diff --git a/.docsearch.README.md b/.docsearch.README.md
new file mode 100644
index 00000000000..2310abe4854
--- /dev/null
+++ b/.docsearch.README.md
@@ -0,0 +1,111 @@
+# DocSearch Configuration
+
+This directory contains the Algolia DocSearch configuration for the Apache Camel website.
+
+## Overview
+
+The `.docsearch.config.json` file defines how Algolia's crawler indexes the Camel website for search functionality. This configuration ensures that all relevant content is discoverable through the site search, including:
+
+- All component documentation (not just canonical versions)
+- Tables with component specifications and supported models
+- Metadata sections and inline code
+- Multiple documentation versions (next, latest, and release branches)
+
+## Key Configuration Elements
+
+### Index Settings (`index`)
+- **name**: `apache_camel` - The Algolia index where content is stored
+- **startUrls**: Entry points for the crawler
+- **pathsToMatch**: URL patterns to include in indexing
+- **pathsToIgnore**: URLs to skip (search pages, error pages, etc.)
+- **includeHeadingLevels**: All heading levels (h1-h6) are indexed for better navigation
+
+### Content Selectors (`selectors`)
+
+These CSS selectors define what content gets indexed:
+
+- **lvl0-lvl5**: Heading hierarchy (h1-h6) used to build the breadcrumb structure
+- **text**: Main content to index including:
+ - Paragraphs (`p`), list items (`li`)
+ - Table cells (`td`, `th`) - **Important for component specs**
+ - Definition terms (`dt`, `dd`)
+ - Code blocks (`code`, `pre`)
+
+This ensures keywords like "PyTorch" in Model Zoo tables are indexed, fixing issue #1209.
+
+### Exclusions (`selectors_exclude`)
+
+Navigation, sidebars, footers, and other non-content elements are excluded to improve search quality:
+- `.no_index`, `[data-no-index]` - Custom exclusion attributes
+- Navigation elements: `nav`, `.navbar`, `.menu`, `.sidebar`, `.toc`
+- Footer and copyright
+- Hidden elements: `.hidden`, `[aria-hidden='true']`
+
+### Crawling Rules (`crawler`)
+
+- **maxDepth**: 20 - Allows deep navigation through component docs
+- **maxUrls**: 50,000 - Sufficient for Camel's comprehensive documentation
+- **sitemapUrls**: Uses sitemap for efficient crawling
+- **timeoutMs**: 30,000 - Adequate for large pages with tables
+
+### Multi-Version Support (`start_urls`)
+
+The configuration crawls multiple documentation versions:
+
+1. **next** (page_rank: 5) - Development version
+2. **latest** (page_rank: 5) - Latest stable
+3. **\d+\.\d+\.\x** (page_rank: 4) - Release branches (4.4.x, 4.10.x, etc.)
+4. **manual** (page_rank: 7) - Core documentation (highest priority)
+5. **docs** (page_rank: 6) - General documentation
+6. **blog** (page_rank: 3) - Blog posts
+
+This addresses the issue where only canonical (4.4.x) pages were indexed.
+
+### Search Behavior (`custom_settings`)
+
+- **searchableAttributes**: Fields available for full-text search
+- **separatorsToIndex**: Include underscores, dots, and dashes in search (important for component names like `camel-k`)
+- **attributeForDistinctResults**: Deduplicate results by URL to avoid showing the same page multiple times
+
+## Maintenance
+
+When making changes to this configuration:
+
+1. **Test locally** - Build the site and verify crawling works
+2. **Document changes** - Explain why selectors or URLs were modified
+3. **Consider impacts** - Changes affect search indexing across all users
+4. **Verify coverage** - Use Algolia dashboard to check what's indexed
+
+### Common Modifications
+
+**Adding new documentation sections:**
+```json
+{
+ "url": "https://camel.apache.org/new-section/",
+ "page_rank": 5
+}
+```
+
+**Excluding problematic content:**
+```json
+"selectors_exclude": [
+ ".no_index",
+ ".problematic-element"
+]
+```
+
+**Adjusting content extraction:**
+Modify the `text` selector in the `selectors` section to include additional elements.
+
+## Related Issue
+
+- **Issue #1209**: "The search is not finding several fields"
+ - Problem: Keywords like Bradley, firmata, PyTorch not indexed from component documentation
+ - Root cause: Missing configuration for table content and non-canonical versions
+ - Solution: This configuration file with improved selectors and multi-version crawling
+
+## References
+
+- [Algolia DocSearch Documentation](https://docsearch.algolia.com/)
+- [Camel Website GitHub](https://github.com/apache/camel-website)
+- [Issue #1209](https://github.com/apache/camel-website/issues/1209)
diff --git a/.docsearch.config.json b/.docsearch.config.json
new file mode 100644
index 00000000000..557fd70e84f
--- /dev/null
+++ b/.docsearch.config.json
@@ -0,0 +1,125 @@
+{
+ "index": {
+ "name": "apache_camel",
+ "startUrls": [
+ "https://camel.apache.org/"
+ ],
+ "ignoreCanonicalTo": false,
+ "pathsToMatch": [
+ "https://camel.apache.org/**"
+ ],
+ "pathsToIgnore": [
+ "https://camel.apache.org/search",
+ "https://camel.apache.org/404.html"
+ ],
+ "includeHeadingLevels": [1, 2, 3, 4, 5, 6],
+ "stripQueryParameters": true
+ },
+ "crawler": {
+ "userAgent": "Algolia Crawler",
+ "maxDepth": 20,
+ "maxUrls": 50000,
+ "waitUntilFired": true,
+ "timeoutMs": 30000,
+ "sitemapUrls": [
+ "https://camel.apache.org/sitemap.xml"
+ ],
+ "ignoreRobotsTxt": false,
+ "allowedDomains": [
+ "camel.apache.org"
+ ]
+ },
+ "selectors": {
+ "lvl0": {
+ "selector": "h1",
+ "global": true,
+ "default_value": "Documentation"
+ },
+ "lvl1": "h2",
+ "lvl2": "h3",
+ "lvl3": "h4",
+ "lvl4": "h5",
+ "lvl5": "h6",
+ "text": "p, li, td, th, dt, dd, span:not(.tooltip), div:not([class*='hidden']), table tbody, code, pre"
+ },
+ "selectors_exclude": [
+ ".no_index",
+ "[data-no-index]",
+ ".sidebar",
+ ".breadcrumb",
+ "nav",
+ ".navbar",
+ ".menu",
+ ".toc",
+ "footer",
+ ".footer",
+ ".copyright",
+ ".hide",
+ ".hidden",
+ "[aria-hidden='true']",
+ "script",
+ "style",
+ ".language-toggle",
+ ".sidebar-toggle"
+ ],
+ "min_indexed_level": 1,
+ "only_content_level": false,
+ "start_urls": [
+ {
+ "url": "https://camel.apache.org/components/next/",
+ "page_rank": 5
+ },
+ {
+ "url": "https://camel.apache.org/components/latest/",
+ "page_rank": 5
+ },
+ {
+ "url": "https://camel.apache.org/components/(\\d+)\\.(\\d+)\\.x/",
+ "page_rank": 4
+ },
+ {
+ "url": "https://camel.apache.org/manual/",
+ "page_rank": 7
+ },
+ {
+ "url": "https://camel.apache.org/docs/",
+ "page_rank": 6
+ },
+ {
+ "url": "https://camel.apache.org/blog/",
+ "page_rank": 3
+ },
+ {
+ "url": "https://camel.apache.org/",
+ "page_rank": 8
+ }
+ ],
+ "stop_urls": [
+ "\\?",
+ "#"
+ ],
+ "custom_settings": {
+ "separatorsToIndex": "_.-",
+ "attributesForFaceting": [
+ "version"
+ ],
+ "attributesToIndex": [
+ "hierarchy",
+ "content",
+ "url"
+ ],
+ "minWordSizefor1Typo": 4,
+ "minWordSizefor2Typos": 8,
+ "exactOnSingleWordQuery": "none",
+ "attributeForDistinctResults": "url",
+ "searchableAttributes": [
+ "hierarchy.lvl0",
+ "hierarchy.lvl1",
+ "hierarchy.lvl2",
+ "hierarchy.lvl3",
+ "hierarchy.lvl4",
+ "hierarchy.lvl5",
+ "content"
+ ]
+ }
+}
diff --git a/README.md b/README.md
index 62a0c645f5b..81c42a8ceb2 100644
--- a/README.md
+++ b/README.md
@@ -453,6 +453,35 @@ all generated sources in the project first.
Of course this then takes some more time than an optimized rebuild (time to grab another coffee!).
+## Search Indexing Configuration
+
+The website uses [Algolia DocSearch](https://docsearch.algolia.com/) to provide site-wide search functionality. The search configuration is defined in [`.docsearch.config.json`](.docsearch.config.json).
+
+### What is indexed
+
+The configuration ensures that Algolia's crawler indexes:
+- All documentation versions (development `next`, latest, and release branches like `4.4.x`)
+- Component specifications and tables (fixing issue #1209)
+- All heading levels and content blocks
+- Code blocks and inline code snippets
+
+### Maintaining the search configuration
+
+If you need to modify what gets indexed or how content is crawled:
+
+1. Edit [`.docsearch.config.json`](.docsearch.config.json) to change selectors or crawling rules
+2. Review the detailed documentation in [`.docsearch.README.md`](.docsearch.README.md)
+3. Test your changes by building the site locally: `yarn build`
+4. Verify content is indexable by visiting the search functionality in the preview
+
+Key elements to be aware of:
+- **Selectors** define what HTML elements are indexed (headings, paragraphs, tables, code)
+- **start_urls** control which parts of the site are crawled and their search priority
+- **selectors_exclude** specify elements to skip (navigation, sidebars, footers)
+- **custom_settings** control search behavior and index settings
+
+For more details, see [`.docsearch.README.md`](.docsearch.README.md).
+
# Checks, publishing the website
The content of the website, as built by the [Camel.website](https://ci-builds.apache.org/job/Camel/job/Camel.website/job/main/)
diff --git a/antora-ui-camel/src/css/header.css b/antora-ui-camel/src/css/header.css
index 8b265ec82da..9eccac2bd0a 100644
--- a/antora-ui-camel/src/css/header.css
+++ b/antora-ui-camel/src/css/header.css
@@ -303,6 +303,8 @@ html:not([data-scroll='0']) .navbar {
margin-right: 10px;
overflow-y: auto;
max-height: 80vh;
+ max-width: min(600px, 90vw);
+ min-width: 300px;
scrollbar-width: thin; /* Firefox */
}
diff --git a/antora-ui-camel/src/js/vendor/algoliasearch.bundle.js b/antora-ui-camel/src/js/vendor/algoliasearch.bundle.js
index a4b692ab8df..5951d366b5c 100644
--- a/antora-ui-camel/src/js/vendor/algoliasearch.bundle.js
+++ b/antora-ui-camel/src/js/vendor/algoliasearch.bundle.js
@@ -5,6 +5,7 @@
const MAX_SNIPPET_LENGTH = 200
const RESULTS_LIMIT = 10
+ const MAX_INPUT_LENGTH = 200
// Sub-projects to exclude from main search - users can browse these directly
const EXCLUDED_SUBPROJECTS = [
@@ -16,6 +17,15 @@
'/camel-karaf/',
]
+ // Core docs patterns - these should rank higher than component pages
+ const CORE_DOCS_PATTERNS = [
+ '/manual/',
+ '/user-guide/',
+ '/architecture/',
+ '/getting-started/',
+ '/faq/',
+ ]
+
// Check if a URL belongs to a sub-project that should be filtered out
function isSubProjectUrl (url) {
if (!url) return false
@@ -24,6 +34,125 @@
})
}
+ // Check if a URL points to core documentation (should rank higher)
+ function isCoreDocsUrl (url) {
+ if (!url) return false
+ return CORE_DOCS_PATTERNS.some(function (pattern) {
+ return url.indexOf(pattern) !== -1
+ })
+ }
+
+ // Check if a URL points to component documentation
+ function isComponentUrl (url) {
+ if (!url) return false
+ return url.indexOf('/components/') !== -1
+ }
+
+ // Sort hits to prioritize core docs over components
+ function sortByCoreDocs (hits) {
+ return hits.sort(function (a, b) {
+ var aIsCore = isCoreDocsUrl(a.url)
+ var bIsCore = isCoreDocsUrl(b.url)
+ var aIsComponent = isComponentUrl(a.url)
+ var bIsComponent = isComponentUrl(b.url)
+
+ // Core docs first
+ if (aIsCore && !bIsCore) return -1
+ if (!aIsCore && bIsCore) return 1
+
+ // Components last
+ if (aIsComponent && !bIsComponent) return 1
+ if (!aIsComponent && bIsComponent) return -1
+
+ return 0
+ })
+ }
+
+ // Extract the parent page path from a URL (removes anchor and trailing segments)
+ function getParentPagePath (url) {
+ if (!url) return ''
+ // Remove anchor fragment
+ var path = url.split('#')[0]
+ // Normalize trailing slash
+ if (path.endsWith('/')) {
+ path = path.slice(0, -1)
+ }
+ return path
+ }
+
+ // Check if hit represents a sub-section of a page (has anchor or deeper hierarchy)
+ function isSubSection (hit) {
+ if (!hit || !hit.url) return false
+ return hit.url.indexOf('#') !== -1
+ }
+
+ // Get the breadcrumb depth (number of hierarchy levels)
+ function getBreadcrumbDepth (hit) {
+ if (!hit || !hit.hierarchy) return 0
+ return Object.values(hit.hierarchy).filter(function (lvl) {
+ return lvl !== null
+ }).length
+ }
+
+ // Remove duplicate results for the same parent page
+ // When parent page is a direct match, exclude its sub-sections
+ function deduplicateHits (hits, query) {
+ var seenPages = {}
+ var parentMatches = {}
+ var queryLower = (query || '').toLowerCase().trim()
+
+ // First pass: identify parent pages that match the query directly
+ hits.forEach(function (hit) {
+ var parentPath = getParentPagePath(hit.url)
+ var hierarchy = hit.hierarchy || {}
+
+ // Check if any top-level hierarchy matches the search query
+ var lvl1 = (hierarchy.lvl1 || '').toLowerCase()
+ var lvl0 = (hierarchy.lvl0 || '').toLowerCase()
+
+ if (lvl1 && lvl1.indexOf(queryLower) !== -1) {
+ parentMatches[parentPath] = true
+ }
+ if (lvl0 && lvl0.indexOf(queryLower) !== -1) {
+ parentMatches[parentPath] = true
+ }
+ })
+
+ // Second pass: filter out sub-sections when parent is already matched
+ return hits.filter(function (hit) {
+ var parentPath = getParentPagePath(hit.url)
+ var isSubSec = isSubSection(hit)
+ var depth = getBreadcrumbDepth(hit)
+
+ // If this is a sub-section and parent page already matched, skip it
+ if (isSubSec && parentMatches[parentPath]) {
+ // Only keep the main page hit, not sub-sections
+ if (seenPages[parentPath]) {
+ return false
+ }
+ }
+
+ // For component pages, only show the main entry (depth <= 2)
+ if (isComponentUrl(hit.url) && depth > 2 && seenPages[parentPath]) {
+ return false
+ }
+
+ // Track that we've seen this parent page
+ if (!seenPages[parentPath]) {
+ seenPages[parentPath] = { depth: depth, hit: hit }
+ return true
+ }
+
+ // If we already have this page, only keep if it's a better match (shallower)
+ if (depth < seenPages[parentPath].depth) {
+ seenPages[parentPath] = { depth: depth, hit: hit }
+ return true
+ }
+
+ return false
+ })
+ }
+
function truncateHighlightedHtml (html, maxChars) {
if (!html || maxChars <= 0) return ''
@@ -104,6 +233,13 @@
e.stopPropagation()
})
+ // Enforce max input length as backup to HTML maxlength attribute
+ search.addEventListener('input', function () {
+ if (search.value.length > MAX_INPUT_LENGTH) {
+ search.value = search.value.substring(0, MAX_INPUT_LENGTH)
+ }
+ })
+
search.addEventListener(
'keyup',
debounce((key) => {
@@ -119,16 +255,21 @@
return
}
cancel.style.display = 'block'
+ var searchQuery = search.value
index
- .search(search.value, {
- hitsPerPage: 10,
+ .search(searchQuery, {
+ hitsPerPage: 50,
})
.then((results) => {
// Filter out sub-project results to focus on camel-core documentation
const filteredHits = results.hits.filter(function (hit) {
return !isSubProjectUrl(hit.url)
- }).slice(0, RESULTS_LIMIT)
- const data = filteredHits.reduce((data, hit) => {
+ })
+ // Remove duplicate results for the same parent page
+ const dedupedHits = deduplicateHits(filteredHits, searchQuery)
+ // Sort to prioritize core docs over components and limit results
+ const sortedHits = sortByCoreDocs(dedupedHits).slice(0, RESULTS_LIMIT)
+ const data = sortedHits.reduce((data, hit) => {
const section = hit.hierarchy.lvl0
const sectionKey = `${section}-${hit.version || ''}`
diff --git a/antora-ui-camel/src/partials/header-content.hbs b/antora-ui-camel/src/partials/header-content.hbs
index 3173cfc98e4..4ae105d550c 100644
--- a/antora-ui-camel/src/partials/header-content.hbs
+++ b/antora-ui-camel/src/partials/header-content.hbs
@@ -17,7 +17,7 @@
