Padrino Model from JSON Data

Padrino model from json data

Yep, you can use provides and each response object will call to_json i.e:

get :action, :provides => :json do
  @colletion = MyCollection.all
  render @collection # will call @collection.to_json
end

Here an example of an ugly code that fills certain models.

# Gemfile
gem 'json' # note that there are better and faster gems like yajl

# controller
post "/update/:model/:id", :provides => :json do
  if %w(Account Post Category).include?(params[:model])
    klass = params[:model].constantize
    klass.find(params[:id])
    klass.update_attributes(JSON.parse(params[:attributes]))
  end
end

Finally if you POST a request like:

attributes = { :name => "Foo", :category_id => 2 }.to_json
http://localhost:3000/Account/12?attributes=#{attributes}

You'll be able to update record 12 of the Account Model.

Storing json object in Padrino CouchDB model

It's not necessary to stringify json, just parse it and store as an array using the JSON standard lib.

original JSON data example:
json_data = [{ prop1: val1, prop2: val2, prop3: val3 }] #store the JSON data from wherever it came
parsed = JSON.parse(json_data.to_json) #store the parsed JSON data into an Array

foo = MyModel.find(1) #find whatever collection you wanna update
foo.properties = parsed
# => [{ prop1: val1, prop2: val2, prop3: val3 }]
foo.save

Your MyModel should then look like the following:

{
  "_id" : "your_generated_id",
  "_rev" : "your_generated_rev",
  "name" : "UniqueName",
  "properties" : [
         { 
           prop1: val1, 
           prop2: val2, 
           prop3: val3 
         }
  ]
}

Remember CouchDB is "schemaless". More details here.

Why is Padrino trying to use a template when rendering a json response?

So I got it to work by replacing render @records, layout: false with @records.to_json.

Pretty sure that's not how it's supposed to work, so if anyone knows a better way I'm all ears, otherwise I'll accept this answer.

Padrino app with REST API

Blog::App.controllers :posts, map: '/api/v1/posts' do

  get :index do
    ...
  end

end

And then, if you want add new versions of that controller

Blog::App.controllers :v2_posts, map: '/api/v2/posts' do

  get :index do
    ...
  end

end

(Yeah, it seems you can't have several files with the same controller with different map values.)

So, this won't work (sorry if that works, that doesn't when I tried) correctly and will cause issues :

Blog::App.controllers :posts, map: '/api/v1/posts' do

  get :index do
    ...
  end

end

Blog::App.controllers :posts, map: '/api/v2/posts' do

  get :index do
    ...
  end

end

How can I clean up this Padrino AJAX call?

first of all, you should protect your model from mass assignments using attr_accessible and attr_protected aka mass assignment.

Then I highly suggest to use "forms" so your site can work without javascript enabled.

So using unobtrusive javascripts code can be also nicer.

// Live watch events on form submit
$(document).on('submit', 'form[data-remote]', function(e){
  e.preventDefault();
  self = $(this);
  $.ajax({
    url: self.attr('action'),
    data: self.serializeArray(),
    type: self.attr('method'),
    dataType: 'json',
    success: function(res){ console.log(res) }
  })
});

Here the form:

/* Here the form, for dates I suggest to use a calendar */ 
/* like https://github.com/arshaw/fullcalendar */
- form_for @trip, url(:trip, :index), :'data-remote' => true do |f|
  == f.error_messages
  == f.text_field :title
  == f.text_area :description
  == f.text_field :departure_date
  == f.text_field :return_data
  == f.submit 'Send'

Here the controller:

provides :json # if all your actions in your controller respond to #json it is more dry

get :index do
  @trip = Trip.new(params[:trip])

  if @trip.save
    render :success => true, :attributes => @trip.to_json
  else
    render :success => false, :errors => @trip.errors.full_messages
  end
end

LoadError on Windows when trying to generate a model

Put sqlite3.dll to C:\Ruby193\bin

---

Related Topics