A flexible toolkit for customize how Django creates the databse objects for the application schema
Django DB adapter is a flexible schema editor backend built to solve the following problems:
- Generate SQL statements for projects working on database-first approach
- All objects created (including created from Django) must have a particular name pattern, like add prefixes and suffixes
- All table columns should be commented
- Inline constraints (primary/foreign keys, unique/check constraints and indexes) are not allowed
- The database user of application is not the owner of the objects and has no privilege to create/alter/drop them (
python manage.py migrate
will not work for this user). All DDL statements generated should include a grant of manipulation privileges (select/insert/update/delete) on created objects for this user/role - The order of SQL statements matters
- Python (3.6, 3.7, 3.8, 3.9)
- Django (1.11, 2.2)
We highly recommend and only officially support the latest patch release of each Python and Django series.
Install using pip
...
pip install django-db-adapter
Add 'db_adapter'
to your INSTALLED_APPS
setting.
INSTALLED_APPS = [
...
'db_adapter',
]
Let's take a look at a quick example of using DB adapter to customize the DDL statements generated by Django.
This example model defines a Person
, which has a first_name
and last_name
:
from django.db import models
class Person(models.Model):
first_name = models.CharField(max_length=30)
last_name = models.CharField(max_length=30, help_text="It's your last name")
class Meta:
db_table = 'person'
Add the following to your settings.py
module:
INSTALLED_APPS = [
... # Make sure to include the default installed apps here.
'db_adapter',
]
DATABASES = {
'default': {
# Make sure to include `db_adapter.db.backends.oracle` as database
# engine for schema customization
'ENGINE': 'db_adapter.db.backends.oracle',
'NAME': 'xe',
'USER': 'a_user',
'PASSWORD': 'a_password',
'HOST': 'dbprod01ned.mycompany.com',
'PORT': '1540',
}
}
DB_ADAPTER = {
'DEFAULT_ROLE_NAME': 'rl_example',
# Apply this pattern for all tables
'DEFAULT_DB_TABLE_PATTERN': '"example"."tb_{table_name}"',
# Ignore some patterns from normalization
'IGNORE_DB_TABLE_PATTERNS': [
'"{}"."{}"', # Tables with already declared namespace
'django_migrations', # Django migrations table
],
'DEFAULT_OBJECT_NAME_PATTERNS': {
'SEQUENCE': 'sq_{table_name}',
'TRIGGER': 'tg_{table_name}_b',
'INDEX': 'ix_{name}',
'PRIMARY_KEY': 'cp_{name}',
'FOREIGN_KEY': 'ce_{name}',
'UNIQUE': 'ct_{name}_uq',
'CHECK': 'ct_{name}{qualifier}',
},
'SQL_FORMAT_OPTIONS': {
'unquote': True,
'identifier_case': 'lower',
'keyword_case': 'lower',
},
# Order of SQL statements
'SQL_STATEMENTS_ORDER': [
'PRIMARY_KEY',
'UNIQUE',
'FOREIGN_KEY',
'CHECK',
'INDEX',
'COMMENT',
'CONTROL', # Grant/revoke table privileges for specified role (if exists)
'AUTOINCREMENT', # Sequence and triggers for auto-incremented fields
]
}
The above Person
model would create a database table like this:
create table example.tb_person (
id number(11),
first_name nvarchar2(30),
last_name nvarchar2(30)
);
/
alter table example.tb_person
add constraint cp_person_id
primary key (id);
/
alter table example.tb_person
add constraint ct_person_id_nn
check (id is not null);
/
alter table example.tb_person
add constraint ct_person_first_name_nn
check (first_name is not null);
/
alter table example.tb_person
add constraint ct_person_last_name_nn
check (last_name is not null);
/
comment on column example.tb_person.last_name
is 'It''s your last name';
/
grant select, insert, update, delete
on example.tb_person
to rl_example;
/
create sequence example.sq_person
minvalue 1
maxvalue 99999999999
start with 1
increment by 1
cache 20;
/
grant select
on example.sq_person
to rl_example;
/
create or replace trigger example.tg_person_b
before insert on example.tb_person
for each row
when (new.id is null)
begin
select example.sq_person.nextval
into :new.id from dual;
end;
/
v1.0.0
- Apr 16, 2018 - First releasev1.0.1
- Apr 16, 2018 - Rename package and fix setup issuesv1.0.2
- Apr 17, 2018 - Fix documentation previewv2.0.0
- Mar 1, 2021 - Recreate the entire schema editor backend with more flexible featuresv2.0.1
- Mar 22, 2021 - Escape single quotes on column comments