Tina4-Python ORM#

🔥 Hot Tips

You never need the ORM. Raw Database is perfect for 90% of cases
But when you want it... it feels like writing plain Python classes
Zero boilerplate: no __init__, no save() method to write
Works with SQLite, PostgreSQL, MySQL, MSSQL, Firebird, all the same

:::

Quick Start - 8 lines {#quick-start}#

python

from tina4_python import ORM, orm, Databasefrom tina4_python import IntegerField, StringField, DateTimeFieldclass User(ORM):    id         = IntegerField(primary_key=True, auto_increment=True)    name       = StringField()    email      = StringField(unique=True)    created_at = DateTimeField()# Connect once (usually in your main.py)orm(Database("sqlite3:test.db"))# That's it - table is auto-created on first useuser = User({"name": "Alice", "email": "alice@example.com"})user.save()                    # → INSERTprint(user.id)                 # → 1user.name = "Alice Wonder"user.save()                    # → UPDATEuser = User()user.load("email = ?", ["alice@example.com"])print(user.name)               # → Alice Wonder

Field Types {#field-types}#

Field Type	Python type	SQL example	Notes
`IntegerField()`	`int`	`INTEGER`
`StringField()`	`str`	`VARCHAR(255)`	`length=500` optional
`TextField()`	`str`	`TEXT`	unlimited
`DateTimeField()`	`datetime`	`TIMESTAMP`
`NumericField()`	`Decimal`	`NUMERIC(10,2)`
`BlobField()`	`bytes`	`BLOB`
`JSONBField()`	`dict/list`	`JSON` / `JSONB`	auto serialize/deserialize
`ForeignKeyField(...)`	`int`	`INTEGER REFERENCES table(id)`	import from `tina4_python.FieldTypes`

Field Options#

python

name = StringField(    default="Anonymous",    null=False,    length=100)created_at = DateTimeField()status = IntegerField(default=1)

Foreign Keys - Beautiful & Simple {#foreign-keys}#

python

from tina4_python.FieldTypes import ForeignKeyFieldclass Post(ORM):    id        = IntegerField(primary_key=True, auto_increment=True)    title     = StringField()    author_id = ForeignKeyField(references=User)

Usage:

python

post = Post({"title": "Hello World", "author_id": 1})post.save()# Loadpost = Post()post.load("id = ?", [1])print(post.title)

Core Methods {#core-methods}#

Method	Example	What it does	Returns
`.save()`	`user.save()`	INSERT or UPDATE
`.load(where, params)`	`user.load("id = ?", [1])`	Load single record into instance	`bool`
`.select()` / `.fetch()`	`User().select()`	Returns records (default limit 10)	`DatabaseResult`
`.fetch_one()`	`User().fetch_one()`	Fetches one record	`DatabaseResult`
`.delete()`	`user.delete()`	Delete record
`.create_table()`	`User().create_table()`	Auto-create table
`.to_dict()`	`user.to_dict()`	Convert instance to dict	`dict`

Migrations - One command {#migrations}#

python

from tina4_python.Migration import migratemigrate(Database("sqlite3:test.db"))  # creates/updates all ORM tables

Full Example - Real Project Ready {#full-example}#

python

from tina4_python import ORM, orm, Databasefrom tina4_python import IntegerField, StringField, TextField, DateTimeFieldfrom tina4_python.FieldTypes import ForeignKeyFieldorm(Database("sqlite3:app.db"))  # ← one line to rule them allclass Category(ORM):    id   = IntegerField(primary_key=True, auto_increment=True)    name = StringField(unique=True)class Article(ORM):    id          = IntegerField(primary_key=True, auto_increment=True)    title       = StringField()    content     = TextField()    category_id = ForeignKeyField(references=Category)    created_at  = DateTimeField()# Auto-create tablesCategory().create_table()Article().create_table()# Use itcat = Category({"name": "Tech"})cat.save()article = Article({"title": "Tina4 is awesome", "content": "Great framework", "category_id": 1})article.save()# Queryresult = Article().select(filter="category_id = ?", params=[1])for a in result.records:    print(a["title"])

Summary - The Tina4 ORM Philosophy {#summary}#

No __init__ needed
No save() method to write
No migration files
No sessions
No query builder hell

Just:

python

user = User({"name": "Bob"})user.save()@post("/api/users")async def post_api_users(request, response):    user = User(request.body)    user.save()    return response(user.to_dict())