Making address: https://github.com/honmaple/maple-json
SQLAlchemy Object serialized to JSON
Inspired by the Django REST Framework
More than one instance
posts = Post.query.all()
serializer = Seralizer(posts,many=True)
data = serializer.data
A single instance
post = Post.query.first()
serializer = Seralizer(post,many=False)
data = serializer.data
Rule out field
serializer = Seralizer(post,exclude=['title'])
Including fields only
serializer = Seralizer(post,include=['title'])
Relational query depth
serializer = Seralizer(post,depth=3)
-
depth
The default value is 2
Add some custom functions
serializer = Serializer(post,extra=['get_post_count'])
Post
class Post(Model):
......
def get_post_count(self):
return 11
A function that can pass arguments
class PostSerializer(Serializer):
count = Field(source = 'get_post_count',args={'name':'hello'},default=20)
class Meta:
include = []
depth = 2
include = []
exclude = []
extra = ['count']
Use SQLAlchemy just like Django ORM
Why do many people think Django ORM is better than SQLAlchemy
Basic query (implemented)
-
gt
-
lt
-
lte
-
gte
-
contains
-
in
-
exact
-
iexact
-
startswith
-
istartswith
-
iendswith
-
endswith
-
isnull
-
range
-
year
-
month
-
day
Example:
Post.query.filter_by(title__contains = 'sql').all()
Post.query.exclude_by(title__contains = 'sql').all()
Relational query
Post.query.filter_by(tags__name__contains = 'sql').all()
other
Post.query.filter_by(tags__name__contains = 'sql').or(Post.id == 1,Post.id == 2).all()
Post.query.filter_by(tags__name__contains = 'sql').and(Post.id == 1,Post.id == 2).all()
Post.query.filter_by(tags__name__contains = 'sql').exists()
Post.query.load_only('title')
Remove some of the duplicated work of SQLAlchemy
In the case of Flask-SQLAlchemy, you can remove some of the duplication by inheriting mixins in models.py
ModelMixin
Auto-increment ID – ID
POST = POST (····) POST. SAVE () # POST. DELETE () # POST
The batch operation
-
bulk_insert
-
bulk_update
-
bulk_save
ModelTimeMixin
Add two fields
-
created_at
Data creation time
-
updated_at
Data update time
ModelUserMixin
The User table is associated with a many-to-one relationship with User (i.e., one User has multiple posts).
class Post(ModelUserMixin, Model):
user_related_name = 'posts'
titile = ...