TIG-254: Move selected tagger into backend code

.env.example

@@ -5,7 +5,7 @@
 
 # ----- PostgreSQL (Docker) -----
 POSTGRES_USER=forum
-POSTGRES_PASSWORD=forum_password
+POSTGRES_PASSWORD=YOUR_DB_PASSWORD_HERE
 POSTGRES_DB=the_forum
 POSTGRES_PORT=5434
 
@@ -16,17 +16,17 @@ DATABASE_URL=postgresql://forum:forum_password@localhost:5434/the_forum
 AUTH_SECRET=your-auth-secret-here
 AUTH_AZURE_AD_CLIENT_ID=your-client-id
 AUTH_AZURE_AD_CLIENT_SECRET=your-client-secret
-AUTH_AZURE_AD_TENANT_ID=2ff60116-7431-425d-b5af-077d7791bda4
+AUTH_AZURE_AD_TENANT_ID=YOUR_TENANT_ID_HERE
 
 # ----- AWS S3 (image uploads) -----
 # AWS_S3_BUCKET=the-forum-uploads
 # AWS_REGION=us-east-1
 
 # ----- Next.js public vars (prefix with NEXT_PUBLIC_) -----
 # NEXT_PUBLIC_API_URL=http://localhost:8000
-NEXT_PUBLIC_MAPBOX_TOKEN=pk.your-mapbox-public-token
-NEXT_PUBLIC_CAMPUS_MAP_TOKEN=pk.campus-map-public-token
-NEXT_PUBLIC_CAMPUS_MAP_STYLE=mapbox://styles/account/style-id
+NEXT_PUBLIC_MAPBOX_TOKEN=YOUR_CAMPUS_MAPBOX_TOKEN_HERE
+NEXT_PUBLIC_CAMPUS_MAP_TOKEN=YOUR_CAMPUS_MAP_TOKEN_HERE
+NEXT_PUBLIC_CAMPUS_MAP_STYLE=YOUR_CAMPUS_MAP_STYLE_HERE
 
 # ----- FastAPI -----
 # FASTAPI_SECRET_KEY=changeme
@@ -44,4 +44,4 @@ NEXT_PUBLIC_CAMPUS_MAP_STYLE=mapbox://styles/account/style-id
 
 # ----- Listserv Scraper -----
 LISTSERV_EMAIL=tigerapp@princeton.edu
-LISTSERV_PASSWORD=your-listserv-password
+LISTSERV_PASSWORD=YOUR_LISTSERV_PASSWORD_HERE

.github/workflows/scrape_listserv.yml

@@ -0,0 +1,41 @@
+name: Scrape WhitmanWire
+
+on:
+  schedule:
+    # Runs at 8:00 AM UTC (4:00 AM Eastern Time) every day
+    - cron: '0 8 * * *'
+  workflow_dispatch: # Allows you to click a "Run workflow" button in GitHub's UI
+
+jobs:
+  scrape-and-commit:
+    runs-on: ubuntu-latest
+    # This permission is required so the bot can push the JSON file back to your repo
+    permissions:
+      contents: write 
+
+    steps:
+      - name: Checkout repository
+        uses: actions/checkout@v4
+
+      - name: Set up Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: '3.11' # The script uses standard libraries, so this is fine
+
+      - name: Run Scraper
+        env:
+          # Pulls securely from GitHub Secrets
+          LISTSERV_EMAIL: ${{ secrets.LISTSERV_EMAIL }}
+          LISTSERV_PASSWORD: ${{ secrets.LISTSERV_PASSWORD }}
+        run: |
+          python3 apps/listserv-scraper/src/scrape_listserv.py --list WHITMANWIRE --limit 100 --fetch-bodies
+
+      - name: Commit and push changes
+        run: |
+          git config --global user.name "github-actions[bot]"
+          git config --global user.email "41898282+github-actions[bot]@users.noreply.github.com"
+          # Stage the specific data folder where the script saves its output
+          git add apps/listserv-scraper/data/
+          # Commit the changes; if there are no new emails, it fails gracefully instead of crashing
+          git commit -m "chore: update WhitmanWire scraped data" || echo "No changes to commit"
+          git push

.gitignore

@@ -19,6 +19,7 @@ build/
 !.env.example
 !**/.env.example
 !**/env.local.example
+.forum_venv/
 
 # ---- Database ----
 drizzle/

apps/database/package.json

@@ -8,9 +8,9 @@
   "scripts": {
     "db:generate": "drizzle-kit generate",
     "db:migrate": "drizzle-kit migrate",
-    "db:push": "drizzle-kit push",
+    "db:push": "drizzle-kit push --force",
     "db:studio": "drizzle-kit studio",
-    "db:seed": "tsx src/seed.ts",
+    "db:seed": "bun src/seed.ts",
     "lint": "biome check .",
     "check-types": "tsc --noEmit"
   },

apps/database/src/schema/index.ts

@@ -10,29 +10,35 @@ import {
   timestamp,
   uuid,
   varchar,
+  vector,
 } from "drizzle-orm/pg-core";
 
 // ── Enums ─────────────────────────────────────────────────
 
+// YUBI has modified event tags
 export const eventTagEnum = pgEnum("event_tag", [
-  "free-food",
-  "workshop",
-  "performance",
-  "speaker",
-  "social",
+  "free food",
   "career",
-  "sports",
-  "music",
-  "art",
-  "academic",
-  "cultural",
-  "community-service",
-  "religious",
-  "political",
+  "research",
+  "academics",
   "tech",
+  "entrepreneurship",
+  "politics",
+  "visual arts",
+  "performing arts",
+  "literature",
+  "culture",
+  "music",
   "gaming",
-  "outdoor",
+  "athletics",
+  "religion",
+  "sustainability",
+  "outdoors",
   "wellness",
+  "community service",
+  "speaker event",
+  "social event",
+  "stem",
 ]);
 
 export const campusRegionEnum = pgEnum("campus_region", [
@@ -59,14 +65,14 @@ export const notificationTypeEnum = pgEnum("notification_type", [
 export const orgCategoryEnum = pgEnum("org_category", [
   "career",
   "affinity",
-  "performance",
-  "academic",
-  "athletic",
-  "social",
-  "cultural",
-  "religious",
-  "political",
-  "service",
+  "performing arts",
+  "academics",
+  "athletics",
+  "social event",
+  "culture",
+  "religion",
+  "politics",
+  "community service",
 ]);
 
 export const orgRoleEnum = pgEnum("org_role", ["owner", "officer", "member"]);
@@ -219,6 +225,11 @@ export const eventTags = pgTable(
   (t) => [primaryKey({ columns: [t.eventId, t.tag] })],
 );
 
+export const eventTagEmbeddings = pgTable("event_tag_embeddings", {
+  tagName: eventTagEnum("tag_name").primaryKey(),
+  embedding: vector("embedding", { dimensions: 1536 }).notNull(),
+});
+
 export const rsvps = pgTable(
   "rsvps",
   {
@@ -278,6 +289,7 @@ export const pipelineLogStatusEnum = pgEnum("pipeline_log_status", [
   "skipped_not_event",
   "duplicate",
   "error",
+  "needs_review",
 ]);
 
 export const listservConfigs = pgTable("listserv_configs", {