feat: add web FAQs schema and router

[BIA3IA]

No description provided.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro

Run ID: efe83103-5985-44c8-87dc-ed5522c9eb89

📥 Commits

Reviewing files that changed from the base of the PR and between b3243da and db961e8.

📒 Files selected for processing (1)

• src/routers/web/faqs.ts

🚧 Files skipped from review as they are similar to previous changes (1)

• src/routers/web/faqs.ts

---

Walkthrough

Adds migrations and Drizzle snapshots, TypeScript table schemas, tRPC CRUD routes, and router wiring to implement a bilingual (IT/EN) FAQ system with audit fields and cascade deletes.

Changes

FAQ Management Feature

Layer / File(s)	Summary
Database schema migrations and snapshots drizzle/0013_oval_dreaming_celestial.sql, drizzle/0014_remarkable_synch.sql, drizzle/meta/0013_snapshot.json, drizzle/meta/0014_snapshot.json, drizzle/meta/_journal.json	Creates web_faq_categories and web_faqs with identity PKs, audit timestamps, FKs to tg_permissions, and cascading category deletes; renames Italian fields to *_it and adds required English *_en columns; snapshots and journal record migration history.
TypeScript Drizzle ORM schema definitions src/db/schema/web/faqs.ts, src/db/schema/web/index.ts	Adds faqCategories and faqs Drizzle table schemas with localized fields, creator/modifier FKs, shared time columns, and cascade category relationship; merges them into the web schema export.
FAQ CRUD API endpoints src/routers/web/faqs.ts	New tRPC router with getAllFaqs (categories with nested FAQs), addFaqs/addFaqsCategory (insert), editFaqs/editFaqsCategory (update by id), and deleteFaqs/deleteFaqsCategory (delete by id); inputs validated with Zod and DB ops return rows or { error: "NOT_FOUND" }/{ error: null }.
Main application router integration src/routers/web/index.ts, src/routers/index.ts	Composes webRouter exposing faqs and registers it on the main appRouter as the web sub-router, making endpoints available under web.faqs.*.

🚥 Pre-merge checks | ✅ 4

✅ Passed checks (4 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title directly summarizes the main change: adding FAQ schema and router with multilingual support and CRUD operations across database migrations, Drizzle schemas, and tRPC procedures.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

---

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.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[coderabbitai]

Actionable comments posted: 2

🧹 Nitpick comments (2)

src/routers/tg/groups.ts (1)

176-176: 💤 Low value

Consider using > 0 for cache invalidation defensively.

The === 1 check assumes telegramId uniqueness. While this is likely correct, using > 0 for invalidation would be more defensive against data anomalies, while still returning === 1 for the API contract.

♻️ Suggested defensive invalidation

 const rows = await DB.delete(GROUPS).where(eq(GROUPS.telegramId, input.telegramId)).returning()
-if (rows.length === 1) invalidateGroupsCache()
+if (rows.length > 0) invalidateGroupsCache()
 return rows.length === 1

Same pattern for setHide:

-if (rows.length === 1) invalidateGroupsCache()
+if (rows.length > 0) invalidateGroupsCache()
 return rows.length === 1

Also applies to: 193-193

🤖 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 `@src/routers/tg/groups.ts` at line 176, The cache invalidation currently
checks for rows.length === 1 which assumes telegramId uniqueness; change the
condition to rows.length > 0 so any deleted/updated rows cause
invalidateGroupsCache() to run defensively; update both occurrences where
invalidateGroupsCache() is called after DB operations (the block with if
(rows.length === 1) invalidateGroupsCache() and the similar check in setHide) to
use > 0 while keeping the API contract that callers still expect a single-row
result where appropriate.

drizzle/0013_oval_dreaming_celestial.sql (1)

15-25: ⚡ Quick win

Index web_faqs.category_id to prevent FK/cascade scan bottlenecks.

category_id is constrained by FK, but it is not indexed. On larger datasets, category reads and category deletions with cascade can trigger slower scans.

💡 Suggested migration addition

 ALTER TABLE "web_faqs" ADD CONSTRAINT "web_faqs_modified_by_id_tg_permissions_user_id_fk" FOREIGN KEY ("modified_by_id") REFERENCES "public"."tg_permissions"("user_id") ON DELETE no action ON UPDATE no action;
+--> statement-breakpoint
+CREATE INDEX "web_faqs_category_id_idx" ON "web_faqs" ("category_id");

🤖 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 `@drizzle/0013_oval_dreaming_celestial.sql` around lines 15 - 25, Add an index
on web_faqs.category_id to avoid FK/cascade scan bottlenecks: create an index
(preferably concurrently in production) on the column referenced by the foreign
key (web_faqs.category_id) so operations involving the constraint
web_faqs_category_id_web_faq_categories_id_fk run efficiently; include the index
creation in the migration that introduces or follows the FK (name it clearly,
e.g., idx_web_faqs_category_id) and ensure the migration handles existence
checks if rerunnable.

🤖 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 `@src/routers/web/faqs.ts`:
- Around line 12-27: The response schema for getAllFaqs currently omits category
and faq IDs; update the output Zod schemas used in getAllFaqs to include id
fields (e.g., add id: z.string() or z.number() as appropriate) on the category
object and on each faq object so edit/delete mutations can target records; apply
the same addition in both schema blocks referenced (the top-level categories
array and the nested faqs array) and keep nullable/optional types consistent
with your DB model.
- Around line 50-187: All mutating endpoints (addFaqs, addFaqsCategory,
editFaqs, editFaqsCategory, deleteFaqs, deleteFaqsCategory) are currently
public; add explicit auth/role checks at the start of each mutation to require
an authenticated user and appropriate role (e.g., admin or faq-manager) by
inspecting the request context (ctx.user or ctx.session) and fail fast with an
authorization error (throw TRPCError with code "UNAUTHORIZED" or return { error:
"UNAUTHORIZED" }). Also ensure createdBy/modifiedBy values come from ctx.user.id
(or validate they match) instead of trusting client input. Implement the checks
consistently in each mutation before performing DB operations.

---

Nitpick comments:
In `@drizzle/0013_oval_dreaming_celestial.sql`:
- Around line 15-25: Add an index on web_faqs.category_id to avoid FK/cascade
scan bottlenecks: create an index (preferably concurrently in production) on the
column referenced by the foreign key (web_faqs.category_id) so operations
involving the constraint web_faqs_category_id_web_faq_categories_id_fk run
efficiently; include the index creation in the migration that introduces or
follows the FK (name it clearly, e.g., idx_web_faqs_category_id) and ensure the
migration handles existence checks if rerunnable.

In `@src/routers/tg/groups.ts`:
- Line 176: The cache invalidation currently checks for rows.length === 1 which
assumes telegramId uniqueness; change the condition to rows.length > 0 so any
deleted/updated rows cause invalidateGroupsCache() to run defensively; update
both occurrences where invalidateGroupsCache() is called after DB operations
(the block with if (rows.length === 1) invalidateGroupsCache() and the similar
check in setHide) to use > 0 while keeping the API contract that callers still
expect a single-row result where appropriate.

🪄 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: 2702614c-ded4-4aad-b0d5-23d8785c3b87

📥 Commits

Reviewing files that changed from the base of the PR and between 07aa3c1 and 73f4f0b.

📒 Files selected for processing (11)

• drizzle/0013_oval_dreaming_celestial.sql
• drizzle/0014_remarkable_synch.sql
• drizzle/meta/0013_snapshot.json
• drizzle/meta/0014_snapshot.json
• drizzle/meta/_journal.json
• src/db/schema/web/faqs.ts
• src/db/schema/web/index.ts
• src/routers/index.ts
• src/routers/tg/groups.ts
• src/routers/web/faqs.ts
• src/routers/web/index.ts