pythonflaskflask-sqlalchemyflask-restless

Modifying a Flask-Restless result array


I have something like this:

{
  "id": 1, 
  "username": "plasmy",
  "userdetails": [
    {
      "detail": "A Name", 
      "detail_name": "Full Name", 
      "id": 1, 
      "user_id": 1
    }, 
    {
      "detail": "an@email.com", 
      "detail_name": "Email", 
      "id": 2, 
      "user_id": 1
    }, 
    {
      "detail": "An Address", 
      "detail_name": "Address", 
      "id": 3, 
      "user_id": 1
    }, 
    {
      "detail": "999-999-9999", 
      "detail_name": "Phone Number", 
      "id": 4, 
      "user_id": 1
    }
  ]
}

This comes as a result from using Flask_Restless and SQLAlchemy. There is a table for users and a table for userdetails, which are put in the userdetails part of that JSON. What I want to do is, find a way in which the data can look like this:

{
  "id": 1, 
  "username": "plasmy",
  "userdetails": {
      "Full Name": "A Name",
      "Email": "an@email.com",
      "Address": "An Address",
      "Phone Number": "A Phone Number"
    }
}

See how I removed the ids and I used the field "detail_name" as the key and "detail" as value. I tried using preprocessors but they didn't work or maybe I'm using them wrong. I put the preprocessor in the "child" table.

This is what I tried doing (but didn't work):

def detail_sort(results):
    return {'user_details': results['userdetails']}


manager.create_api(User, methods=['GET', 'POST'])
manager.create_api(UserDetails, methods=['GET', 'POST'],
                   preprocessors={
                       'GET_COLLECTION': [detail_sort]
                   })

I tried GET_COLLECTION, GET_SINGLE and GET_MANY. Any help on this will be greatly appreciated.

UPDATE: Here is the new code I tried based on the answer

from flask import Blueprint
from medinv import manager
from medinv.User.models import User, UserDetails

blueprint = Blueprint('blueprint', __name__)


@blueprint.route('/')
@blueprint.route('/home')
def home():
    return "Welcome."


def detail_sort(results):
    print(results)
    results['userdetails'] = {item['detail_name']: item['detail'] for item in results['userdetails']}
    return results['userdetails']


manager.create_api(User, methods=['GET', 'POST'])
manager.create_api(UserDetails, methods=['GET', 'POST'],
                   postprocessors={
                       'GET_COLLECTION': [detail_sort]
                   })

Solution

  • I think you need to use postproccessors since you need to modify the json response before sending it back to the client.

    OK, I reproduced your problem. Now it's working. Here is my code:

    import flask
    import flask_sqlalchemy
    import flask_restless
    
    # Create the Flask application and the Flask-SQLAlchemy object.
    app = flask.Flask(__name__)
    app.config['DEBUG'] = True
    app.config['SQLALCHEMY_DATABASE_URI'] = 'sqlite:////tmp/test.db'
    db = flask_sqlalchemy.SQLAlchemy(app)
    
    
    # Create your Flask-SQLALchemy models as usual but with the following
    # restriction: they must have an __init__ method that accepts keyword
    # arguments for all columns (the constructor in
    # flask_sqlalchemy.SQLAlchemy.Model supplies such a method, so you
    # don't need to declare a new one).
    
    class User(db.Model):
    
        __tablename__ = 'user'
    
        id = db.Column(db.Integer, primary_key=True)
        username = db.Column(db.String)
        userdetails = db.relationship('UserDetails', backref='User', lazy='dynamic')
    
    
    class UserDetails(db.Model):
    
        __tablename__ = 'user_details'
    
        id = db.Column(db.Integer, primary_key=True)
        detail = db.Column(db.String)
        detail_name = db.Column(db.String)
        user_id = db.Column(db.Integer, db.ForeignKey("user.id"), nullable=False)
    
    # Create the database tables.
    db.create_all()
    
    # Create the Flask-Restless API manager.
    manager = flask_restless.APIManager(app, flask_sqlalchemy_db=db)
    
    user = User(username='plasmy')
    userdetail_0 = UserDetails(detail='A name', detail_name='Full Name' )
    userdetail_1 = UserDetails(detail='an@email.com', detail_name='Email')
    userdetail_2 = UserDetails(detail='An Address', detail_name='Address')
    userdetail_3 = UserDetails(detail='999-999-9999', detail_name='Phone Number')
    
    
    user.userdetails.append(userdetail_0)
    user.userdetails.append(userdetail_1)
    user.userdetails.append(userdetail_2)
    user.userdetails.append(userdetail_3)
    
    db.session.add(user)
    db.session.commit()
    
    print('USER CREATED')
    
    def detail_sort(result, **kw):
        print('detail_sort called')
        print(result)
    
        for entry in result['objects']:
            entry['userdetails'] = {item['detail_name']: item['detail'] for item in
                                    entry['userdetails']}
        print('MODIFIED JSON: ', result)
    
    # Create API endpoints, which will be available at /api/<tablename> by
    # default. Allowed HTTP methods can be specified as well.
    # manager.create_api(Person, methods=['GET', 'POST', 'DELETE'])
    # manager.create_api(Article, methods=['GET'])
    
    manager.create_api(User, methods=['GET', 'POST', 'DELETE'],
                       postprocessors={
                           'GET_MANY': [detail_sort]
                       })
    manager.create_api(UserDetails, methods=['GET'], )
    
    # start the flask loop
    app.run(use_reloader=False)
    

    Note that you need to use GET_MANY and look how detail_sort is implemented.

    Without using postprocessor the response is like this:

    {
      "num_results": 1, 
      "objects": [
        {
          "id": 1, 
          "userdetails": [
            {
              "detail": "A name", 
              "detail_name": "Full Name", 
              "id": 1, 
              "user_id": 1
            }, 
            {
              "detail": "an@email.com", 
              "detail_name": "Email", 
              "id": 2, 
              "user_id": 1
            }, 
            {
              "detail": "An Address", 
              "detail_name": "Address", 
              "id": 3, 
              "user_id": 1
            }, 
            {
              "detail": "999-999-9999", 
              "detail_name": "Phone Number", 
              "id": 4, 
              "user_id": 1
            }
          ], 
          "username": "plasmy"
        }
      ], 
      "page": 1, 
      "total_pages": 1
    }
    

    With postprocessor the response looks like this:

    {
      "num_results": 1, 
      "objects": [
        {
          "id": 1, 
          "userdetails": {
            "Address": "An Address", 
            "Email": "an@email.com", 
            "Full Name": "A name", 
            "Phone Number": "999-999-9999"
          }, 
          "username": "plasmy"
        }
      ], 
      "page": 1, 
      "total_pages": 1
    }
    

    Hope this helps.