hytale and projects (D-15 ordering)
- Template v-for iterations unchanged — new link auto-propagates to
desktop nav + mobile slideover
- Label resolved via t(`nav.${link.key}`) → uses nav.blog key
added in Task 2.1
- Add `draft: true` to frontmatter of both test-kotlin-syntax.md files
so they are excluded from all `queryCollection(...).where('draft', '=', false)`
listings (D-14).
- Articles remain accessible via direct URL (no draft filter on `.path(x).first()`),
keeping them available for internal rendering tests.
- Listings will be empty until real Hytale seed articles land in Phase 8 —
the empty state will render per D-16 ("Hytale articles coming soon" CTA).
- Body content untouched (only frontmatter +1 line each).
- Pure synchronous helper returning minutes (>= 1) from either a pre-computed
word count (number) or raw text (string, tokenized on whitespace).
- Client-side safety net when `article.minutes` isn't yet populated
(e.g., dev hot-reload before the Nitro hook re-parsed). Source of truth
remains the Nitro `content:file:afterParse` hook (D-19).
- Same 200 wpm formula as server-side hook — ensures listing ↔ article parity.
- Auto-imported by Nuxt thanks to `use*` naming convention.
- Register `content:file:afterParse` hook to inject `wordCount` + `minutes`
on every parsed markdown content object (D-19: 200 wpm, floor 1 min).
- Import pure util `countWordsInMinimalBody` from app/utils/countWords.
- Guard against non-`.md` files (defensive — hook fires on all sources).
- Values persist in @nuxt/content SQLite DB and are queryable via
queryCollection thanks to matching Zod fields (content.config.ts).
- Pure AST traversal of @nuxt/content v3 minimal body shape
- Skips code and pre tags (code snippets are not readable prose)
- Zero dependency, zero import, reused by Nitro hook
- Add draft: z.boolean().optional().default(false) to allow .where('draft','=',false)
- Add wordCount + minutes as optional (injected by Nitro hook at parse time)
- Collections blog_fr/blog_en unchanged (schema is referenced by variable)
- content/fr/blog/test-kotlin-syntax.md: FR test article covering all 4 validation criteria
- content/en/blog/test-kotlin-syntax.md: EN version with same slug
- app/pages/test.vue: temporary page at /test for visual checkpoint verification
- Both articles contain: kotlin code block, NuxtImg image, markdown table, 4 callout types
- Introduced a new document outlining the configuration and component patterns for integrating @nuxt/content.
- Included mappings for `nuxt.config.ts`, `content.config.ts`, and new components `ProseImg.vue` and `Alert.vue`.
- Added example markdown content for testing syntax highlighting and layout.
- Documented critical patterns and anti-patterns to follow during implementation.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
kayjaydee