@redpanda-data/docs-extensions-and-macros 4.10.7 → 4.11.0

package/extensions/convert-to-markdown.js

@@ -0,0 +1,373 @@
+const path = require('path')
+const os = require('os')
+const TurndownService = require('turndown')
+const turndownPluginGfm = require('turndown-plugin-gfm')
+const { gfm } = turndownPluginGfm
+
+module.exports.register = function () {
+  const logger = this.getLogger('convert-to-markdown-extension')
+  let playbook
+
+  // Shared Turndown configuration
+  const baseConfig = {
+    headingStyle: 'atx',
+    codeBlockStyle: 'fenced',
+    bulletListMarker: '-',
+    linkReferenceStyle: 'full',
+  }
+
+  // Factory: create a configured Turndown instance
+  function createTurndownBase() {
+    const td = new TurndownService(baseConfig)
+    td.use(gfm)
+
+    // Remove unwanted global elements (footers, modals, feedback, etc.)
+    td.addRule('remove-unwanted', {
+      filter: (node) => {
+        if (!node || !node.getAttribute) return false
+
+        const classAttr = (node.getAttribute('class') || '').toLowerCase()
+        const idAttr = (node.getAttribute('id') || '').toLowerCase()
+        const tag = node.nodeName.toLowerCase()
+
+        // Remove by tag
+        if (['script', 'style', 'footer', 'nav'].includes(tag)) return true
+
+        // Remove tracking or hidden images
+        if (
+          tag === 'img' &&
+          (classAttr.includes('tracking') ||
+            idAttr.includes('scarf') ||
+            node.getAttribute('role') === 'presentation' ||
+            node.style?.display === 'none')
+        ) {
+          return true
+        }
+
+        // Remove by class or id
+        const toRemove = [
+          'thumbs',
+          'back-to-top',
+          'contributors-modal',
+          'feedback-section',
+          'feedback-toast',
+          'pagination',
+          'footer',
+          'nav-expand',
+          'banner-container'
+        ]
+        return toRemove.some(
+          (x) => classAttr.includes(x) || idAttr.includes(x)
+        )
+      },
+      replacement: () => '',
+    })
+
+    // Keep critical content blocks only
+    td.keep(['div.openblock.tabs', 'article.doc'])
+    return td
+  }
+
+  // Factory: create page-specific Turndown converter
+  function createTurndownForPage(page) {
+    const outerTurndown = createTurndownBase()
+    const nestedTurndown = createTurndownBase()
+
+    // Helper to add custom rules
+    function addCustomRules(turndownInstance, isInner = false) {
+      // Determine heading depth for tab conversion
+      function findNearestHeadingLevel(el) {
+        let current = el.previousElementSibling
+        while (current) {
+          if (/^H[1-6]$/i.test(current.nodeName))
+            return parseInt(current.nodeName.substring(1))
+          current = current.previousElementSibling
+        }
+        let parent = el.parentElement
+        while (parent) {
+          const headings = Array.from(
+            parent.querySelectorAll('h1,h2,h3,h4,h5,h6')
+          )
+          if (headings.length > 0) {
+            const last = headings[headings.length - 1]
+            return parseInt(last.nodeName.substring(1))
+          }
+          parent = parent.parentElement
+        }
+        return 2
+      }
+
+      // Asciidoctor tab conversion
+      turndownInstance.addRule('asciidoctor-tabs', {
+        filter: (node) => {
+          if (node.nodeName !== 'DIV') return false
+          const classAttr = node.getAttribute?.('class') || node.className || ''
+          return classAttr.includes('openblock') && classAttr.includes('tabs')
+        },
+        replacement: function (_, node) {
+          function processTabGroup(group, parentHeadingLevel = null) {
+            const contentDiv = group.querySelector('.content') || group
+            const tabList = contentDiv.querySelectorAll('li.tab')
+            if (!tabList.length) return ''
+
+            const nearestLevel =
+              parentHeadingLevel != null
+                ? parentHeadingLevel + 1
+                : findNearestHeadingLevel(group) + 1
+            const tabHeadingLevel = Math.min(nearestLevel, 6)
+            const headingPrefix = '#'.repeat(tabHeadingLevel)
+
+            let markdown = ''
+            tabList.forEach((tab) => {
+              const title =
+                tab.querySelector('p')?.textContent.trim() ||
+                tab.textContent.trim()
+
+              let panelId = tab.getAttribute('aria-controls')
+              if (!panelId && tab.id) panelId = tab.id + '--panel'
+              const panel = group.querySelector(`#${panelId}`)
+              if (!panel) return
+
+              const nestedTabs = panel.querySelectorAll('.openblock.tabs')
+              let nestedMdCombined = ''
+              nestedTabs.forEach((nested) => {
+                nestedMdCombined +=
+                  '\n' + processTabGroup(nested, tabHeadingLevel) + '\n'
+                nested.remove()
+              })
+
+              const innerHtml = panel.innerHTML || ''
+              let md = ''
+              try {
+                const converter = isInner ? nestedTurndown : turndownInstance
+                md = converter.turndown(innerHtml)
+              } catch (e) {
+                logger.warn(`Turndown failed in nested tab: ${e.message}`)
+              }
+
+              markdown += `${headingPrefix} ${title}\n\n${md.trim()}\n${nestedMdCombined.trim()}\n\n`
+            })
+
+            return markdown.trim()
+          }
+
+          return '\n' + processTabGroup(node, null) + '\n'
+        },
+      })
+
+      // Admonition block conversion
+      turndownInstance.addRule('admonition', {
+        filter: (node) =>
+          node.nodeName === 'TABLE' &&
+          node.querySelector('td.icon') &&
+          node.querySelector('td.content'),
+        replacement: function (_, node) {
+          const iconCell = node.querySelector('td.icon')
+          const contentCell = node.querySelector('td.content')
+          if (!iconCell || !contentCell) return ''
+
+          const iconEl = iconCell.querySelector('i')
+          const classAttr = iconEl?.className || ''
+          const match = classAttr.match(/icon-([a-z]+)/i)
+          const type = match ? match[1].toUpperCase() : 'NOTE'
+
+          const titleEl =
+            node.querySelector('.title') ||
+            contentCell.querySelector('.title') ||
+            iconEl?.getAttribute('title')
+          const customTitle =
+            typeof titleEl === 'string'
+              ? titleEl.trim()
+              : titleEl?.textContent?.trim() || ''
+
+          const emojiMap = {
+            CAUTION: '⚠️',
+            WARNING: '⚠️',
+            TIP: '💡',
+            NOTE: '📝',
+            IMPORTANT: '❗',
+          }
+          const emoji = emojiMap[type] || '📘'
+
+          const innerHtml = contentCell.innerHTML || ''
+          let innerMd = ''
+          try {
+            const converter = isInner ? nestedTurndown : turndownInstance
+            innerMd = converter.turndown(innerHtml).trim()
+          } catch (e) {
+            logger.warn(`Turndown failed in admonition: ${e.message}`)
+          }
+
+          const titleLower = customTitle.toLowerCase()
+          const typeLower = type.toLowerCase()
+          const header =
+            customTitle && titleLower !== typeLower
+              ? `${emoji} **${type}: ${customTitle}**`
+              : `${emoji} **${type}**`
+
+          const quoted = innerMd
+            .split('\n')
+            .map((line) => (line.startsWith('>') ? line : `> ${line}`))
+            .join('\n')
+
+          return `\n> ${header}\n>\n${quoted}\n`
+        },
+      })
+
+      // Markdown table conversion
+      turndownInstance.addRule('tables', {
+        filter: (node) => {
+          if (node.nodeName !== 'TABLE') return false
+          if (node.querySelector('td.icon') && node.querySelector('td.content'))
+            return false
+          return true
+        },
+        replacement: function (content, node) {
+          const rows = Array.from(node.querySelectorAll('tr'))
+          if (!rows.length) return content
+          const tableRows = []
+          rows.forEach((row, index) => {
+            const cells = Array.from(row.querySelectorAll('th, td'))
+            const cellContents = cells.map((cell) =>
+              (cell.textContent || '').trim().replace(/\s+/g, ' ')
+            )
+            if (!cellContents.length) return
+            const rowLine = '| ' + cellContents.join(' | ') + ' |'
+            tableRows.push(rowLine)
+            if (index === 0) {
+              const separator =
+                '| ' + cellContents.map(() => '---').join(' | ') + ' |'
+              tableRows.push(separator)
+            }
+          })
+          return '\n' + tableRows.join('\n') + '\n'
+        },
+      })
+    }
+
+    addCustomRules(outerTurndown, false)
+    addCustomRules(nestedTurndown, true)
+    return outerTurndown
+  }
+
+  // Conversion pipeline
+  this.on('pagesComposed', async ({ playbook: pb, contentCatalog }) => {
+    playbook = pb
+    const siteUrl = playbook.site?.url || ''
+    const pages = contentCatalog.getPages()
+    logger.info(
+      `Converting ${pages.length} pages to Markdown${
+        siteUrl ? ` (site.url=${siteUrl})` : ''
+      }...`
+    )
+
+    const concurrency = Math.max(2, Math.floor(os.cpus().length / 2))
+    const queue = [...pages]
+    let convertedCount = 0
+
+    async function processQueue() {
+      while (queue.length) {
+        const page = queue.shift()
+        if (!page?.contents) continue
+
+        try {
+          const html = page.contents.toString().trim()
+          if (!html) continue
+
+          // Extract only the <article class="doc"> portion
+          const match = html.match(
+            /<article[^>]*class=["'][^"']*\bdoc\b[^"']*["'][^>]*>([\s\S]*?)<\/article>/i
+          )
+          if (!match || !match[1]) {
+            logger.info(`No <article class="doc"> found for ${page.src?.path}`)
+            continue
+          }
+          const articleHtml = match[1]
+
+          // Convert with Turndown
+          const td = createTurndownForPage(page)
+          let markdown = td.turndown(articleHtml).trim()
+
+          // Canonical source link
+          let canonicalUrl = ''
+          try {
+            if (siteUrl && page.pub?.url) {
+              const htmlStyle = playbook?.urls?.htmlExtensionStyle
+              const isIndexify = htmlStyle === 'indexify'
+              const baseUrl = new URL(page.pub.url, siteUrl)
+              let pathname = baseUrl.pathname
+
+              if (isIndexify) {
+                const looksLikeDir =
+                  pathname.endsWith('/') ||
+                  !path.basename(pathname).includes('.')
+                baseUrl.pathname = looksLikeDir
+                  ? pathname.replace(/\/?$/, '/index.md')
+                  : pathname.replace(/\.html$/, '.md')
+              } else {
+                baseUrl.pathname = pathname.replace(/\.html$/, '.md')
+              }
+
+              canonicalUrl = baseUrl.toString()
+            }
+          } catch (e) {
+            logger.debug(
+              `Failed to build canonical URL for ${page.src?.path}: ${e.message}`
+            )
+          }
+
+          // Prepend Markdown source reference and URL construction hint
+          if (canonicalUrl) {
+            const urlHint = `<!-- Note for AI: Links in this doc are relative to the current page and use indexify format. Add /index.md to directory-style links for the Markdown version. -->`
+            
+            markdown = `<!-- Source: ${canonicalUrl} -->\n${urlHint}\n\n${markdown}`
+          }
+
+          // Clean up unnecessary whitespace
+          if (markdown) {
+            // Remove excessive blank lines (more than 2 consecutive newlines)
+            markdown = markdown.replace(/\n{3,}/g, '\n\n')
+            // Remove trailing whitespace from lines
+            markdown = markdown.replace(/[ \t]+$/gm, '')
+            // Remove leading/trailing whitespace from the entire document
+            markdown = markdown.trim()
+          }
+
+          if (markdown) {
+            page.markdownContents = Buffer.from(markdown, 'utf8')
+            convertedCount++
+          }
+        } catch (err) {
+          logger.error(
+            `Error converting ${page.src?.path || 'unknown'}: ${err.message}`
+          )
+          logger.debug(err.stack)
+        }
+      }
+    }
+
+    const workers = Array.from({ length: concurrency }, processQueue)
+    await Promise.all(workers)
+    logger.info(`Converted ${convertedCount} Markdown files.`)
+  })
+
+  // Add Markdown files to site catalog
+  this.on('beforePublish', ({ siteCatalog, contentCatalog }) => {
+    const pages = contentCatalog.getPages((p) => p.markdownContents)
+    if (!pages.length) {
+      logger.info('No Markdown files to publish.')
+      return
+    }
+    logger.info(`Adding ${pages.length} Markdown files to site catalog...`)
+    for (const page of pages) {
+      const htmlOut = page.out?.path
+      if (!htmlOut) continue
+      const mdOutPath = htmlOut.replace(/\.html$/, '.md')
+      siteCatalog.addFile({
+        contents: page.markdownContents,
+        out: { path: mdOutPath },
+      })
+      logger.debug(`Added Markdown: ${mdOutPath}`)
+    }
+  })
+}

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@redpanda-data/docs-extensions-and-macros",
-  "version": "4.10.7",
+  "version": "4.11.0",
   "description": "Antora extensions and macros developed for Redpanda documentation.",
   "keywords": [
     "antora",
@@ -43,6 +43,7 @@
     "./extensions/process-context-switcher": "./extensions/process-context-switcher.js",
     "./extensions/archive-attachments": "./extensions/archive-attachments.js",
     "./extensions/add-pages-to-root": "./extensions/add-pages-to-root.js",
+    "./extensions/convert-to-markdown": "./extensions/convert-to-markdown.js",
     "./extensions/collect-bloblang-samples": "./extensions/collect-bloblang-samples.js",
     "./extensions/compute-end-of-life": "./extensions/compute-end-of-life.js",
     "./extensions/generate-rp-connect-categories": "./extensions/generate-rp-connect-categories.js",
@@ -85,8 +86,10 @@
     "@octokit/core": "^6.1.2",
     "@octokit/plugin-retry": "^7.1.1",
     "@octokit/rest": "^21.0.1",
+    "@redocly/cli": "^2.2.0",
     "algoliasearch": "^4.17.0",
     "chalk": "4.1.2",
+    "cheerio": "^1.1.2",
     "commander": "^14.0.0",
     "gulp": "^4.0.2",
     "gulp-connect": "^5.7.0",
@@ -103,9 +106,10 @@
     "sync-request": "^6.1.0",
     "tar": "^7.4.3",
     "tree-sitter": "^0.22.4",
+    "turndown": "^7.2.2",
+    "turndown-plugin-gfm": "^1.0.2",
     "yaml": "^2.7.1",
-    "yargs": "^17.7.2",
-    "@redocly/cli": "^2.2.0"
+    "yargs": "^17.7.2"
   },
   "devDependencies": {
     "@antora/cli": "3.1.4",

package/tools/property-extractor/generate-handlebars-docs.js

@@ -22,7 +22,7 @@ const helpers = require('./helpers');
  * CLI Usage: node generate-handlebars-docs.js <input-file> <output-dir>
  */
 
-// Register all helpers
+// Register helpers
 Object.entries(helpers).forEach(([name, fn]) => {
   if (typeof fn !== 'function') {
     console.error(`❌ Helper "${name}" is not a function`);
@@ -62,11 +62,6 @@ function getTemplatePath(defaultPath, envVar) {
 
 /**
  * Register Handlebars partials used to render property documentation.
- *
- * Registers:
- *  - "property"
- *  - "topic-property"
- *  - "deprecated-property"
  */
 function registerPartials() {
   const templatesDir = path.join(__dirname, 'templates');
@@ -112,7 +107,7 @@ function registerPartials() {
 /**
  * Generate consolidated AsciiDoc partials for properties grouped by type.
  */
-function generatePropertyPartials(properties, partialsDir) {
+function generatePropertyPartials(properties, partialsDir, onRender) {
   console.log(`📝 Generating consolidated property partials in ${partialsDir}…`);
 
   const propertyTemplate = handlebars.compile(
@@ -129,11 +124,24 @@ function generatePropertyPartials(properties, partialsDir) {
 
   Object.values(properties).forEach(prop => {
     if (!prop.name || !prop.config_scope) return;
-    if (prop.config_scope === 'topic') propertyGroups.topic.push(prop);
-    else if (prop.config_scope === 'broker') propertyGroups.broker.push(prop);
-    else if (prop.config_scope === 'cluster') {
-      if (isObjectStorageProperty(prop)) propertyGroups['object-storage'].push(prop);
-      else propertyGroups.cluster.push(prop);
+
+    switch (prop.config_scope) {
+      case 'topic':
+        propertyGroups.topic.push(prop);
+        break;
+      case 'broker':
+        propertyGroups.broker.push(prop);
+        break;
+      case 'cluster':
+        if (isObjectStorageProperty(prop)) propertyGroups['object-storage'].push(prop);
+        else propertyGroups.cluster.push(prop);
+        break;
+      case 'object-storage':
+        propertyGroups['object-storage'].push(prop);
+        break;
+      default:
+        console.warn(`⚠️ Unknown config_scope: ${prop.config_scope} for ${prop.name}`);
+        break;
     }
   });
 
@@ -143,9 +151,16 @@ function generatePropertyPartials(properties, partialsDir) {
     if (props.length === 0) return;
     props.sort((a, b) => String(a.name || '').localeCompare(String(b.name || '')));
     const selectedTemplate = type === 'topic' ? topicTemplate : propertyTemplate;
-    const content = props.map(p => selectedTemplate(p)).join('\n');
+    const pieces = [];
+    props.forEach(p => {
+      if (typeof onRender === 'function') {
+        try { onRender(p.name); } catch (err) { /* swallow callback errors */ }
+      }
+      pieces.push(selectedTemplate(p));
+    });
+    const content = pieces.join('\n');
     const filename = `${type}-properties.adoc`;
-  fs.writeFileSync(path.join(propertiesPartialsDir, filename), AUTOGEN_NOTICE + content, 'utf8');
+    fs.writeFileSync(path.join(propertiesPartialsDir, filename), AUTOGEN_NOTICE + content, 'utf8');
     console.log(`✅ Generated ${filename} (${props.length} properties)`);
     totalCount += props.length;
   });
@@ -178,19 +193,18 @@ function generateDeprecatedDocs(properties, outputDir) {
     clusterProperties: clusterProperties.length ? clusterProperties : null
   };
 
-  const output = template(data);
   const outputPath = process.env.OUTPUT_PARTIALS_DIR
     ? path.join(process.env.OUTPUT_PARTIALS_DIR, 'deprecated', 'deprecated-properties.adoc')
     : path.join(outputDir, 'partials', 'deprecated', 'deprecated-properties.adoc');
 
   fs.mkdirSync(path.dirname(outputPath), { recursive: true });
-  fs.writeFileSync(outputPath, AUTOGEN_NOTICE + output, 'utf8');
+  fs.writeFileSync(outputPath, AUTOGEN_NOTICE + template(data), 'utf8');
   console.log(`✅ Generated ${outputPath}`);
   return deprecatedProperties.length;
 }
 
 /**
- * Generate topic-property-mappings.adoc using the mappings template and topic properties.
+ * Generate topic-property-mappings.adoc
  */
 function generateTopicPropertyMappings(properties, partialsDir) {
   const templatesDir = path.join(__dirname, 'templates');
@@ -218,31 +232,42 @@ function generateTopicPropertyMappings(properties, partialsDir) {
 }
 
 /**
- * Generate error reports for missing descriptions and deprecated properties.
+ * Generate error reports for missing descriptions, deprecated, and undocumented properties.
  */
-function generateErrorReports(properties) {
+function generateErrorReports(properties, documentedProperties = []) {
   const emptyDescriptions = [];
   const deprecatedProperties = [];
+  const allKeys = Object.keys(properties);
 
-  Object.values(properties).forEach(p => {
-    if (!p.description || !p.description.trim()) emptyDescriptions.push(p.name);
-    if (p.is_deprecated) deprecatedProperties.push(p.name);
+  // Use documentedProperties array (property names that were rendered into partials)
+  const documentedSet = new Set(documentedProperties);
+  const undocumented = [];
+
+  Object.entries(properties).forEach(([key, p]) => {
+    const name = p.name || key;
+    if (!p.description || !p.description.trim()) emptyDescriptions.push(name);
+    if (p.is_deprecated) deprecatedProperties.push(name);
+    if (!documentedSet.has(name)) undocumented.push(name);
   });
 
-  const total = Object.keys(properties).length;
+  const total = allKeys.length;
   const pctEmpty = total ? ((emptyDescriptions.length / total) * 100).toFixed(2) : '0.00';
   const pctDeprecated = total ? ((deprecatedProperties.length / total) * 100).toFixed(2) : '0.00';
-  console.log(`Empty descriptions: ${emptyDescriptions.length} (${pctEmpty}%)`);
-  console.log(`Deprecated: ${deprecatedProperties.length} (${pctDeprecated}%)`);
+  const pctUndocumented = total ? ((undocumented.length / total) * 100).toFixed(2) : '0.00';
+
+  console.log(`📉 Empty descriptions: ${emptyDescriptions.length} (${pctEmpty}%)`);
+  console.log(`🕸️ Deprecated: ${deprecatedProperties.length} (${pctDeprecated}%)`);
+  console.log(`🚫 Not documented: ${undocumented.length} (${pctUndocumented}%)`);
 
   return {
     empty_descriptions: emptyDescriptions.sort(),
-    deprecated_properties: deprecatedProperties.sort()
+    deprecated_properties: deprecatedProperties.sort(),
+    undocumented_properties: undocumented.sort(),
   };
 }
 
 /**
- * Main generator — only supports partials and deprecated docs.
+ * Main generator
  */
 function generateAllDocs(inputFile, outputDir) {
   const data = JSON.parse(fs.readFileSync(inputFile, 'utf8'));
@@ -252,13 +277,14 @@ function generateAllDocs(inputFile, outputDir) {
 
   let partialsCount = 0;
   let deprecatedCount = 0;
+  const documentedProps = []; // Track which property names were rendered
 
   if (process.env.GENERATE_PARTIALS === '1' && process.env.OUTPUT_PARTIALS_DIR) {
     console.log('📄 Generating property partials and deprecated docs...');
     deprecatedCount = generateDeprecatedDocs(properties, outputDir);
-    partialsCount = generatePropertyPartials(properties, process.env.OUTPUT_PARTIALS_DIR);
 
-    // Generate topic-property-mappings.adoc
+    // Generate property partials using the shared helper and collect names via callback
+    partialsCount = generatePropertyPartials(properties, process.env.OUTPUT_PARTIALS_DIR, name => documentedProps.push(name));
     try {
       generateTopicPropertyMappings(properties, process.env.OUTPUT_PARTIALS_DIR);
     } catch (err) {
@@ -268,24 +294,34 @@ function generateAllDocs(inputFile, outputDir) {
     console.log('📄 Skipping partial generation (set GENERATE_PARTIALS=1 and OUTPUT_PARTIALS_DIR to enable)');
   }
 
-  const errors = generateErrorReports(properties);
-  const inputData = JSON.parse(fs.readFileSync(inputFile, 'utf8'));
-  inputData.empty_descriptions = errors.empty_descriptions;
-  inputData.deprecated_properties = errors.deprecated_properties;
-  fs.writeFileSync(inputFile, JSON.stringify(inputData, null, 2), 'utf8');
+  const errors = generateErrorReports(properties, documentedProps);
 
-  console.log('📊 Summary:');
-  console.log(`   Total properties: ${Object.keys(properties).length}`);
-  console.log(`   Total partials generated: ${partialsCount}`);
-  console.log(`   Deprecated properties: ${deprecatedCount}`);
+  const totalProperties = Object.keys(properties).length;
+  const notRendered = errors.undocumented_properties.length;
+  const pctRendered = totalProperties
+    ? ((partialsCount / totalProperties) * 100).toFixed(2)
+    : '0.00';
+
+  console.log('\n📊 Summary:');
+  console.log(`   Total properties found:      ${totalProperties}`);
+  console.log(`   Property partials generated: ${partialsCount} (${pctRendered}% of total)`);
+  console.log(`   Not documented:              ${notRendered}`);
+  console.log(`   Deprecated properties:       ${deprecatedCount}`);
+
+  if (notRendered > 0) {
+    console.log('⚠️ Undocumented properties:\n   ' + errors.undocumented_properties.join('\n   '));
+  }
 
   return {
-    totalProperties: Object.keys(properties).length,
-    propertyPartials: partialsCount,
-    deprecatedProperties: deprecatedCount
+    totalProperties,
+    generatedPartials: partialsCount,
+    undocumentedProperties: errors.undocumented_properties,
+    deprecatedProperties: deprecatedCount,
+    percentageRendered: pctRendered
   };
 }
 
+
 module.exports = {
   generateAllDocs,
   generateDeprecatedDocs,

package/tools/property-extractor/helpers/allTopicsConditional.js

@@ -0,0 +1,19 @@
+// Returns 'cloud' if all topics are cloud-only, 'self-managed' if all are self-managed-only, else 'normal'
+module.exports = function allTopicsConditional(related_topics) {
+  if (!Array.isArray(related_topics) || related_topics.length === 0) return null;
+  let allCloud = true;
+  let allSelfManaged = true;
+  for (const t of related_topics) {
+    if (typeof t !== 'string') {
+      allCloud = false;
+      allSelfManaged = false;
+      break;
+    }
+    const trimmed = t.trim();
+    if (!trimmed.startsWith('cloud-only:')) allCloud = false;
+    if (!trimmed.startsWith('self-managed-only:')) allSelfManaged = false;
+  }
+  if (allCloud) return 'cloud';
+  if (allSelfManaged) return 'self-managed';
+  return 'normal';
+};

package/tools/property-extractor/helpers/formatPropertyValue.js

@@ -28,6 +28,10 @@ function processDefaults(inputString, suffix) {
     return inputString;
   }
 
+  // Remove C++ digit separators (apostrophes) from numbers/durations
+  // e.g. 30'000ms -> 30000ms, 1'500 -> 1500
+  inputString = inputString.replace(/(?<=\d)'(?=\d)/g, '');
+
   // Test for ip:port in vector: std::vector<net::unresolved_address>({{...}})
   const vectorMatch = inputString.match(/std::vector<net::unresolved_address>\(\{\{("([\d.]+)",\s*(\d+))\}\}\)/);
   if (vectorMatch) {

package/tools/property-extractor/helpers/index.js

@@ -11,4 +11,6 @@ module.exports = {
   renderPropertyExample: require('./renderPropertyExample.js'),
   formatUnits: require('./formatUnits.js'),
   anchorName: require('./anchorName.js'),
+  parseRelatedTopic: require('./parseRelatedTopic.js'),
+  allTopicsConditional: require('./allTopicsConditional.js'),
 };

package/tools/property-extractor/helpers/parseRelatedTopic.js

@@ -0,0 +1,13 @@
+// Returns an object with type and value for a related topic
+// type: 'cloud', 'self-managed', or 'normal'
+module.exports = function parseRelatedTopic(topic) {
+  if (typeof topic !== 'string') return { type: 'normal', value: topic };
+  const trimmed = topic.trim();
+  if (trimmed.startsWith('cloud-only:')) {
+    return { type: 'cloud', value: trimmed.replace(/^cloud-only:/, '').trim() };
+  }
+  if (trimmed.startsWith('self-managed-only:')) {
+    return { type: 'self-managed', value: trimmed.replace(/^self-managed-only:/, '').trim() };
+  }
+  return { type: 'normal', value: trimmed };
+};

package/tools/property-extractor/property_extractor.py

@@ -1010,6 +1010,9 @@ def resolve_type_and_default(properties, definitions):
             processed (str): A string representing the JSON-ready value (for example: '"value"', 'null', '0', or the original input when no mapping applied).
         """
         arg_str = arg_str.strip()
+        # Remove C++ digit separators (apostrophes) that may appear in numeric literals
+        # Example: "30'000ms" -> "30000ms". Use conservative replace only between digits.
+        arg_str = re.sub(r"(?<=\d)'(?=\d)", '', arg_str)
         
         # Handle std::nullopt -> null
         if arg_str == "std::nullopt":
@@ -1023,9 +1026,27 @@ def resolve_type_and_default(properties, definitions):
             resolved_value = resolve_cpp_function_call(function_name)
             if resolved_value is not None:
                 return f'"{resolved_value}"'
+
+        # Handle std::chrono literals like std::chrono::minutes{5} -> "5min"
+        chrono_match = re.match(r'std::chrono::([a-zA-Z]+)\s*\{\s*(\d+)\s*\}', arg_str)
+        if chrono_match:
+            unit = chrono_match.group(1)
+            value = chrono_match.group(2)
+            unit_map = {
+                'hours': 'h',
+                'minutes': 'min',
+                'seconds': 's',
+                'milliseconds': 'ms',
+                'microseconds': 'us',
+                'nanoseconds': 'ns'
+            }
+            short = unit_map.get(unit.lower(), unit)
+            return f'"{value} {short}"'
         
-        # Handle enum-like patterns (such as fips_mode_flag::disabled -> "disabled")
-        enum_match = re.match(r'[a-zA-Z0-9_:]+::([a-zA-Z0-9_]+)', arg_str)
+        # Handle enum-like patterns (such as fips_mode_flag::disabled -> "disabled").
+        # Only treat bare 'X::Y' tokens as enums — do not match when the token
+        # is followed by constructor braces/parentheses (e.g. std::chrono::minutes{5}).
+        enum_match = re.match(r'[a-zA-Z0-9_:]+::([a-zA-Z0-9_]+)\s*$', arg_str)
         if enum_match:
             enum_value = enum_match.group(1)
             return f'"{enum_value}"'

package/tools/property-extractor/templates/property.hbs

@@ -23,6 +23,7 @@ endif::[]
 {{else}}
 
 No description available.
+
 {{/if}}
 {{#if is_enterprise}}
 
@@ -31,9 +32,11 @@ ifndef::env-cloud[]
 endif::[]
 {{/if}}
 {{#if cloud_byoc_only}}
+
 ifdef::env-cloud[]
 NOTE: This property is available only in Redpanda Cloud BYOC deployments.
 endif::[]
+
 {{/if}}
 {{#if units}}
 
@@ -50,6 +53,10 @@ endif::[]
 *Requires restart:* {{#if needs_restart}}Yes{{else}}No{{/if}}
 {{/if}}
 {{/if}}
+
+ifndef::env-cloud[]
+*Restored during xref:manage:whole-cluster-restore.adoc[Whole Cluster Restore]:* {{#if (ne gets_restored false)}}Yes{{else}}No{{/if}}
+endif::[]
 {{#if visibility}}
 
 // tag::self-managed-only[]
@@ -89,13 +96,49 @@ endif::[]
 {{{renderPropertyExample this}}}
 {{/if}}
 {{#if related_topics}}
+{{#with (allTopicsConditional related_topics) as |sectionType|}}
+
+{{#if (eq sectionType "cloud")}}
+ifdef::env-cloud[]
+*Related topics:*
+
+{{#each ../related_topics}}
+{{#with (parseRelatedTopic this)}}
+* {{{value}}}
+{{/with}}
+{{/each}}
+endif::[]
+{{else if (eq sectionType "self-managed")}}
+ifndef::env-cloud[]
+*Related topics:*
 
+{{#each ../related_topics}}
+{{#with (parseRelatedTopic this)}}
+* {{{value}}}
+{{/with}}
+{{/each}}
+endif::[]
+{{else}}
 *Related topics:*
 
-{{#each related_topics}}
-* {{{this}}}
+{{#each ../related_topics}}
+{{#with (parseRelatedTopic this)}}
+{{#if (eq type "cloud")}}
+ifdef::env-cloud[]
+* {{{value}}}
+endif::[]
+{{else if (eq type "self-managed")}}
+ifndef::env-cloud[]
+* {{{value}}}
+endif::[]
+{{else}}
+* {{{value}}}
+{{/if}}
+{{/with}}
 {{/each}}
 {{/if}}
+{{/with}}
+{{/if}}
 {{#if aliases}}
 
 // tag::self-managed-only[]

package/tools/property-extractor/templates/topic-property.hbs

@@ -74,17 +74,57 @@ endif::[]
 {{/if}}
 
 *Nullable:* {{#if nullable}}Yes{{else}}No{{/if}}
+
+ifndef::env-cloud[]
+*Restored during xref:manage:whole-cluster-restore.adoc[Whole Cluster Restore]:* {{#if (ne gets_restored false)}}Yes{{else}}No{{/if}}
+endif::[]
 {{#if example}}
 
 {{{renderPropertyExample this}}}
 {{/if}}
 {{#if related_topics}}
+{{#with (allTopicsConditional related_topics) as |sectionType|}}
 
+{{#if (eq sectionType "cloud")}}
+ifdef::env-cloud[]
 *Related topics:*
 
-{{#each related_topics}}
-* {{{this}}}
+{{#each ../related_topics}}
+{{#with (parseRelatedTopic this)}}
+* {{{value}}}
+{{/with}}
 {{/each}}
+endif::[]
+{{else if (eq sectionType "self-managed")}}
+ifndef::env-cloud[]
+*Related topics:*
+
+{{#each ../related_topics}}
+{{#with (parseRelatedTopic this)}}
+* {{{value}}}
+{{/with}}
+{{/each}}
+endif::[]
+{{else}}
+*Related topics:*
+
+{{#each ../related_topics}}
+{{#with (parseRelatedTopic this)}}
+{{#if (eq type "cloud")}}
+ifdef::env-cloud[]
+* {{{value}}}
+endif::[]
+{{else if (eq type "self-managed")}}
+ifndef::env-cloud[]
+* {{{value}}}
+endif::[]
+{{else}}
+* {{{value}}}
+{{/if}}
+{{/with}}
+{{/each}}
+{{/if}}
+{{/with}}
 {{/if}}
 {{#if aliases}}
 

package/tools/property-extractor/topic_property_extractor.py

@@ -304,11 +304,29 @@ class TopicPropertyExtractor:
             
     def _determine_property_type(self, property_name: str) -> str:
         """Determine the type of a property based on its name and usage patterns"""
-        
-        # Type mapping based on property name patterns
+        # Explicit exceptions / overrides for properties whose name contains
+        # keywords that would otherwise map to boolean but which are actually
+        # string-valued (for example, a bucket name).
+        if property_name == "redpanda.remote.readreplica":
+            # This topic property contains the read-replica bucket identifier
+            # and should be treated as a string (not a boolean).
+            return "string"
+        # Explicit override: iceberg.delete is a boolean (whether to delete
+        # the corresponding Iceberg table when the topic is deleted).
+        if property_name == "redpanda.iceberg.delete":
+            return "boolean"
+
+        # Type mapping based on property name patterns (heuristic)
         if any(keyword in property_name for keyword in ["caching", "recovery", "read", "write", "delete"]):
-            if property_name in ["write.caching", "redpanda.remote.recovery", "redpanda.remote.write", 
-                               "redpanda.remote.read", "redpanda.remote.delete", "redpanda.remote.readreplica"]:
+            # Known boolean topic properties (keep list conservative)
+            boolean_props = [
+                "write.caching",
+                "redpanda.remote.recovery",
+                "redpanda.remote.write",
+                "redpanda.remote.read",
+                "redpanda.remote.delete",
+            ]
+            if property_name in boolean_props:
                 return "boolean"
                 
         elif any(suffix in property_name for suffix in [".bytes", ".ms", ".factor", ".lag.ms"]):
@@ -583,7 +601,8 @@ NOTE: All topic properties take effect immediately after being set.
 *Type:* {prop_type}
 
 """
-            if acceptable_values:
+            # If the property type is boolean, never include an Accepted values section
+            if acceptable_values and str(prop_type).lower() not in ("boolean", "bool"):
                 adoc_content += f"*Accepted values:* {acceptable_values}\n\n"
             
             adoc_content += "*Default:* null\n\n"

package/tools/property-extractor/transformers.py

@@ -145,15 +145,19 @@ class IsArrayTransformer:
 
 class NeedsRestartTransformer:
     def accepts(self, info, file_pair):
-        return True
-    
+        # Only accept when the params blob exists and contains a needs_restart entry
+        return (
+            len(info.get("params", [])) > 2
+            and isinstance(info["params"][2].get("value"), dict)
+            and "needs_restart" in info["params"][2]["value"]
+        )
+
     def parse(self, property, info, file_pair):
-        needs_restart = "yes"
-        if len(info["params"]) > 2 and "needs_restart" in info["params"][2]["value"]:
-            needs_restart = re.sub(
-                r"^.*::", "", info["params"][2]["value"]["needs_restart"]
-            )
-        property["needs_restart"] = needs_restart != "no"  # True by default, unless we find "no"
+        # We only get here if accepts(...) returned True, so the metadata blob is present
+        raw = info["params"][2]["value"]["needs_restart"]
+        flag = re.sub(r"^.*::", "", raw)
+        # Store as boolean; do not set any default when metadata is absent
+        property["needs_restart"] = (flag != "no")
 
 class GetsRestoredTransformer:
     def accepts(self, info, file_pair):