Hi! I'm Usama Walayat, a GSoC 2026 applicant for the Wikifile-Transfer project. I'd like to claim this task. Here's my deep analysis and proposed solution:
Problem:
The WMCZ Tracker (a Django-based expense and document tracking tool for Wikimedia CZ) has a global search feature. Currently when users search for anything, the search query also scans through boolean/checkbox fields like "Wage?" and "Paid?". This makes no sense — these are Yes/No fields and searching them with text produces irrelevant or misleading results, polluting the search output with unintended matches.
Root Cause:
In the Django Tracker codebase, the search view likely uses a broad Q() query that includes ALL model fields via icontains — including boolean fields like is_wage and is_paid. Django does not raise an error when you apply icontains to a BooleanField but it produces meaningless results.
My Proposed Solution:

Locate the search view in the Tracker codebase (likely in views.py or search_indexes.py)
Identify the Q() query object that builds the search filter
Explicitly exclude boolean fields (wage, paid) from the search query by removing them from the Q() chain
Ensure only text-based fields like title, description, author, document_number remain in the search
Add a test case verifying that searching "True" or "False" does not return results based on the Wage/Paid fields

I have Python and Django experience including writing search views with Q() objects. May I claim this task?