feat: route report and drift check at server start

- the module writes #seo-routes (static paths and route patterns) from the
  resolved pages; options seo.routeReport, seo.pagination and
  seo.htmlSitemap reach the public runtime config
- seo-route-report: POST /registry/routes once per process with the
  route-report token, one retry after 60 s on failure, nothing without a
  token, with the layer off or with routeReport false
- seo-drift: one "[vue-seo-layer] drift:" warning per mismatch between
  the CMS's site answer (URL patterns, pagination, HTML sitemap paths)
  and the build
- GET /api/__seo__/site and useSeoSite()
- smoke grammar routereport and log, fixture registry route, drift run
This commit is contained in:
Jakub Zych
2026-11-04 11:49:21 +01:00
parent 9cc2cda924
commit 3abc986f42
21 changed files with 599 additions and 20 deletions

36
lib/options.ts Normal file
View File

@@ -0,0 +1,36 @@
// The layer's pagination and HTML-sitemap module options normalised for
// the runtime config (lib/seo-module.ts) and their readers in app and
// server code. Pure, so `node --test` drives it.
export interface SeoPaginationOption {
form: 'query' | 'path'
param: string
segments: Record<string, string>
}
export type SeoHtmlSitemapOption = { paths: Record<string, string> } | false
/** The pagination option with its defaults, as the runtime config carries it. */
export function paginationOption(value: unknown): SeoPaginationOption {
const v = value !== null && typeof value === 'object' ? (value as Record<string, unknown>) : {}
const segments: Record<string, string> = {}
const raw = v.segments !== null && typeof v.segments === 'object' ? (v.segments as Record<string, unknown>) : {}
for (const [locale, segment] of Object.entries(raw)) if (typeof segment === 'string' && /^[\p{L}\p{N}_-]+$/u.test(segment)) segments[locale] = segment
return {
form: v.form === 'path' ? 'path' : 'query',
param: typeof v.param === 'string' && /^[A-Za-z0-9_-]+$/.test(v.param) ? v.param : 'page',
segments,
}
}
/** The HTML sitemap option: false, or the paths (each '/' plus at least one character). */
export function htmlSitemapOption(value: unknown): SeoHtmlSitemapOption {
if (value === false) return false
const v = value !== null && typeof value === 'object' ? (value as Record<string, unknown>) : {}
const paths: Record<string, string> = {}
const raw = v.paths !== null && typeof v.paths === 'object' ? (v.paths as Record<string, unknown>) : {}
for (const [locale, path] of Object.entries(raw)) {
if (typeof path === 'string' && /^\/[^?#\s]+$/.test(path) && !path.includes(':') && !path.includes('*')) paths[locale] = path.replace(/\/+$/, '')
}
return { paths }
}

View File

@@ -5,6 +5,21 @@
// for the route report and the drift check. Pure TypeScript, so
// `node --test` drives it.
import { flattenPages, isNoindex, pagePattern, type GuardPage } from './guard.ts'
export interface RoutePage extends GuardPage {
name?: string
meta?: Record<string, unknown>
children?: RoutePage[]
}
export interface RoutePattern {
/** A route-rule style pattern: '/blog/:slug', '/docs/**', '/'. */
pattern: string
/** True when the route ends in a catch-all ('[...slug]'). */
catchAll: boolean
}
/** True when a robots value keeps the page out of the index. */
export const robotsNoindex = (robots: unknown) => typeof robots === 'string' && /(^|[\s,])(noindex|none)($|[\s,])/i.test(robots)
@@ -23,3 +38,47 @@ export function pageMetaRobots(meta: Record<string, unknown> | undefined | null)
if (meta.requiresAuth === true) return 'noindex, nofollow'
return null
}
/** The layer's own HTML sitemap pages, which the CMS knows already. */
const isLayerPage = (name: unknown) => typeof name === 'string' && name.startsWith('seo-sitemap-')
/**
* The paths without parameters that may be indexed, sorted: no page whose
* meta keeps it out of the index (requiresAuth, a noindex seo.robots), none
* under seo.noindexRoutes, none of the layer's HTML sitemap pages and no
* client-only page.
*/
export function staticRoutes(pages: readonly RoutePage[], noindexRoutes: readonly string[]): string[] {
const out = new Set<string>()
for (const page of flattenPages(pages) as Array<RoutePage & { fullPath: string }>) {
// A parent with children renders through one of them.
if (Array.isArray(page.children) && page.children.some((c) => c.path === '')) continue
const path = page.fullPath
if (path.includes(':') || path.includes('*')) continue
if (page.mode === 'client' || isLayerPage(page.name)) continue
if (robotsNoindex(pageMetaRobots(page.meta))) continue
if (isNoindex(path, noindexRoutes)) continue
out.add(path)
}
return [...out].sort()
}
/** Every page's route pattern ('/posts/:slug', '/**' for a catch-all), sorted. */
export function routePatterns(pages: readonly RoutePage[]): RoutePattern[] {
const seen = new Map<string, RoutePattern>()
for (const page of flattenPages(pages)) {
const pattern = pagePattern(page.fullPath)
if (!seen.has(pattern)) seen.set(pattern, { pattern, catchAll: pattern.endsWith('/**') })
}
return [...seen.values()].sort((a, b) => a.pattern.localeCompare(b.pattern))
}
/** The body of the server template #seo-routes. */
export function routesTemplate(staticPaths: readonly string[], patterns: readonly RoutePattern[]): string {
return [
'// Generated by vue-seo-layer (lib/routes.ts) from the resolved pages.',
`export const staticRoutes = ${JSON.stringify(staticPaths)}`,
`export const routePatterns = ${JSON.stringify(patterns)}`,
'',
].join('\n')
}

View File

@@ -8,8 +8,10 @@
// app still uses it: each module reads only its own keys, so the layer's
// options never use a name nuxt-seo-utils has (enabled, fallbackTitle …).
import { defineNuxtModule, useLogger } from 'nuxt/kit'
import { addServerTemplate, defineNuxtModule, useLogger } from 'nuxt/kit'
import { findOffenders, guardMessage, noindexRule, type GuardPage } from './guard.ts'
import { htmlSitemapOption, paginationOption, type SeoHtmlSitemapOption, type SeoPaginationOption } from './options.ts'
import { routePatterns, routesTemplate, staticRoutes, type RoutePage } from './routes.ts'
export interface SeoModuleOptions {
/**
@@ -45,6 +47,25 @@ export interface SeoModuleOptions {
* build guard then checks nothing. Default true.
*/
indexable: boolean
/**
* Report the site's static routes to the CMS's page registry at server
* start (D-33). Needs the route-report token (NUXT_SEO_ROUTE_REPORT_TOKEN,
* else NUXT_SEO_NOTFOUND_TOKEN). Default true.
*/
routeReport: boolean
/**
* The pagination URL form the site's routes are built for (D-40): query
* (?page=2) or path (/blog/strona/2, the segment per locale). The CMS's
* resolve answer wins at runtime; the startup drift check logs a
* difference.
*/
pagination: SeoPaginationOption
/**
* The HTML sitemap pages (D-34), by locale: the overview at the path, a
* group at <path>/<group>. false registers none. A path an app page
* already has is left to the app.
*/
htmlSitemap: SeoHtmlSitemapOption
}
export const SEO_MODULE_DEFAULTS: SeoModuleOptions = {
@@ -55,6 +76,9 @@ export const SEO_MODULE_DEFAULTS: SeoModuleOptions = {
resolveFiles: ['pdf'],
noindexRoutes: [],
indexable: true,
routeReport: true,
pagination: { form: 'query', param: 'page', segments: { pl: 'strona', en: 'page' } },
htmlSitemap: { paths: { pl: '/mapa-strony', en: '/sitemap' } },
}
const strings = (v: unknown): string[] => (Array.isArray(v) ? v.filter((e): e is string => typeof e === 'string' && e !== '') : [])
@@ -72,6 +96,8 @@ export default defineNuxtModule<SeoModuleOptions>({
// a global ssr: false into a '/**' route rule.
const ssr = nuxt.options.ssr !== false
const inlineRouteRules = nuxt.options.experimental?.inlineRouteRules === true
const pagination = paginationOption(options.pagination)
const htmlSitemap = htmlSitemapOption(options.htmlSitemap)
// requiresAuth and seo are read from definePageMeta at build time, so
// the route report leaves signed-in and noindex pages out (D-23).
@@ -123,6 +149,17 @@ export default defineNuxtModule<SeoModuleOptions>({
pages = resolved as GuardPage[]
guard()
})
// The route artefact (#seo-routes) for the route report and the drift
// check: written from the resolved pages when Nitro bundles the
// server, so no network call happens at build time.
addServerTemplate({
filename: '#seo-routes',
getContents: () => {
const resolved = (pages ?? []) as RoutePage[]
return routesTemplate(staticRoutes(resolved, noindexRoutes), routePatterns(resolved))
},
})
nuxt.hook('nitro:build:before', (nitro) => {
nitroRules = nitro.options.routeRules as Record<string, unknown>
if (!checked || nuxt.options.dev) guard()
@@ -141,6 +178,9 @@ export default defineNuxtModule<SeoModuleOptions>({
resolveFiles: Array.isArray(options.resolveFiles) ? [...options.resolveFiles] : [...SEO_MODULE_DEFAULTS.resolveFiles],
noindexRoutes,
indexable,
routeReport: options.routeReport !== false,
pagination,
htmlSitemap,
}
},
})