A Flask extension for creating standard resource searches
Project description
Flask-Filter
Filtering Extension for Flask / SQLAlchemy
Flask-Filter is a simple Flask extension for standardizing behavior of REST API resource search endpoints. It is designed to integrate with the Flask-SQLAlchemy extension and Marshmallow, a popular serialization library. Check out our GitHub Pages site for the full documentation.
Out-of-the-box, Flask-Filter provides search functionality on top-level object fields via an array of filter objects provided in the JSON body of a POST request. For configuring filtering on derived or nested fields see the "Filtering on Nested Fields" section of the documentation.
Installation
Flask-Filter is available on PyPi. To use this library, we recommend you install it via pip:
(venv)$ pip install flask-filter
Default Filters
Flask-Filter supports searching resources based on an array of filters, JSON objects with the following structure:
{"field": "<field_name>", "op": "<operator>", "value": "<some_value>"}
The built-in filters support the following operators:
symbol | operator | python filter class |
---|---|---|
< | less-than | LTFilter |
<= | less-than or equal to | LTEFilter |
= | equal to | EqualsFilter |
> | greater-than | GTFilter |
>= | greater-than or equal to | GTEFilter |
in | in | InFilter |
!= | not equal to | NotEqualsFilter |
like | like | LikeFilter |
contains | many-to-many associated | ContainsFilter |
Note: Be careful with typing around comparator operators. This version does not provide rigorous type-checking, which could cause problems for a user who submits a search like "find Pets with name greater than 'Fido'"
Many-to-many associations can be searched using the contains
operator.
For a Dog object with a many-to-many relationship with "favorite toys"
defined as Dog.toys = [Toy(), Toy()], you can set the field to "toys.name",
the operator to "contains" and the value to "Tennis Ball". This will perform
a SQL "any" search on that field / value and return any Dog objects who like
tennis balls.
Examples
This section demonstrates simplified use-cases for Flask-Filter. For
a complete example app (a Pet Store API), see the /example
folder.
Note: examples in this readme define simple /search
endpoints that
assume a working Flask app has already been initialized, and other
required classes have been defined in a pet_store
directory. To see
a full implementation, go to /examples/pet_store
Example 1: Manually implementing filters in a flask view
Using the FilterSchema
class directly, you can deserialize an
array of JSON filters into a list of flask_filter.Filter
objects
and directly apply the filters using Filter.apply
to craft a
SQLAlchemy query with a complex set of filters.
filter_schema = FilterSchema()
pet_schema = PetSchema()
@app.route('/api/v1/pets/search', methods=['POST'])
def pet_search():
filters = filter_schema.load(request.json.get("filters"), many=True)
query = Pet.query
for f in filters:
query = f.apply(query, Pet, PetSchema)
return jsonify(pet_schema.dump(query.all())), 200
Example 2: Automatically filtering using the query_with_filters
function
from flask_filter import query_with_filters
pet_schema = PetSchema()
@app.route('/api/v1/pets/search', methods=['POST']
def pet_search():
pets = query_with_filters(Pet, request.json.get("filters"), PetSchema)
return jsonify(pet_schema.dump(pets)), 200
Example 3: Initializing and using the Flask extension object
from flask import Flask
from pet_store import Pet, PetSchema # Model defined as subclass of `db.Model`
from pet_store.extensions import db, filtr # SQLAlchemy and FlaskFilter objects
app = Flask(__name__)
db.init_app(app)
filtr.init_app(app)
@app.route('/api/v1/pets/search', methods=['POST']
def pet_search():
pets = filtr.search(Pet, request.json.get("filters"), PetSchema)
return jsonify(pet_schema.dump(pets)), 200
or alternatively, if you pre-register the Model and Schema with the
FlaskFilter
object you do not need to pass the Schema
directly to
the search
method:
filtr.register_model(Dog, DogSchema) # Register in the app factory
followed by the search execution (without an explicitly-defined schema):
pets = filtr.search(Pet, request.json.get("filters"))
Project details
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
Hashes for Flask_Filter-0.1.0.dev3-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | df040e1a6bb6c840ac8230c0d3c6e937208445832b8ce00bfa41d2650b005e99 |
|
MD5 | 75911eee6cf7f880b445aad13235af0a |
|
BLAKE2b-256 | ff18ce306fa3e6d2942464ec1acbda5c99ef4b87267e186ec68edfc92c1490bd |