It converts django models to a sqlalchemy orm/expression objects.
Project description
Requirements
Python: 2.7.15 or later, 3.4 or later.
Tested with 2.7.15, 3.6
Django: 1.9 or later.
Tested with 1.11, 2.0, 2.1
SQLAlchemy: 0.9 or later.
Tested with 1.2
Installation
$ pip install d2aUsage
Auto loading
Just add d2a to settings.INSTALLED_APPS.
INSTALLED_APPS = [
'django.contrib.admin',
'django.contrib.auth',
'django.contrib.contenttypes',
'django.contrib.sessions',
'django.contrib.messages',
'django.contrib.staticfiles',
# top or here
'd2a',
# example apps
'books',
'sales',
]Then models_sqla (default) in all apps become possible to be imported as a module.
>>> from books import models_sqla
>>> models_sqla. # tab completion
models_sqla.Author( models_sqla.BookCategory( models_sqla.CategoryRelation( models_sqla.transfer(
models_sqla.Book( models_sqla.Category( models_sqla.models
>>> models_sqla.Book
<class 'd2a.book'>
>>> models_sqla.Book. # tab completion
models_sqla.Book.author models_sqla.Book.content models_sqla.Book.metadata models_sqla.Book.tags
models_sqla.Book.author_id models_sqla.Book.description models_sqla.Book.mro( models_sqla.Book.title
models_sqla.Book.category models_sqla.Book.id models_sqla.Book.price
# SQL Expression schema
>>> models_sqla.Book.__table__
Table(
'book', MetaData(bind=None),
Column('id', UUID(), table=<book>, primary_key=True, nullable=False, default=ColumnDefault(<function uuid4 at 0x7f3cebe7e598>)),
Column('price', JSON(astext_type=Text()), table=<book>, nullable=False),
Column('title', VARCHAR(length=255), table=<book>, nullable=False),
Column('description', TEXT(), table=<book>),
Column('author_id', INTEGER(), ForeignKey('author.id'), table=<book>),
Column('content', BYTEA(), table=<book>, nullable=False),
Column('tags', ARRAY(VARCHAR()), table=<book>, nullable=False),
schema=None
)Also, it can extract model declared implicitly depending on m2m field. (in this case, BookCategory)
Per models module
If you want to create a module manually, create a models_sqla.py in the apps.
Write like the following to it:
from d2a import transfer
from . import models
transfer(models, globals())models_sqla.py exists, auto module creation will be omitted.
And if you create every models_sqla.py manually, it is unnecessary to set d2a to settings.INSTALLED_APPS.
Example:
You can omit specifying db_type, then it automatically detects a database type from settings.DATABASES['default'].
Now postgresql, mysql and oracle are allowed, the other types will be converted to the following types as default type: sqlalchemy/types.py
Per model
If you just want to convert one model, you should use declare function.
>>> from d2a import declare
>>> from sales.models import Sales
>>> sales = declare(Sales)
>>> sales
<class 'd2a.sales'>
>>> sales.__table__
Table(
'sales', MetaData(bind=None),
Column('id', BIGINT(), table=<sales>, primary_key=True, nullable=False),
Column('book_id', UUID(), ForeignKey('book.id'), table=<sales>, nullable=False),
Column('sold', TIMESTAMP(), table=<sales>, nullable=False),
Column('reservation', INTERVAL(), table=<sales>),
Column('source', INET(), table=<sales>),
schema=None
)
>>> sales.
sales.book sales.id sales.mro( sales.sold
sales.book_id sales.metadata sales.reservation sales.sourceCustom fields
If you are using customized field (not built-in), you can register the field as the other field using alias or alias_dict method.
from django.db.models import ImageField
class ExtendedImageField(ImageField):
"""something customizing"""
from d2a import alias
alias(ExtendedImageField, ImageField)
# or
alias_dict({
ExtendedImageField: ImageField,
})Querying shortcut
ORM
There is a function named make_session for ORM mode.
>>> from d2a import make_session
>>> from books.models_sqla import Author
>>> with make_session() as session:
... # it commits and flushes automatically when the scope exits.
... a = Author()
... a.name = 'righ'
... a.age = 30
... session.add(a)
...
>>> with make_session() as session:
... # when the session was rolled back or causes some exception in the context,
... # it won't register records in the session.
... a = Author()
... a.name = 'teruhiko'
... a.age = 85
... session.add(a)
... session.rollback()
...
>>> with make_session() as session:
... session.query(Author.name, Author.age).all()
...
[('righ', 30)]It receives the following arguments, all arguments can be omitted.
- engine:
engine object or database-type (string) (default: None). When it is omitted, it guesses database type and gets an engine automatically.
- autoflush:
It is the same as sessionmaker (default: True)
- autocommit:
It is the same as sessionmaker (default: False)
- expire_on_commit:
It is the same as sessionmaker (default: True)
- info:
It is the same as sessionmaker (default: None)
Expression
There are two functions.
- query_expression:
It is for getting SELECT results, and returns a list containing record.
- execute_expression:
It is for executing INSERT, DELETE, UPDATE statements, and returns num of records having been affected.
>>> from sqlalchemy import (
... select,
... insert,
... )
>>> from d2a import query_expression, execute_expression
>>> from books.models_sqla import Author
>>>
>>> AuthorTable = Author.__table__
>>> records = [
... {'name': 'a', 'age': 10},
... {'name': 'b', 'age': 30},
... {'name': 'c', 'age': 20},
... ]
>>> # insert
>>> stmt = insert(AuthorTable).values(records)
>>> execute_expression(stmt)
3
>>> # select
>>> stmt = select([
... AuthorTable.c.id,
... AuthorTable.c.name,
... AuthorTable.c.age,
... ]).select_from(AuthorTable).order_by(AuthorTable.c.age)
>>> query_expression(stmt)
[
OrderedDict([('id', 12), ('name', 'a'), ('age', 10)]),
OrderedDict([('id', 14), ('name', 'c'), ('age', 20)]),
OrderedDict([('id', 13), ('name', 'b'), ('age', 30)])
]
>>> # record as tuple
>>> query_expression(stmt, as_dict=False)
[(12, 'a', 10), (14, 'c', 20), (13, 'b', 30)]Demo
start up environment
$ git clone git@github.com:righ/d2a.git
$ cd d2a
$ docker-compose uppreparation
$ docker exec -it d2a_app_1 /bin/bash
# python -m venv venv # only first time
# source venv/bin/activate
(venv) # cd project_postgresql/ # (or mysql)
(venv) project_postgresql # ./manage.py migrateexecute
(venv) project_postgresql # ./manage.py shell>>> from books import models_sqla
>>> book = models_sqla.Book()
>>> author = models_sqla.Author()
>>> book.author = author
>>> author.books
[<d2a.book object at 0x7f3cec539358>]
# And do something you want do ;)Links
History
- 2.0.0:
Added a shortcut function for executing in ORM mode.
Added two shortcut functions for executing in EXPRESSION mode.
- 1.1.x:
(2019-02-17)
Added a function to load all models automatically.
- 1.0.2:
(2018-07-10)
Improved a little.
- 1.0.1:
(2018-07-06)
Fixed a bug, that it will be provided None even though it’s not specified default argument.
- 1.0.0:
(2018-07-05)
Fixed bugs.
Added unit tests.
- 0.0.6:
Fixed a bug that abstract models become the targets.
Deleted install_requires.
- 0.0.5:
added alias method.
- 0.0.4:
fixed bugs.
- 0.0.3:
it got easy to declare custom field.
transfer method can define secondary table.
- 0.0.2:
it supported m2m field.
it limited django version less than 1.9.
- 0.0.1:
first release (2017-12-27)
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
File details
Details for the file d2a-2.0.0.tar.gz.
File metadata
- Download URL: d2a-2.0.0.tar.gz
- Upload date:
- Size: 14.9 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/1.13.0 pkginfo/1.5.0.1 requests/2.21.0 setuptools/40.8.0 requests-toolbelt/0.9.1 tqdm/4.31.1 CPython/3.7.0
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 | a6a53009df6f2e27a1f60069518ab23543aff5dc67903f618350f4200bc1a75c |
|
| MD5 | 27a396510119c74cbe4bdb0f77070c3a |
|
| BLAKE2b-256 | 8c0d607b8120f7e02b0c72310e227908285b48d9db6c4000379877ddac6e0b86 |
File details
Details for the file d2a-2.0.0-py3-none-any.whl.
File metadata
- Download URL: d2a-2.0.0-py3-none-any.whl
- Upload date:
- Size: 12.3 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/1.13.0 pkginfo/1.5.0.1 requests/2.21.0 setuptools/40.8.0 requests-toolbelt/0.9.1 tqdm/4.31.1 CPython/3.7.0
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 | ac7eb11b238c211ece6da8bad5907b6e933730253917cc4595bdbb239eba6712 |
|
| MD5 | 09097e8ca9699a568a2d712d3d9cefb6 |
|
| BLAKE2b-256 | 28d75faa28e8593030959d47de79b62dbb92ddf789095814ba34c13ef63cca3e |
