django-fox-query 0.1.0

Static ORM query analyzer for Django — detects N+1, missing indexes, unused queries and more

django-fox-query 🦊

Static ORM query analyzer for Django. Scans your codebase and detects query performance issues before they hit production.

Installation

pip install django-fox-query

Add to your INSTALLED_APPS:

INSTALLED_APPS = [
    ...
    "django_fox_query",
]

Usage

python manage.py find_query

That's it. No configuration needed, no runtime overhead — it's pure static analysis.

What it detects

N+1 queries

Detects querysets on models that have ForeignKey, OneToOneField or ManyToManyField relations without select_related() or prefetch_related().

# ⚠️ detected
books = Book.objects.all()

# ✅ ok
books = Book.objects.all().select_related('author')

---

Missing indexes on filter()

Cross-references the fields used in .filter() and .exclude() with the model definition to detect fields without db_index=True or unique=True.

# ⚠️ detected — title has no index
books = Book.objects.filter(title="Django")

---

Unnecessary field loading

Tracks which fields are actually accessed on a queryset and suggests .values() or .values_list() when only a subset is used.

# ⚠️ detected — only author__name is used
books = Book.objects.all()
for book in books:
    print(book.author.name)

# ✅ suggestion
books = Book.objects.values('author__name')

---

Unused querysets

Detects querysets assigned to a variable that is never used in the scope.

# 🗑️ detected — authors is never used
authors = Author.objects.filter(name="Victor Hugo")

---

len() instead of .count()

Detects len(queryset) calls which load all objects into memory instead of issuing a single COUNT query.

# ⚡ detected
count = len(Book.objects.all())

# ✅ suggestion
count = Book.objects.count()

Example output

📄 /app/core/views.py:4
   scope  → my_view
   model  → Book
   method → .objects.all()
   code   → books = Book.objects.all()
   ⚠️  warning → relation 'author' (ForeignKey) → utilise .select_related('author')
   💡 values → seuls 'author__name' sont utilisés → envisage .values('author__name')

📄 /app/core/views.py:5
   scope  → my_view
   model  → Author
   method → .objects.filter()
   code   → authors = Author.objects.filter(name="Victor Hugo")
   🗑️  unused → 'authors' n'est jamais utilisé → query inutile, envisage la suppression
   🔍 filter → filter sur 'name' sans index → ajoute db_index=True sur Author.name

── len() vs count() ──────────────────
📄 /app/core/views.py:12
   code   → count = len(books)
   ⚡ count → len(books) sur un queryset → utilise books.count()

⚠️  3 N+1 potentiel(s) détecté(s)
4 queries trouvées

How it works

find_query parses your Python files using the ast module — no code is executed, no database connection is needed. It builds a map of your ORM calls and your model definitions, then cross-references the two to surface issues.

Migrations and site-packages are automatically ignored.

Requirements

• Python 3.10+
• Django 4.0+