Merge branch 'main' into add-topics-langchain-core

Author: hamsehmahamud

.github/workflows/explore-triage-commenter-writer.yml

@@ -1,16 +1,11 @@
 name: Explore PR Triage Commenter
 
-# Posts a sticky comment on PRs that touch topic or collection pages,
-# surfacing the facts maintainers normally look up by hand:
-#   - topics: repo count for the topic
-#   - collections: per-item stars, last push, owner type, plus a flag if
-#     the PR author looks like one of the item owners (self-submission)
-#
-# Edit-in-place: subsequent runs (synchronize, reopen) update the same
-# comment instead of posting a new one. Marker: <!-- explore-triage-comment -->
+# Computes maintainer triage data for Explore PRs in an unprivileged
+# pull_request workflow. A separate workflow_run workflow writes the sticky
+# comment after re-fetching PR state; no privileged job checks out PR code.
 
 on:
-  pull_request_target:
+  pull_request:
     types: [opened, synchronize, reopened]
     paths:
       - 'topics/**'
@@ -22,41 +17,53 @@ concurrency:
 
 permissions:
   contents: read
-  pull-requests: write
+  pull-requests: read
 
 jobs:
-  triage:
+  build-comment-data:
     runs-on: ubuntu-latest
     steps:
-      - uses: actions/github-script@v9
+      - name: Build triage comment data
+        uses: actions/github-script@v9
         env:
-          MARKER: '<!-- explore-triage-comment -->'
+          OUTPUT_PATH: ${{ runner.temp }}/explore-triage-comment.json
         with:
           script: |
-            const marker = process.env.MARKER;
+            const fs = require('fs');
+
             const pr = context.payload.pull_request;
+            const baseOwner = context.repo.owner;
+            const baseRepo = context.repo.repo;
             const prNumber = pr.number;
             const prAuthor = pr.user.login.toLowerCase();
             const headSha = pr.head.sha;
-            const baseOwner = context.repo.owner;
-            const baseRepo = context.repo.repo;
 
-            // List files in the PR (paginated).
+            const payload = {
+              schema: 'explore-triage-comment/v1',
+              owner: baseOwner,
+              repo: baseRepo,
+              prNumber,
+              headSha,
+              baseRepoFullName: pr.base.repo.full_name,
+              headRepoFullName: pr.head.repo && pr.head.repo.full_name,
+              hasChanges: false,
+              topics: [],
+              collections: [],
+            };
+
             const files = await github.paginate(github.rest.pulls.listFiles, {
               owner: baseOwner,
               repo: baseRepo,
               pull_number: prNumber,
               per_page: 100,
             });
 
-            // Detect topic and collection slugs touched.
-            // Skip removed files; only validate slug shape we'd ever expect on disk.
             const SLUG = /^[a-z0-9](?:[a-z0-9-]{0,80}[a-z0-9])?$/i;
             const topics = new Set();
             const collections = new Set();
             for (const f of files) {
               if (f.status === 'removed') continue;
-              const m = f.filename.match(/^(topics|collections)\/([^\/]+)\//);
+              const m = f.filename.match(/^(topics|collections)\/([^/]+)\//);
               if (!m) continue;
               const slug = m[2];
               if (!SLUG.test(slug)) continue;
@@ -66,142 +73,121 @@ jobs:
 
             if (topics.size === 0 && collections.size === 0) {
               core.info('No topic or collection changes detected; nothing to do.');
+              writePayload(payload);
               return;
             }
 
-            const sections = [];
+            payload.hasChanges = true;
+
+            for (const slug of [...topics].sort()) {
+              const topic = { slug, count: null };
+              try {
+                const res = await github.rest.search.repos({
+                  q: `topic:${slug}`,
+                  per_page: 1,
+                });
+                topic.count = res.data.total_count;
+              } catch (err) {
+                core.warning(`Search failed for topic '${slug}': ${err.message}`);
+              }
+              payload.topics.push(topic);
+            }
+
+            for (const slug of [...collections].sort()) {
+              const collection = {
+                slug,
+                readStatus: 'ok',
+                errorStatus: null,
+                items: [],
+              };
+
+              let content;
+              try {
+                content = await readCollectionIndex(slug);
+              } catch (err) {
+                collection.readStatus = err.status === 404 ? 'not-found' : 'error';
+                collection.errorStatus = String(err.status || 'error');
+                payload.collections.push(collection);
+                continue;
+              }
+
+              const items = parseCollectionItems(content);
+              for (const item of items) {
+                if (!/^[\w.-]+\/[\w.-]+$/.test(item)) {
+                  collection.items.push({ name: item, valid: false });
+                  continue;
+                }
 
-            // ---- Topic section ----
-            if (topics.size > 0) {
-              const lines = ['### Topics', ''];
-              for (const slug of topics) {
-                let count = null;
+                const [owner, repo] = item.split('/');
                 try {
-                  const res = await github.rest.search.repos({
-                    q: `topic:${slug}`,
-                    per_page: 1,
+                  const r = await github.rest.repos.get({ owner, repo });
+                  const notes = [];
+                  if (owner.toLowerCase() === prAuthor) notes.push('possible-self-submission');
+                  if (r.data.archived) notes.push('archived');
+                  if (r.data.disabled) notes.push('disabled');
+                  collection.items.push({
+                    name: item,
+                    valid: true,
+                    lookupStatus: 'ok',
+                    stars: r.data.stargazers_count,
+                    pushed: r.data.pushed_at ? r.data.pushed_at.slice(0, 10) : null,
+                    ownerType: r.data.owner.type,
+                    notes,
                   });
-                  count = res.data.total_count;
                 } catch (err) {
-                  core.warning(`Search failed for topic '${slug}': ${err.message}`);
-                }
-                const url = `https://github.com/topics/${encodeURIComponent(slug)}`;
-                if (count == null) {
-                  lines.push(`- **${slug}** — [topic page](${url}) _(repo count lookup failed)_`);
-                } else {
-                  lines.push(`- **${slug}** — ${count.toLocaleString()} repositories — [topic page](${url})`);
+                  collection.items.push({
+                    name: item,
+                    valid: true,
+                    lookupStatus: err.status === 404 ? 'not-found' : 'error',
+                    errorStatus: String(err.status || 'error'),
+                  });
                 }
               }
-              sections.push(lines.join('\n'));
+
+              payload.collections.push(collection);
             }
 
-            // ---- Collection section ----
-            if (collections.size > 0) {
-              for (const slug of collections) {
-                const lines = [`### Collection \`${slug}\``, ''];
+            writePayload(payload);
+
+            async function readCollectionIndex(slug) {
+              const attempts = [];
+              if (pr.head.repo) {
+                attempts.push({
+                  owner: pr.head.repo.owner.login,
+                  repo: pr.head.repo.name,
+                  ref: headSha,
+                });
+              }
+              attempts.push({ owner: baseOwner, repo: baseRepo, ref: headSha });
 
-                // Read collection's index.md at the PR head SHA.
-                // PR commits from forks are mirrored into the base repo's network,
-                // so we can fetch from the base repo with the head SHA — simpler
-                // and avoids any cross-repo token concerns.
-                let content;
+              let lastError;
+              for (const attempt of attempts) {
                 try {
                   const res = await github.rest.repos.getContent({
-                    owner: baseOwner,
-                    repo: baseRepo,
+                    ...attempt,
                     path: `collections/${slug}/index.md`,
-                    ref: headSha,
                   });
-                  content = Buffer.from(res.data.content, 'base64').toString('utf8');
-                } catch (err) {
-                  lines.push(`_Could not read \`collections/${slug}/index.md\` at PR head (\`${err.status || 'error'}\`)._`);
-                  sections.push(lines.join('\n'));
-                  continue;
-                }
-
-                const items = parseCollectionItems(content);
-                if (items.length === 0) {
-                  lines.push('_No `items:` list found in frontmatter._');
-                  sections.push(lines.join('\n'));
-                  continue;
-                }
-
-                lines.push('| Item | Stars | Last push | Owner type | Notes |');
-                lines.push('| --- | ---: | --- | --- | --- |');
-
-                for (const item of items) {
-                  if (!/^[\w.-]+\/[\w.-]+$/.test(item)) {
-                    const safeItem = item.replace(/`/g, "'").replace(/\\/g, '\\\\').replace(/\|/g, '\\|');
-                    lines.push(`| \`${safeItem}\` | – | – | – | invalid format |`);
-                    continue;
-                  }
-                  const [owner, repo] = item.split('/');
-                  try {
-                    const r = await github.rest.repos.get({ owner, repo });
-                    const stars = r.data.stargazers_count.toLocaleString();
-                    const pushed = r.data.pushed_at ? r.data.pushed_at.slice(0, 10) : '–';
-                    const ownerType = r.data.owner.type;
-                    const notes = [];
-                    if (owner.toLowerCase() === prAuthor) notes.push('⚠️ possible self-submission');
-                    if (r.data.archived) notes.push('archived');
-                    if (r.data.disabled) notes.push('disabled');
-                    lines.push(`| [\`${item}\`](https://github.com/${item}) | ${stars} | ${pushed} | ${ownerType} | ${notes.join(', ') || '–'} |`);
-                  } catch (err) {
-                    const note = err.status === 404 ? 'not found' : `error (${err.status || '?'})`;
-                    lines.push(`| \`${item}\` | – | – | – | ${note} |`);
+                  if (Array.isArray(res.data) || res.data.type !== 'file' || !res.data.content) {
+                    const err = new Error('Collection index is not a file');
+                    err.status = 'invalid';
+                    throw err;
                   }
+                  return Buffer.from(res.data.content, 'base64').toString('utf8');
+                } catch (err) {
+                  lastError = err;
                 }
-                lines.push('');
-                sections.push(lines.join('\n'));
               }
-            }
-
-            const body = [
-              marker,
-              '<!-- Maintained by .github/workflows/explore-triage-commenter.yml. Edits will be overwritten. -->',
-              '',
-              '## Maintainer triage',
-              '',
-              ...sections,
-            ].join('\n');
-
-            // Edit-in-place via marker.
-            const comments = await github.paginate(github.rest.issues.listComments, {
-              owner: baseOwner,
-              repo: baseRepo,
-              issue_number: prNumber,
-              per_page: 100,
-            });
-            const existing = comments.find(c => c.body && c.body.startsWith(marker));
-
-            if (existing) {
-              await github.rest.issues.updateComment({
-                owner: baseOwner,
-                repo: baseRepo,
-                comment_id: existing.id,
-                body,
-              });
-              core.info(`Updated comment ${existing.id}`);
-            } else {
-              await github.rest.issues.createComment({
-                owner: baseOwner,
-                repo: baseRepo,
-                issue_number: prNumber,
-                body,
-              });
-              core.info('Created new comment');
+              throw lastError;
             }
 
             function parseCollectionItems(text) {
-              // Frontmatter between leading --- lines.
               const fmMatch = text.match(/^---\n([\s\S]*?)\n---/);
               if (!fmMatch) return [];
               const lines = fmMatch[1].split('\n');
               const items = [];
               let inItems = false;
               for (const line of lines) {
                 if (/^items:\s*$/.test(line)) { inItems = true; continue; }
-                // Next top-level key ends the items block.
                 if (inItems && /^[a-zA-Z_]\w*\s*:/.test(line)) break;
                 if (inItems) {
                   const m = line.match(/^\s*-\s*([^\s#]+)/);
@@ -210,3 +196,16 @@ jobs:
               }
               return items;
             }
+
+            function writePayload(data) {
+              fs.writeFileSync(process.env.OUTPUT_PATH, JSON.stringify(data, null, 2));
+              core.info(`Wrote ${process.env.OUTPUT_PATH}`);
+            }
+
+      - name: Upload triage comment data
+        uses: actions/upload-artifact@v4
+        with:
+          name: explore-triage-comment
+          path: ${{ runner.temp }}/explore-triage-comment.json
+          if-no-files-found: error
+          retention-days: 1

.github/workflows/explore-triage-commenter.yml

@@ -7,7 +7,7 @@ gem "json", "2.21.2"
 gem "language_server-protocol", "3.17.0.6"
 gem "nokogiri", "~> 1.19.4"
 gem "rake", "13.4.2"
-gem "rubocop", "1.88.2"
+gem "rubocop", "1.89.0"
 
 group :test do
   gem "fastimage"

Gemfile

@@ -286,7 +286,7 @@ GEM
       faraday (>= 1, < 3)
       sawyer (~> 0.9)
     parallel (2.1.0)
-    parser (3.3.11.1)
+    parser (3.3.12.0)
       ast (~> 2.4.1)
       racc
     pathutil (0.16.2)
@@ -308,7 +308,7 @@ GEM
       io-console (~> 0.5)
     rexml (3.4.2)
     rouge (3.30.0)
-    rubocop (1.88.2)
+    rubocop (1.89.0)
       json (~> 2.3)
       language_server-protocol (~> 3.17.0.2)
       lint_roller (~> 1.1.0)
@@ -379,7 +379,7 @@ DEPENDENCIES
   octokit
   pry
   rake (= 13.4.2)
-  rubocop (= 1.88.2)
+  rubocop (= 1.89.0)
   rubocop-performance
   safe_yaml
   webrick

Gemfile.lock

@@ -6,6 +6,7 @@ items:
  - microsoft/semantic-kernel
  - modelcontextprotocol/servers
  - gfernandf/agent-skills
+ - tamish-max/embercore
 display_name: AI Agents
 created_by: gfernandf
 ---

collections/ai-agents/index.md

@@ -0,0 +1,13 @@
+---
+items:
+ - ajay-dhangar/algo
+ - codeharborhub/dsa
+ - TheAlgorithms/Python
+ - TheAlgorithms/Java
+ - trekhleb/javascript-algorithms
+ - jwasham/coding-interview-university
+display_name: Algorithms & Data Structures
+created_by: ajay-dhangar
+---
+
+A curated collection of open-source algorithms, data structures, and implementation guides to help developers master computational problem-solving and technical interviews.