debugorm 0.1.0

A minimal, plugin-driven ORM for understanding and debugging SQL

DebugORM

A minimal, plugin-driven ORM for SQLite whose primary goal is helping developers understand and debug SQL queries, not just execute them.

---

Architecture

Model → QuerySet → Query → Pipeline → [Plugins] → SQLCompiler → Executor → Result

Key objects

Object	Responsibility
Field	Declares a column; validates and converts values
Model	Base class; provides save(), delete(), create_table()
Manager	Attached as Model.objects; creates QuerySets
QuerySet	Chainable, lazy query builder
Query	Structured, SQL-free representation of a query (the internal AST)
SQLCompiler	Translates Query → (sql, params)
QueryPipeline	Orchestrates the full lifecycle and calls plugin hooks
Plugin	Base class for all debug extensions

Plugin hook order

before_compile(query, ctx)
    ↓
SQLCompiler.compile(query) → CompiledQuery
    ↓
after_compile(compiled, ctx)
    ↓
before_execute(compiled, ctx)   ← return QueryResult here to skip execution
    ↓
QueryExecutor.execute(compiled) → QueryResult
    ↓
after_execute(result, ctx)

---

Quick start

from debugorm import configure, Model
from debugorm.fields import IntegerField, StringField

configure(":memory:")       # or a file path: "myapp.db"

class User(Model):
    id   = IntegerField(primary_key=True)
    name = StringField(max_length=100)
    age  = IntegerField()

User.create_table()

CRUD

# Create
alice = User(name="Alice", age=25)
alice.save()                        # INSERT

alice.age = 26
alice.save()                        # UPDATE

alice.delete()                      # DELETE

# Read
User.objects.all()
User.objects.filter(age__gte=18).order_by("-age").all()
User.objects.get(name="Bob")
User.objects.filter(age__gt=20).count()
User.objects.order_by("-age").first()

Supported lookup operators

Suffix	SQL
age__eq=18 or age=18	age = 18
age__ne=18	age != 18
age__gt=18	age > 18
age__gte=18	age >= 18
age__lt=18	age < 18
age__lte=18	age <= 18
name__like="A%"	name LIKE 'A%'
name__in=["Alice","Bob"]	name IN ('Alice', 'Bob')
name__isnull=True	name IS NULL

---

Debug plugins

Attach plugins on the fly with .debug():

User.objects.debug(
    explain=True,       # EXPLAIN QUERY PLAN
    dry_run=True,       # show SQL without executing
    log=True,           # log SQL + timing via Python logging
    visualize=True,     # print query tree
    pretty=True,        # pretty-format SQL
).filter(age__gt=18).all()

ExplainPlugin

Runs EXPLAIN QUERY PLAN after query execution and prints how SQLite plans to process it.

╔────────────────────────────────────────────────────────────╗
║  [ExplainPlugin]
╠────────────────────────────────────────────────────────────╣
║  SQL  : SELECT * FROM users WHERE age > ? ORDER BY age DESC
╠────────────────────────────────────────────────────────────╣
║  QUERY PLAN
║    └── SCAN users
║    └── USE TEMP B-TREE FOR ORDER BY
╚────────────────────────────────────────────────────────────╝

DryRunPlugin

Prevents execution; estimates how many rows would be affected.

╔────────────────────────────────────────────────────────────╗
║  [DryRunPlugin]  ⚠  NOT EXECUTED
╠────────────────────────────────────────────────────────────╣
║  SQL   : SELECT * FROM users WHERE age >= ? ORDER BY name ASC
║  Interpolated: SELECT * FROM users WHERE age >= 18 ORDER BY name ASC
╠────────────────────────────────────────────────────────────╣
║  Estimated rows affected: ~3
╚────────────────────────────────────────────────────────────╝

LoggingPlugin

Uses Python's logging module — integrates with any existing log config.

DEBUG  debugorm.sql › SQL ▶ SELECT * FROM users WHERE age > ?  params=(20,)
DEBUG  debugorm.sql › SQL ◀ rows=3  time=0.12ms

VisualizePlugin

Renders the query as a tree before compilation (shows intent, not SQL).

╔────────────────────────────────────────────────────────────╗
║  [VisualizePlugin]  Query Structure
╠────────────────────────────────────────────────────────────╣
║  User (SELECT)
║   ├── filter: age > 18
║   ├── filter: name LIKE 'A%'
║   ├── order_by: name ASC
║   └── limit: 10
╚────────────────────────────────────────────────────────────╝

Pass show_json=True to also get the JSON representation.

PrettyPrintPlugin

Formats the compiled SQL for human readability.

SELECT *
  FROM users
  WHERE age >= ?
  ORDER BY age DESC
  LIMIT ?

---

Query Diff

Compare two queries at the internal representation level (not string comparison):

q1 = User.objects.filter(age__gt=18)
q2 = User.objects.filter(age__gt=21)

print(q1.diff(q2))
# - age > 18
# + age > 21

q3 = User.objects.filter(age__gte=18).order_by("-age").limit(10)
q4 = User.objects.filter(age__gte=18).order_by("name").limit(5)

print(q3.diff(q4))
# - ORDER BY age DESC
# + ORDER BY name ASC
# - LIMIT 10
# + LIMIT 5

print(q1.diff(q1))
# Queries are identical

---

Global plugin configuration via DebugORM facade

from debugorm import DebugORM, LoggingPlugin, PrettyPrintPlugin

orm = DebugORM(":memory:", plugins=[LoggingPlugin(), PrettyPrintPlugin()])
orm.install()   # applies the pipeline to every already-defined model

---

Project structure

debugorm/
├── __init__.py          public API
├── core/
│   ├── model.py         Model base class + ModelMeta
│   ├── query.py         Query AST + FilterCondition + Query.diff()
│   ├── compiler.py      SQLCompiler → CompiledQuery
│   ├── pipeline.py      QueryPipeline + QueryExecutor + QueryResult
│   └── manager.py       Manager + QuerySet
├── fields/
│   ├── base.py
│   ├── integer.py
│   └── string.py
├── plugins/
│   ├── base.py          Plugin base class + hook signatures
│   ├── explain.py       ExplainPlugin
│   ├── dry_run.py       DryRunPlugin
│   ├── logging_plugin.py LoggingPlugin
│   ├── visualize.py     VisualizePlugin
│   └── pretty_print.py  PrettyPrintPlugin
└── db/
    └── connection.py    Connection + configure()
demo.py                  end-to-end usage example

---

Writing a custom plugin

from debugorm.plugins.base import Plugin

class TimingPlugin(Plugin):
    name = "timing"

    def before_execute(self, compiled, ctx):
        import time
        ctx.extra["t0"] = time.perf_counter()
        return None   # don't intercept

    def after_execute(self, result, ctx):
        elapsed = (time.perf_counter() - ctx.extra["t0"]) * 1000
        print(f"[TimingPlugin] {elapsed:.1f}ms — {result.rowcount} rows")


# Use it:
User.objects.debug().filter(age__gt=18)   # ... or
from debugorm import QueryPipeline
User.objects.pipeline = QueryPipeline(plugins=[TimingPlugin()])

---

Requirements

• Python 3.10+
• No external dependencies (standard library only: sqlite3, logging, dataclasses, copy, re, time, json)