feat: CMD-187 autocreate search page

[wesleyboar]

Overview

Support a Google Search page with less effort.

What Effort?

Every Time:

1. Create search engine.
2. Configure search engine to match others.
3. Set owners of search engine.
4. Copy search code → ID.
5. Set relevant SEARCH settings.

Previously:

6. Restart the server.
7. Create /search page.
8. Create snippet on page.
9. Paste search code in snippet.
10. Duplicate-or-Neglect search styles from TACC

Now, Instead:

6. Restart the server.

Important

Incomplete because there is a suggestion (which I agree with) to use a custom migration.
Originally built off code from release/v4.36.X when it had been main branch.

Caution

Do not merge.

To Do:

• Rebuild off of main branch.
• Use new solution.

Warning

Successful run requires two deploys. Help wanted.

Related

• CMD-187
• replaces feat: CMD-187 autocreate search page #885

Instances:

• CTRN Search
  ^{Private, so no results.}
• LCCF Search
  ^{Private, but has results from public LCCF.}
• WTCS Search
  ^{Private still, so no results.}

Changes

• added search_page app
• added SEARCH_AUTO_SETUP setting
• added GOOGLE_SEARCH_ENGINE_ID setting

Testing

0. Register a Google Programmable Search Engine.
1. Add real GOOGLE_SEARCH_ENGINE_ID setting e.g.
   ^{GOOGLE_SEARCH_ENGINE_ID = 'b099996c09ebd4ece'}
2. Add SEARCH_AUTO_SETUP = True setting.
3. Run/Deploy website that has no search page.
4. Verify "Search (Auto-Generated)" page exists.
5. Search using search bar.
6. ✅ Verify search page shows with results.
7. Remove GOOGLE_SEARCH_ENGINE_ID setting.
8. Run/Deploy site.
9. Revisit search page.
10. ✅ Verify warning appears on page.

UI

with GOOGLE_SEARCH_ENGINE_ID	sans GOOGLE_SEARCH_ENGINE_ID

[qodo-code-review]

Review Summary by Qodo

Auto-create Google Search page with Django CMS integration

✨ Enhancement

Walkthroughs

Description

• Automatically create and manage Google Search page with minimal configuration
• Add new search_page Django app with CMS integration and apphook support
• Introduce SEARCH_PAGE_AUTO_SETUP and GOOGLE_SEARCH_ENGINE_ID settings
• Implement search template with Google Custom Search Engine styling and fallback warning
• Update search bar to use dynamically resolved search page URL via template tag

Diagram

flowchart LR
  A["Settings<br/>SEARCH_PAGE_AUTO_SETUP<br/>GOOGLE_SEARCH_ENGINE_ID"] -->|triggers| B["SearchPageConfig.ready()"]
  B -->|calls| C["create_page()"]
  C -->|creates CMS Page<br/>with apphook| D["SearchPageApphook"]
  D -->|routes to| E["SearchPageView"]
  E -->|renders| F["search_page.html<br/>with Google CSE"]
  G["nav_search.raw.html"] -->|uses| H["search_page_url tag"]
  H -->|resolves to| D

Loading

File Changes

1. apps/search_page/apps.py ✨ Enhancement +10/-0

App config with auto-setup trigger

apps/search_page/apps.py

---

2. apps/search_page/cms_apps.py ✨ Enhancement +11/-0

CMS apphook for search page routing

apps/search_page/cms_apps.py

---

3. apps/search_page/urls.py ✨ Enhancement +8/-0

URL patterns for search page view

apps/search_page/urls.py

---

View more (10)

4. apps/search_page/views.py ✨ Enhancement +4/-0

Simple view rendering search template

apps/search_page/views.py

---

5. apps/search_page/utils.py ✨ Enhancement +64/-0

Utilities for page creation and retrieval

apps/search_page/utils.py

---

6. apps/search_page/templates/search_page.html ✨ Enhancement +25/-0

Search page template with Google CSE integration

apps/search_page/templates/search_page.html

---

7. apps/search_page/static/search_page/css/google-search.css ✨ Enhancement +152/-0

Comprehensive styling for Google Custom Search

apps/search_page/static/search_page/css/google-search.css

---

8. taccsite_cms/templatetags/search_tags.py ✨ Enhancement +10/-0

Template tag for dynamic search page URL resolution

taccsite_cms/templatetags/search_tags.py

---

9. taccsite_cms/_settings/search.py ⚙️ Configuration changes +6/-2

Add search configuration settings and defaults

taccsite_cms/_settings/search.py

---

10. taccsite_cms/settings/settings.py ⚙️ Configuration changes +2/-0

Register search_page app and expose settings

taccsite_cms/settings/settings.py

---

11. taccsite_cms/settings/settings_custom.example.py 📝 Documentation +3/-2

Document Google Search configuration example

taccsite_cms/settings/settings_custom.example.py

---

12. taccsite_cms/templates/nav_search.raw.html ✨ Enhancement +2/-2

Update search form to use dynamic search URL

taccsite_cms/templates/nav_search.raw.html

---

13. apps/search_page/__init__.py Additional files +0/-0

...

apps/search_page/init.py

---

[qodo-code-review]

Code Review by Qodo

🐞 Bugs (2) 📘 Rule violations (1) 📎 Requirement gaps (0)

Context used

✅ Tickets: CMD-187

1. Unused Page import 📘 Rule violation ⚙ Maintainability

Description

taccsite_cms/templatetags/search_tags.py imports Page but never uses it, which triggers a Flake8
F401 unused import violation. This can fail linting for the taccsite_cms/ module.

Code

taccsite_cms/templatetags/search_tags.py[2]

+from cms.models import Page

Evidence

PR Compliance ID 1 requires no new Flake8 violations in taccsite_cms/. The added file imports
Page on line 2 but does not reference it anywhere in the module, which is an F401 violation.

AGENTS.md
taccsite_cms/templatetags/search_tags.py[1-10]

Agent prompt

The issue below was found during a code review. Follow the provided context and guidance below and implement a solution

## Issue description
`taccsite_cms/templatetags/search_tags.py` imports `Page` but does not use it, which violates Flake8 (`F401`).

## Issue Context
Compliance requires no new Flake8 violations when running `flake8 taccsite_cms/ --max-line-length=120`.

## Fix Focus Areas
- taccsite_cms/templatetags/search_tags.py[1-4]

ⓘ Copy this prompt and use it to remediate the issue with your preferred AI generation tools

---

2. DB write in AppConfig.ready 🐞 Bug ☼ Reliability

Description

SearchPageConfig.ready() calls create_page(), which performs ORM reads/writes during Django app
initialization; this can crash server startup or management commands when CMS tables aren’t
migrated/available yet. The codebase already uses post_migrate for DB bootstrapping, so this is an
inconsistent and unsafe initialization mechanism.

Code

apps/search_page/apps.py[R7-10]

+    def ready(self):
+        if settings.SEARCH_PAGE_AUTO_SETUP:
+            from .utils import create_page
+            create_page()

Evidence

SearchPageConfig.ready() immediately calls create_page(), and create_page() uses the CMS
Page model and create_cms_page() (DB write). In contrast, existing DB bootstrapping in this repo
is deferred to post_migrate, indicating the intended safe lifecycle for DB-touching
initialization.

apps/search_page/apps.py[4-10]
apps/search_page/utils.py[41-59]
taccsite_cms/apps.py[13-23]

Agent prompt

The issue below was found during a code review. Follow the provided context and guidance below and implement a solution

## Issue description
`SearchPageConfig.ready()` triggers DB work (`create_page()`) during app initialization. This runs on every process start and can fail before migrations (or when DB is unavailable), preventing startup.

## Issue Context
The repository already bootstraps DB state using `post_migrate` (see `taccsite_cms/apps.py`). The search page creation should follow the same pattern and be resilient to missing tables.

## Fix Focus Areas
- apps/search_page/apps.py[1-10]
- apps/search_page/utils.py[41-64]
- taccsite_cms/apps.py[13-24]

## Implementation notes
- Replace `ready()` DB call with `post_migrate.connect(...)` (like `taccsite_cms.apps.TaccsiteCmsConfig`).
- Guard access to `settings.SEARCH_PAGE_AUTO_SETUP` via `getattr(settings, 'SEARCH_PAGE_AUTO_SETUP', False)`.
- Ensure the handler is idempotent and wraps DB errors (e.g., `OperationalError`, `ProgrammingError`) with logging instead of crashing startup.

ⓘ Copy this prompt and use it to remediate the issue with your preferred AI generation tools

---

3. Uncaught DB errors in tag 🐞 Bug ☼ Reliability

Description

{% search_page_url %} calls get_page_url()/get_page() which performs a CMS Page query
without effective exception handling, so DB/table-not-ready failures can break rendering of
/cms/nav/search/markup/ instead of falling back to PORTAL_SEARCH_PATH. The current `except
Page.DoesNotExist around filter(...).first()` is unreachable and misleading, and it does not
protect against real DB-level errors for this TemplateView-served markup endpoint relied on by other
services.

Code

taccsite_cms/templates/nav_search.raw.html[R3-6]

+{% load static search_tags %}
<!-- FAQ: This template loads independently at a unique url (see `urls.py`)
          so Portal and User Guide can render this markup into their markup. -->

Evidence

The nav search markup template invokes search_page_url, which calls get_page_url() and then
get_page() to query the CMS Page table via Page.objects.filter(...).first(), meaning it can
hit the database during template rendering. Because .filter().first() does not raise
Page.DoesNotExist (only .get() does), the existing except Page.DoesNotExist handler is dead
code and provides no resilience; additionally, there is no handling for DB-level exceptions such as
missing tables or connection issues, so those exceptions would propagate and break
/cms/nav/search/markup/, which is explicitly exposed via TemplateView in taccsite_cms/urls.py.

taccsite_cms/urls.py[32-40]
taccsite_cms/templates/nav_search.raw.html[3-26]
taccsite_cms/templatetags/search_tags.py[1-10]
apps/search_page/utils.py[18-39]
apps/search_page/utils.py[18-23]

Agent prompt

The issue below was found during a code review. Follow the provided context and guidance below and implement a solution

## Issue description
The `search_page_url` template tag triggers a CMS `Page` database query via `get_page_url()`/`get_page()` during rendering of `/cms/nav/search/markup/`. The current error handling is ineffective because it catches `Page.DoesNotExist` around a `filter(...).first()` query that never raises that exception, and DB/table-not-ready errors (e.g., during deploy/migration windows) can propagate and break this markup endpoint instead of gracefully falling back to `settings.PORTAL_SEARCH_PATH`.

## Issue Context
`/cms/nav/search/markup/` is served via `TemplateView` and is used to return markup to other apps/services, so it should remain reliable even when CMS tables are temporarily unavailable. The code path is: nav search markup template -> `search_page_url` -> `get_page_url()` -> `get_page()` -> `Page.objects.filter(...).first()`, which currently lacks appropriate DB-level exception handling and contains an unreachable `except Page.DoesNotExist`.

## Fix Focus Areas
- taccsite_cms/templates/nav_search.raw.html[3-13]
- taccsite_cms/templatetags/search_tags.py[1-10]
- apps/search_page/utils.py[18-39]
- taccsite_cms/urls.py[32-40]

ⓘ Copy this prompt and use it to remediate the issue with your preferred AI generation tools

---

[qodo-code-review]

1. Unused page import 📘 Rule violation ⚙ Maintainability

taccsite_cms/templatetags/search_tags.py imports Page but never uses it, which triggers a Flake8
F401 unused import violation. This can fail linting for the taccsite_cms/ module.

Agent Prompt

## Issue description
`taccsite_cms/templatetags/search_tags.py` imports `Page` but does not use it, which violates Flake8 (`F401`).

## Issue Context
Compliance requires no new Flake8 violations when running `flake8 taccsite_cms/ --max-line-length=120`.

## Fix Focus Areas
- taccsite_cms/templatetags/search_tags.py[1-4]

ⓘ Copy this prompt and use it to remediate the issue with your preferred AI generation tools

[qodo-code-review]

2. Db write in appconfig.ready 🐞 Bug ☼ Reliability

SearchPageConfig.ready() calls create_page(), which performs ORM reads/writes during Django app
initialization; this can crash server startup or management commands when CMS tables aren’t
migrated/available yet. The codebase already uses post_migrate for DB bootstrapping, so this is an
inconsistent and unsafe initialization mechanism.

Agent Prompt

## Issue description
`SearchPageConfig.ready()` triggers DB work (`create_page()`) during app initialization. This runs on every process start and can fail before migrations (or when DB is unavailable), preventing startup.

## Issue Context
The repository already bootstraps DB state using `post_migrate` (see `taccsite_cms/apps.py`). The search page creation should follow the same pattern and be resilient to missing tables.

## Fix Focus Areas
- apps/search_page/apps.py[1-10]
- apps/search_page/utils.py[41-64]
- taccsite_cms/apps.py[13-24]

## Implementation notes
- Replace `ready()` DB call with `post_migrate.connect(...)` (like `taccsite_cms.apps.TaccsiteCmsConfig`).
- Guard access to `settings.SEARCH_PAGE_AUTO_SETUP` via `getattr(settings, 'SEARCH_PAGE_AUTO_SETUP', False)`.
- Ensure the handler is idempotent and wraps DB errors (e.g., `OperationalError`, `ProgrammingError`) with logging instead of crashing startup.

ⓘ Copy this prompt and use it to remediate the issue with your preferred AI generation tools