Add unified navigation extension and update component references#197
Merged
Conversation
Fixes two bugs in connector documentation generation: 1. CGO-only connectors incorrectly reported as "removed" - When processing intermediate versions, newData wasn't augmented with CGO connectors from binary analysis before diff generation, causing tigerbeetle_cdc, zmq4, ffi, etc. to appear as removed. 2. Cloud metadata inconsistency - New components like gcp_bigquery_write_api had cloudSupported: false in newComponents even when binary analysis correctly identified them as cloud-supported. Changes: - Add augmentConnectorData() helper function for reusable data augmentation - Call augmentation before intermediate diff generation - Save augmented data back to disk for subsequent iterations - Refactor main augmentation to use the new helper function
Previously, cleanup kept versions based on current run's intermediate processing results, which missed files from previous runs that were committed to git. Now cleanup: - Sorts all versions by semver descending - Keeps only the latest 2 versions - Deletes all older JSON and CSV files This ensures docs-data stays clean regardless of intermediate processing history from previous automation runs.
When a connector page exists but the connector has been removed from the CSV data (like the salesforce processor), the component_type_dropdown macro now shows a clear message directing users to the release notes for migration guidance instead of showing 'Available in: (blank)'.
Instead of hardcoding paths like '/redpanda-connect/get-started/whats-new/', use Antora's contentCatalog.resolvePage() to dynamically resolve the whats-new page URL. This generates proper relative URLs from the current page and handles version/component variations automatically. Falls back to hardcoded paths when contentCatalog is unavailable (e.g., during testing or preview builds). Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Tests cover: - Relative URL calculation from connector pages to whats-new - Cloud vs self-managed context page spec generation - Fallback behavior when contentCatalog unavailable - Resolution behavior when contentCatalog available - Macro registration without errors Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Add unified-navigation.js extension for Data Platform unified sidebar - Update ROOT → streaming component references across extensions - Update redpanda-labs → labs component references - Changes in: algolia-indexer, find-related-docs, find-related-labs, and other extensions Part of unified navigation implementation.
Contributor
|
Important Review skippedAuto incremental reviews are disabled on this repository. Please check the settings in the CodeRabbit UI or the ⚙️ Run configurationConfiguration used: Organization UI Review profile: CHILL Plan: Pro Run ID: You can disable this status message by setting the Use the checkbox below for a quick retry:
📝 WalkthroughWalkthroughThis PR modernizes component naming across the documentation extensions ecosystem while introducing a major new unified-navigation feature. Component names are standardized: Estimated code review effort🎯 3 (Moderate) | ⏱️ ~25 minutes Possibly related PRs
Suggested reviewers
🚥 Pre-merge checks | ✅ 4 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (4 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
✅ Deploy Preview for docs-extensions-and-macros ready!
To edit notification comments on pull requests, go to your Netlify project configuration. |
Contributor
Author
|
Jira: DOC-2180 |
Major changes:
- Rename page-header-data to component-metadata throughout
- Read navigation config from page-navigation attribute
- Support hierarchical navigation structure from config
- Compute breadcrumb hierarchy from navigation tree
- Attach page-breadcrumb-hierarchy to pages
- Remove section-based discovery (use explicit config)
- Support version detection for multi-version components
- Fallback to standard Antora nav when no config present
Config format in antora.yml:
component-metadata:
title: "Name"
color: "#hex"
icon: "icon-name"
order: 10
page-navigation:
- component-name
- parent: [child1, child2]
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
Contributor
There was a problem hiding this comment.
Actionable comments posted: 4
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In `@extensions/unified-navigation.js`:
- Line 90: Remove the noisy console.log call and the streaming-specific
logger.warn in extensions/unified-navigation.js: replace console.log(`[config]
Storing ${component.name}'s OWN config (root items:
${Array.from(allComponentsInConfig).join(', ')})`) with the extension's logger
at a trace/debug level (e.g., logger.debug or logger.trace) or remove it
entirely, and change any streaming home-page warning (logger.warn) to
logger.debug/trace so these messages can be silenced via log level; look for
other similar console.log calls and replace or remove them in the
config-resolution and filter helper code paths to avoid uncontrolled build
output.
- Around line 202-204: The summary log for processed config-driven pages is
using logger.warn which incorrectly signals a warning; change the call in the
block that checks configDrivenPages.size (the
Array.from(configDrivenPages).join(...) message) to use logger.info (or
logger.debug if you prefer lower verbosity) so the end-of-run summary is
informational rather than a warning, keeping the same message text and
interpolation.
- Around line 245-266: filterUnpublishedPages currently mutates Antora's shared
navigation by writing back into item.items; instead create and return a new tree
so you don't modify the original objects. Update filterUnpublishedPages (and its
recursive call) to map each item to a new object: if item.url and
publishedUrlsSet.has(item.url) keep a shallow copy of item but with items
replaced by the result of recursively calling filterUnpublishedPages on the
original item.items (when Array.isArray(item.items)); skip (return null) for
unpublished URLs and finally filter out nulls — never assign into item.items
in-place.
- Around line 320-326: The parseNavigationConfig function currently uses
JSON.parse to parse string navConfig which fails for the documented YAML format;
replace JSON.parse(navConfig) with yaml.load(navConfig) from the existing
js-yaml dependency (and add/import the yaml symbol if missing) so YAML strings
like "- streaming" are parsed correctly and still fallback to returning [] when
not an array.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Organization UI
Review profile: CHILL
Plan: Pro
Run ID: ab59c6c9-485f-4464-9dd2-887bd784c666
⛔ Files ignored due to path filters (1)
package-lock.jsonis excluded by!**/package-lock.json
📒 Files selected for processing (13)
__tests__/macros/rp-connect-components.test.jsextensions/algolia-indexer/generate-index.jsextensions/collect-bloblang-samples.jsextensions/find-related-docs.jsextensions/find-related-labs.jsextensions/generate-fields-only-pages.jsextensions/generate-rp-connect-categories.jsextensions/generate-rp-connect-info.jsextensions/unified-navigation.jsextensions/unlisted-pages.jsmacros/rp-connect-components.jspackage.jsontools/redpanda-connect/rpcn-connector-docs-handler.js
| configOwner: component.name, | ||
| allComponents: allComponentsInConfig, | ||
| }) | ||
| console.log(`[config] Storing ${component.name}'s OWN config (root items: ${Array.from(allComponentsInConfig).join(', ')})`) |
Contributor
There was a problem hiding this comment.
Remove debug console.log calls and the streaming-specific debug warning.
console.log statements scattered across the config-resolution and filter helpers (and the targeted logger.warn for the streaming home page) look like leftover debugging that will flood the Antora build output on every run, with no way to silence them. They should go through the extension's logger at debug/trace level, or be removed.
🧹 Suggested cleanup
- console.log(`[config] Storing ${component.name}'s OWN config (root items: ${Array.from(allComponentsInConfig).join(', ')})`)
+ logger.debug(`Storing ${component.name}'s OWN config (root items: ${Array.from(allComponentsInConfig).join(', ')})`)
@@
- console.log(`[config] Updating ${component.name}'s OWN config (better depth)`)
+ logger.debug(`Updating ${component.name}'s OWN config (better depth)`)
@@
- if (page.src.component === 'streaming' && page.src.path.includes('home/index')) {
- logger.warn(
- `Streaming home page: filtered tree has ${relevantTree.length} root items: ${relevantTree.map((t) => t.name).join(', ')}`
- )
- }
@@
- const treeNames = configTree.map(t => t.name).join(', ')
- console.log(`[filterConfigTreeForComponent] Called for ${componentName}, root items: [${treeNames}]`)
@@
- console.log(`[filterConfigTreeForComponent] ${componentName}: FOUND at root level, hasChildren: ${!!item.children}`)
@@
- console.log(`[filterConfigTreeForComponent] ${componentName}: returning parent with showNavItemsOnly + ${item.children.length} children`)
@@
- console.log(`[filterConfigTreeForComponent] ${componentName}: no children, returning original tree`)
@@
- console.log(`[filterConfigTreeForComponent] ${componentName}: NOT found at root, searching nested...`)Also applies to: 102-102, 155-159, 396-396, 401-401, 406-406, 409-409, 414-414
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In `@extensions/unified-navigation.js` at line 90, Remove the noisy console.log
call and the streaming-specific logger.warn in extensions/unified-navigation.js:
replace console.log(`[config] Storing ${component.name}'s OWN config (root
items: ${Array.from(allComponentsInConfig).join(', ')})`) with the extension's
logger at a trace/debug level (e.g., logger.debug or logger.trace) or remove it
entirely, and change any streaming home-page warning (logger.warn) to
logger.debug/trace so these messages can be silenced via log level; look for
other similar console.log calls and replace or remove them in the
config-resolution and filter helper code paths to avoid uncontrolled build
output.
Comment on lines
+202
to
+204
| if (configDrivenPages.size > 0) { | ||
| logger.warn(`Processed config-driven navigation for ${configDrivenPages.size} components: ${Array.from(configDrivenPages).join(', ')}`) | ||
| } |
Contributor
There was a problem hiding this comment.
Use an informational log level for the per-build summary.
This is a normal end-of-run summary, not a warning condition, but it's emitted at warn. In Antora pipelines that fail-on-warning or surface warning counts, this will produce noise and false positives on every build. Demote to info/debug.
- if (configDrivenPages.size > 0) {
- logger.warn(`Processed config-driven navigation for ${configDrivenPages.size} components: ${Array.from(configDrivenPages).join(', ')}`)
- }
+ if (configDrivenPages.size > 0) {
+ logger.info(`Processed config-driven navigation for ${configDrivenPages.size} components: ${Array.from(configDrivenPages).join(', ')}`)
+ }📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
Suggested change
| if (configDrivenPages.size > 0) { | |
| logger.warn(`Processed config-driven navigation for ${configDrivenPages.size} components: ${Array.from(configDrivenPages).join(', ')}`) | |
| } | |
| if (configDrivenPages.size > 0) { | |
| logger.info(`Processed config-driven navigation for ${configDrivenPages.size} components: ${Array.from(configDrivenPages).join(', ')}`) | |
| } |
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In `@extensions/unified-navigation.js` around lines 202 - 204, The summary log for
processed config-driven pages is using logger.warn which incorrectly signals a
warning; change the call in the block that checks configDrivenPages.size (the
Array.from(configDrivenPages).join(...) message) to use logger.info (or
logger.debug if you prefer lower verbosity) so the end-of-run summary is
informational rather than a warning, keeping the same message text and
interpolation.
Comment on lines
+245
to
+266
| function filterUnpublishedPages(items, publishedUrlsSet) { | ||
| if (!Array.isArray(items)) return [] | ||
|
|
||
| return items | ||
| .map((item) => { | ||
| // If item has a URL, check if it's in the published URLs set | ||
| if (item.url) { | ||
| // Skip items whose pages aren't published | ||
| if (!publishedUrlsSet.has(item.url)) { | ||
| return null | ||
| } | ||
| } | ||
|
|
||
| // Recursively filter children | ||
| if (item.items && Array.isArray(item.items)) { | ||
| item.items = filterUnpublishedPages(item.items, publishedUrlsSet) | ||
| } | ||
|
|
||
| return item | ||
| }) | ||
| .filter(Boolean) // Remove null entries | ||
| } |
Contributor
There was a problem hiding this comment.
filterUnpublishedPages mutates Antora's shared navigation in place.
item.items = filterUnpublishedPages(item.items, publishedUrlsSet) on line 260 writes back into the navigation objects that came from version.navigation via componentVersionNavMap. The same navigation array is reused for every page of the same component/version (and is Antora's internal structure), so this side-effects shared state — affecting subsequent navigation consumers (other extensions, the standard sidebar, breadcrumb logic) and making the function order-dependent. Return a new tree instead of mutating in place.
🔒 Suggested fix (non-mutating)
function filterUnpublishedPages(items, publishedUrlsSet) {
if (!Array.isArray(items)) return []
return items
- .map((item) => {
- // If item has a URL, check if it's in the published URLs set
- if (item.url) {
- // Skip items whose pages aren't published
- if (!publishedUrlsSet.has(item.url)) {
- return null
- }
- }
-
- // Recursively filter children
- if (item.items && Array.isArray(item.items)) {
- item.items = filterUnpublishedPages(item.items, publishedUrlsSet)
- }
-
- return item
- })
- .filter(Boolean) // Remove null entries
+ .map((item) => {
+ if (item.url && !publishedUrlsSet.has(item.url)) {
+ return null
+ }
+ if (item.items && Array.isArray(item.items)) {
+ return { ...item, items: filterUnpublishedPages(item.items, publishedUrlsSet) }
+ }
+ return { ...item }
+ })
+ .filter(Boolean)
}📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
Suggested change
| function filterUnpublishedPages(items, publishedUrlsSet) { | |
| if (!Array.isArray(items)) return [] | |
| return items | |
| .map((item) => { | |
| // If item has a URL, check if it's in the published URLs set | |
| if (item.url) { | |
| // Skip items whose pages aren't published | |
| if (!publishedUrlsSet.has(item.url)) { | |
| return null | |
| } | |
| } | |
| // Recursively filter children | |
| if (item.items && Array.isArray(item.items)) { | |
| item.items = filterUnpublishedPages(item.items, publishedUrlsSet) | |
| } | |
| return item | |
| }) | |
| .filter(Boolean) // Remove null entries | |
| } | |
| function filterUnpublishedPages(items, publishedUrlsSet) { | |
| if (!Array.isArray(items)) return [] | |
| return items | |
| .map((item) => { | |
| if (item.url && !publishedUrlsSet.has(item.url)) { | |
| return null | |
| } | |
| if (item.items && Array.isArray(item.items)) { | |
| return { ...item, items: filterUnpublishedPages(item.items, publishedUrlsSet) } | |
| } | |
| return { ...item } | |
| }) | |
| .filter(Boolean) | |
| } |
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In `@extensions/unified-navigation.js` around lines 245 - 266,
filterUnpublishedPages currently mutates Antora's shared navigation by writing
back into item.items; instead create and return a new tree so you don't modify
the original objects. Update filterUnpublishedPages (and its recursive call) to
map each item to a new object: if item.url and publishedUrlsSet.has(item.url)
keep a shallow copy of item but with items replaced by the result of recursively
calling filterUnpublishedPages on the original item.items (when
Array.isArray(item.items)); skip (return null) for unpublished URLs and finally
filter out nulls — never assign into item.items in-place.
Comment on lines
+320
to
+326
| function parseNavigationConfig(navConfig) { | ||
| // If it's already an array (from YAML parsing), use it directly | ||
| const config = typeof navConfig === 'string' ? JSON.parse(navConfig) : navConfig | ||
|
|
||
| if (!Array.isArray(config)) { | ||
| return [] | ||
| } |
Contributor
There was a problem hiding this comment.
🧩 Analysis chain
🏁 Script executed:
#!/bin/bash
# Check how page-navigation is actually supplied in this repo / fixtures and whether js-yaml is already used.
rg -nP -C2 "page-navigation" --type=yml --type=yaml --type=js
rg -nP "require\(['\"]js-yaml['\"]\)" --type=jsRepository: redpanda-data/docs-extensions-and-macros
Length of output: 977
🏁 Script executed:
# First, read the unified-navigation.js file to see the docstring, catch block, and full context
cat -n extensions/unified-navigation.js | head -330 | tail -100Repository: redpanda-data/docs-extensions-and-macros
Length of output: 3734
🏁 Script executed:
# Search for page-navigation without restrictive type filters
rg -n "page-navigation" -C3Repository: redpanda-data/docs-extensions-and-macros
Length of output: 4295
🏁 Script executed:
# Check if there are tests for this function
fd -i "test\|spec" -type f | xargs rg -l "parseNavigationConfig\|page-navigation" 2>/dev/null | head -20Repository: redpanda-data/docs-extensions-and-macros
Length of output: 255
JSON.parse cannot parse the YAML format documented in the docstring.
The function documentation (lines 301–302) and example (lines 305–309) specify YAML input format, but line 322 uses JSON.parse() which only handles JSON. Any YAML string attribute (e.g., - streaming) will fail JSON parsing and be silently caught by the error handler at line 134, disabling navigation for that component without obvious indication.
Use yaml.load() from js-yaml (already a project dependency) instead of JSON.parse() to handle the documented YAML format correctly.
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In `@extensions/unified-navigation.js` around lines 320 - 326, The
parseNavigationConfig function currently uses JSON.parse to parse string
navConfig which fails for the documented YAML format; replace
JSON.parse(navConfig) with yaml.load(navConfig) from the existing js-yaml
dependency (and add/import the yaml symbol if missing) so YAML strings like "-
streaming" are parsed correctly and still fallback to returning [] when not an
array.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Add fallback indexing for landing pages (component-home-v3, data-platform layouts) - Add special tag handling for umbrella components (Home, Data Platform, Self-Managed) - Rename Self-Managed references to Streaming - Filter umbrella components from home page tags - Index landing pages using page metadata (title, description) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
… connect Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
micheleRP
reviewed
May 22, 2026
There was a problem hiding this comment.
Review
Critical
-
In-place mutation of Antora's shared navigation —
extensions/unified-navigation.js:258.filterUnpublishedPageswrites back viaitem.items = filterUnpublishedPages(...)into the sameversion.navigationarrays held bycomponentVersionNavMap, which is reused for every page of the component. Subsequent pages see the already-filtered (possibly empty) tree; other nav consumers see the mutation too. Fix: return a new tree, e.g.:return { ...item, items: filterUnpublishedPages(item.items, publishedUrlsSet) };
-
Streaming-specific debug
logger.warnleft in production —extensions/unified-navigation.js:153-156. Cleanup commit 10 removed twoconsole.logs but kept theif (page.src.component === 'streaming' && page.src.path.includes('home/index')) { logger.warn(...) }
block. Hardcoded component name + spurious WARN per build. Fix: delete the block or drop to
logger.debug.
Suggestions
- Summary log emitted at
warn—extensions/unified-navigation.js:201. Demote toinfo; it isn't a warning, and will inflate warning counts on every build. - No tests for the new extension.
__tests__/extensions/covers four other extensions but not this one.parseNavigationConfig,getComponentDepth,isStandalone,filterConfigTreeForComponent,findComponentPathare all pure and trivially testable. - Stray
logger.warndebug instrumentation inextensions/unpublish-pages.js(lines 19-25, 35) — five lines per page withpublish-only-during-betaplus one per unpublished page. Reads like leftover ADP beta debug. parseNavigationConfigstring branch is dead — input is already parsed by Antora's YAML loader; theJSON.parsepath doesn't fire.- Hardcoded umbrella names in Algolia indexer —
['home', 'data platform', 'self-managed']+['home', 'component-home-v3', 'data-platform']layouts. Extract to constants.
CodeRabbit triage (4 findings)
- Keep: console.log + streaming-specific warn (cleanup commit 10 fixed two of the three; the streaming-home
logger.warnblock remains). Critical #2 above. - Keep: Summary log at
warn. Suggestion #1 above. Demote toinfo. - Keep:
filterUnpublishedPagesmutates shared state. Most important finding. Critical #1 above. - Skip:
page-navigationYAML /js-yamlanalysis — input is already parsed by Antora's YAML loader, so theJSON.parsepath doesn't fire. Dead code, not missing functionality.
Coordination risk
- Breadcrumb computation depends on a component literally named
'home'(extensions/unified-navigation.js:138,componentConfigs.get('home')). If the docs-site playbook doesn't register an umbrella named exactlyhome, breadcrumbs silently no-op. (Verified — docs-site redpanda-data/docs-site#173 does registerhomeas a component.) - Component-name updates across the eight modified extensions are consistent with the rollout (
ROOT → streaming,redpanda-labs → labs).
See docs-site redpanda-data/docs-site#173 for cross-rollout notes (suggested approval order + verification steps).
micheleRP
approved these changes
May 22, 2026
micheleRP
left a comment
micheleRP
left a comment
There was a problem hiding this comment.
Approved on the assumption that the 2 Critical items in the review comment above are addressed before merge: (1) fix the filterUnpublishedPages in-place mutation at line 258 (return a new tree instead of writing back into the shared componentVersionNavMap), (2) delete or logger.debug-demote the streaming-specific debug logger.warn at lines 153-156. CodeRabbit triage + 5 non-blocking suggestions in the comment.
When agentic-data-plane component is not included in the build, this extension automatically redirects the home page to the data-platform landing page. Supports both static HTML redirects and Netlify _redirects file. Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
- Replace console.log/console.error with Antora logger in 8 extensions - Add logger parameter support to version-fetcher utilities for proper logging - Update component references from 'redpanda-connect' to 'connect' - Add conditional-home-redirect to package.json exports - Fix test expectations for renamed component All 379 tests passing. Logging now respects Antora log levels (info, warn, error, debug). Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
- Change streaming home-page logger.warn to logger.debug for less noise - Change config-driven navigation summary from logger.warn to logger.info - Fix filterUnpublishedPages to create new tree instead of mutating original - Replace JSON.parse with yaml.load in parseNavigationConfig for YAML support - Add js-yaml import for proper YAML parsing All 379 tests passing. Navigation tree is now immutable and YAML configs work correctly. Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
Critical fixes:
- Fix filterUnpublishedPages to create new tree without mutating shared state
Previously shallow copy still allowed mutation of componentVersionNavMap
Now uses { ...item, items: recursiveCall() } pattern to ensure immutability
- Remove hardcoded streaming-specific debug block (lines 155-159)
Production code should not have component-specific debug instrumentation
Additional cleanup:
- Change unpublish-pages debug logging from logger.warn to logger.debug (lines 20-25)
- Change unpublish action from logger.warn to logger.info (line 38)
Reduces log noise - only actual unpublish actions are info-level
All 379 tests passing. Navigation tree mutation fixed, debug spam eliminated.
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
Fix: llms.adoc and sitemap pages should not trigger unlisted page warnings Problem: - llms.adoc is flagged as "unlisted page" even though it's intentionally not in navigation - This page is converted to llms.txt at site root by convert-llms-to-txt extension - The unpublish happens on 'beforePublish' event, after 'navigationBuilt' when unlisted-pages runs Solution: - Add exclusion list for technical pages that are intentionally unlisted - Skip pages with stems: llms, sitemap, sitemap-llms - These pages serve technical purposes (AI consumption, search engines) and shouldn't be in navigation Result: No more spurious warnings for technical documentation artifacts All 379 tests passing. Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
Changed 404 errors when fetching Chart.yaml to WARN level instead of ERROR, with a clearer message explaining that the branch may not exist yet for unreleased operator versions. This is expected behavior and not a critical error. Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
1. Filter large data attributes from markdown frontmatter
- Added blocklist in convert-to-markdown.js to exclude massive objects
- Prevents categories, csvData, and other component-level data from
appearing in every page's frontmatter
- Reduces markdown file sizes significantly for LLM consumption
2. Prepend component nav items in unified navigation
- Data platform pages now show their own nav items (Overview, etc.)
above child component buckets
- Fixes missing navigation items for umbrella components
3. Improve operator version error messaging
- Better warning when Docker Hub has version but GitHub branch doesn't exist
- Clarifies this is expected for unreleased versions
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
The previous fix prepended raw nav items to the buckets array, causing the UI to render them with an empty bucket wrapper. Now wrapping parent component's nav items in a pseudo-bucket with showNavItemsOnly: true, which renders items directly without bucket header/wrapper. This makes data-platform's Overview and Install MCP Server links appear as regular nav items above the child component buckets. Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
Major version bump for breaking changes: - Component rename: redpanda-connect → connect - Unified navigation architecture - Markdown frontmatter filtering Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
1. Fix duplication bug where child component nav items appeared twice
- Only top-level parent components (with children) should show their
own nav items above child buckets
- Child components (like streaming) should only show child buckets
- Checks if component appears in relevantTree with children
2. Add better error handling for YAML parsing
- Catch and log YAML parse errors with context
- Helps debug "Unexpected token ':'" errors
Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
Moved page-support-months into attrs object where it belongs. The previous conflict resolution had it as a stray line after Object.assign. Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
The isTopLevelParent check was looking at relevantTree (filtered for current component) instead of the original configTree. This caused it to miss data-platform's children, hiding its nav items. Now uses componentHasChildren() helper to check the full configTree, which correctly identifies data-platform as having children (cloud-data-platform, self-managed) and shows its nav items (Overview, Install MCP Server). Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
Only show component nav items outside buckets if BOTH conditions are true: 1. Component has children (defines child buckets) 2. Component owns the config (defined page-navigation in its own antora.yml) This fixes: - data-platform: has children + owns config → shows nav items ✓ - self-managed: has children but inherits config → nav items stay in bucket ✓ - streaming: no children + inherits config → no duplication ✓ Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
When a component has children and owns its config (like data-platform or self-managed), its nav items should appear outside buckets, not inside. Before: relevantTree included parent + children, buildBucketsFromConfig created buckets for ALL including parent → duplication After: When hasChildren && ownsConfig, extract only children for bucket building, parent's nav items shown outside buckets only Results: - data-platform: nav items above buckets ✓ - self-managed: nav items above buckets, no duplication ✓ - streaming: no duplication ✓ Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
Tests added: ✓ Components without page-navigation use standard nav ✓ Child components don't duplicate nav items ✓ Standalone components use standard nav ✗ Parent component nav items (needs debugging) ✗ Unpublished page filtering (needs setup fix) The parent component test is failing because treeForBuckets extraction isn't working correctly - still includes parent in buckets. Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
Added logs to trace: - Which children are extracted for bucket building - Whether parent nav items condition passes - How many nav items are found and added This will help diagnose why data-platform nav items aren't showing and why self-managed has duplicate Overview. Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
The bug: filterConfigTreeForComponent sets children: undefined on parent nodes, so we couldn't extract children from the filtered relevantTree. The fix: Extract children from the original configData.configTree instead of the filtered tree. Added findNodeInTree() helper to recursively find nodes in the tree. Simplified componentHasChildren() to use findNodeInTree(). This should fix: - data-platform: nav items now extracted properly - self-managed: nav items now extracted properly - No more "couldn't extract them from relevantTree" warnings Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
Cleaned up excessive debug logs added during troubleshooting. Kept only essential error/warning logs. Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
The test was using a standalone component config which doesn't trigger custom navigation. Updated to use a parent-child config so custom nav is applied and unpublished page filtering can be tested. Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Adds the unified navigation extension for Data Platform unified sidebar and updates component references across all extensions.
Changes
New Extension
Component Reference Updates
Updates component names across all extensions to align with renamed components:
ROOT→streamingredpanda-labs→labsModified Extensions
Part of unified navigation implementation.
Related PRs