Notes posted to Ruby on Rails

Moved from ActiveRecord to ActiveModel

Don’t let the depreciation warning scare you, Dirty lives under ActiveModel as of 3.0.0

apidock.com/rails/ActiveModel

Returns a string used for submit button label

This private method returns the label used for the `f.submit` helper.

apidock.com/rails/v5.2.3/ActionView/Helpers/FormBuilder/submit

<%= form_for @post do |f| %>
  <%= f.submit %>
<% end %>

In the example above, if @post is a new record, this method returns “Create Post” as submit button label; otherwise, it uses “Update Post”.

This method also checks for custom language using I18n under the helpers.submit key and using %{model} for translation interpolation:

en:
  helpers:
    submit:
      create: "Create a %{model}"
      update: "Confirm changes to %{model}"

It also searches for a key specific to the given object:

en:
  helpers:
    submit:
      post:
        create: "Add %{model}"

Access just the label text

One of the handiest features of `f.submit` is the auto-generated label text.

Some styling frameworks don’t lend themselves to using `f.submit` without lots of tweaking.

If you’re not using `f.submit` but still want to access the label text, note that there is a private method:

f.send(:submit_default_value)

One example (in Haml, using MaterializeCss):

.row
  .col.s6.center
    %button{ type: 'submit', name: 'commit', data: { disable: { with: 'Submitting...' } }, class: 'waves-effect waves-light btn maroon' }
      = f.send(:submit_default_value)

Text improvement

The statement about not loading a bunch of records is correct because pluck returns an Array of values, not a ActiveRecord::Relation or an Array of Records.

The exact wording though may be:

“…without retrieving from the database unused columns or loading a bunch of records just to grab the attributes you want”

Works even on nested structures

@EdvardM added a note years ago saying this does not symbolize the keys of deeply nested hashes, which may have been the case for whatever version of ruby was available at the time, but in 4+, it definitely does work as described:

hash = {"a" => :a, "b" => {z: [[{"c" => 3}, {"c" => 4}], []]}}
hash.deep_symbolize_keys
=> {:a=>:a, :b=>{:z=>[[{:c=>3}, {:c=>4}], []]}}

expect_any_instance_of(ActiveRecord::Relation).to receive(:create_with) and others do not work

If you’re doing

expect_any_instance_of(ActiveRecord::Relation).to receive(:create_with) 

and it does not work, try:

expect_any_instance_of(ActiveRecord::Associations::CollectionProxy).to receive(:create_with) { |proxy, attributes|
  expect(proxy.klass).to eq(RecordClass)
  expect(attributes[:....]).to eq(...)
  double('find_or_create_by!' => Proc.new {})
}

or when testing “find_or_create_by”

expect_any_instance_of(ActiveRecord::Associations::CollectionProxy).to receive(:find_or_create_by) { |proxy, attributes|
  expect(proxy.klass).to eq(RecordClass)
  expect(attributes[:....]).to eq(...)
}

to_sentence_exclusive

By default, to_sentence combines elements inclusively by using ", and ".

If you want to be exclusive and combine the elements using ", or ", you can either pass in lengthy options to to_sentence or use this handy extension I made to add to_sentence_exclusive to the Array class.

class Array

  # Adds a simple method that overloads `to_sentence` except it uses "or" instead of "and", which is the default.
  # The reason for this is because `to_sentence` is extremely flexible but that results in a lot of code to get
  # the simple result of using "or" instead of "and". This makes it simple.
  #
  def to_sentence_exclusive
    self.to_sentence( two_words_connector: " or ", last_word_connector: ", or " )
  end

end

Just drop this in config/initializers/to_sentence_exclusive.rb, restart your server or console and Bob’s your uncle.

Thanks davinjay!

Your note was SUPER helpful so I wanted to leave more than just a “1 thank”.

Cheers!

This method has been completely removed from Rails 5 onwards

Completely removed from Rails from 5 onwards. See issue: github.com/rails/rails/issues/18336

Just remove from your codebase, or protect with `private` keyword

Upload Files Directly To S3 Using Paperclip And Dropzone.js

Upload Files Directly To S3 Using Paperclip And Dropzone.js

by | Dec 22, 2017 | Technical Articles | 0 comments

It’s usually the small time-consuming tasks that frustrate us the most. Such as uploading a file to S3; the requirement is pretty simple but the method chosen to upload the file will decide the efficiency of the task. As uploading files is a feature that most applications require, RailsCarma has compiled a brief tutorial on one of the best methods of getting this task done efficiently: using Paperclip and Dropzone.js.

Paperclip is a popular choice for uploading images and files as it offers great features to handle the attachments; ‘paperclip’ gem is the go-to option. Paperclip allows you to upload multiple images and files, generate thumbnails and even automatically resize the images. It boasts of a large and active community making it the top choice of most developers.
Dropzone.js is an open source library with file drag & drop (with image preview) features.
Amazon S3 is a simple storage device for data storage. We can use it to retrieve images and all type of files.

Why Paperclip?

Paperclip is a popular file uploading tool for the following reasons:

Supports File Caching:
If a form fails to validate, we don’t want the user to pick his file again and re-upload it. Therefore, file caching is necessary from a UX standpoint. And it also conserves the bandwidth.

Processes Images
Paperclip is able to resize and crop images to several different formats thus allowing the developer to choose the library.

Simplifies The Task!
Paperclip gem does not pollute your code and is easy to test!

Allows File Processing
Paperclip allows file processing for EXIF data extraction and thumbnail creation of uploaded PDFs, PSDs, DOCs, XLSXs.

Provides CDN & Storage-Service Support 
This is a big plus as we want to keep the bandwidth to our servers as low as possible and avoid possible data loss due to server failure.

Offers On-The-Fly Processing
Paperclip processes images and files on a per-request basis. This is an innovative feature that enables developers to create custom content that adapts best to different situations.

What Are The Dropzone Asynchronous Events?

addedfile:  When a file is added to the list.
removedfile: Used whenever a file is removed from the list. You can listen to this and delete the file from your server if you want to.
thumbnail: When the thumbnail has been generated. It receives the data URL as second parameter.
error: An error occurred receives the error message as the second parameter. And if the error was due to xmlhttprequest, the xhr object is received as the third parameter.
processing: When a file is processed (since there is a queue, not all files are processed immediately). This event was previously called processingfile.
drop: The user dropped something onto the drop zone.

How Can We Configure Paperclip In Our Application?

has_attached _file: asset
:storage => :s3
:S3_host_name => ENV[“S3_HOST_NAME”]
:S3_region => ENV[“S3_REGION”]
:S3_protocol => ENV[“S3_PROTOCOL”]
:path => “:account_id/:class/:source_id/:attachment/:file_name”,:s3_headers => {‘ContentDisposition’ => ‘attachment’, ‘content-type’ =>‘application/octet_stream’},
:bucket => ENV[“S3_BUCKET”],
:s3_credentials => Proc.new{|a| a.instance.s3_credentials}
Do_not_validate_attachment_file_type :asset
def s3_credentials
{:access_key_id => ENV[“S3_ACCESS_KEY_ID”], :secret_access_key => ENV[“S3_SECRET_ACCESS_KEY”]}

end 

How Can We Handle Custom Paths In Our Application? Read More from here http://www.railscarma.com/blog/technical-articles/upload-files-directly-s3-using-paperclip-dropzone-js/

If you need to check for overlaps in Times, use the validates_overlap gem.

github.com/robinbortlik/validates_overlap

When using ActionView::Base.new to render templates views

when calling this method to render templates to a string. in order to use any helper methods you need to add them to the view like this

view = ActionView::Base.new(ActionController::Base.view_paths, {})
view.class_eval do  
  # include any needed helpers (for the view)
  include ApplicationHelper
end 

source: http://peden.biz/rendering-a-rails-view-from-a-script/

Each attribute has a `reset_<attribute>!` method on it as well.

So if the attribute is name you can call reset_name! on the object to reset the dirty changes.

method is working until rails 4

deprecation message and rails line (till v 2.3.8) is not correct. Method exist and working until rails 4.

Correction: Getting just the ordinal on Rails 3.

The ordinal method isn’t publicly available in Rails 3 so you can do something like this:

ordinalize(1).last(2) #=> "st"

ordinalize(20).last(2) #=> "th"

Effectively identical to Hash#as_json

As of 5.2.0.beta, there is no ActiveRecord::Relation specific implementation. This will result in Object#as_json , which will convert the relation to a hash and call Hash#as_json .

class Object
  def as_json(options = nil) #:nodoc:
    if respond_to?(:to_hash)
      to_hash.as_json(options)
    else
      instance_values.as_json(options)
    end
  end
end

Counting with select

If you try to write

Model.select('field_one', 'field_two AS something').count

it will fail (at least for Rails 5.0) with the message PG::SyntaxError: ERROR: syntax error at or near “AS”. In order to fix that issue, you should write

Model.select('field_one', 'field_two AS something').count(:all)

Getting just the ordinal on Rails 3.

The ordinal method isn’t publicly available in Rails 3 so you can do something like this:

ordinal(1).last(2) #=> "st"

ordinal(20).last(2) #=> "th"

PAGER DUTY & EXCEPTION NOTIFIER PLUGINS FOR RAILS

RAILS EXCEPTION NOTIFIER

The Exception Notifier plugin provides a mailer object and a default set of templates for sending email notifications when errors occur in a Rails application. It is basically a monitoring tool, which keeps on watching the application and whenever it finds any error, it triggers that error to PagerDuty. To use Exception Notification and PagerDuty in your app, you need to add this gem below:

gem 'exception_notification', '~> 4.1.0'

gem  'pagerduty''

To get the email notifications, you need to include the line below in the development env:

Rails.application.config.middleware.use  ExceptionNotification::Rack,
                 :email => {
                  :email_prefix => "[PREFIX] ",
                  :sender_address => %{"notifier"  <notifier@example.com>},
                  :exception_recipients => %w{exceptions@example.com},
                  :pd => {
                           # simple notifier options
                         }
               }

You can modify sender’s and recipient’s address.

Rails App+PagerDuty

Use the code below in your app with exception notifier to connect with PagerDuty:

require "pagerduty"
module ExceptionNotifier
      Class PdNotifier
          def initialize(options)
            @pagerduty =   Pagerduty.new("0bdcfdacf1b144d7822dfdfa5ed0ab1e")# Service api key
           # do something with the options...
         end
        def call(exception, options={})
           @pagerduty.trigger(exception.message, details:  { backtrace: exception.backtrace })
        end
    end
end

Conclusion

PagerDuty is alert dispatching tool used by operations team/OnCall Engineers to manage the applications and it is popular because of its reliable & rich services(Scheduling,Alerting,Reporting,Call Routing , Feedback & response time).

Create your Free account from app.pagerduty.com/ and integrate with your application to get the flow , how Incident is triggered.

How to use Textacular Gem to search data in your Rails Application

We might have heard about a lot many gems which let us implement search functionality in our rails application; for example: searchkick, elasticsearch-rails, ransack and finally, sunspot to work with solr search engine. All these gems have their own advantages. Both searchkick and elasticsearch use redis to search the data as well as need to perform a ‘reindex’ while inserting new data. In one of my recent projects, I happened to use a gem called as Textacular. It’s simple and very easy to use. Textacular Gem:

It is a gem that provides full text search capabilities for PostgreSQL Database. It basically caters to extend the scope of the work performed by activerecord, in a rather friendly manner. It works on heroku as well. This gem works only on PostgreSQL For working with it, let’s first grab the latest textacular gem from rubygems.org/gems/textacular and add it to the gemfile.

gem 'textacular'
    bundle install

Textacular gem provides us with quite a few methods to search the data. So, all our models have the access to use those methods.

basic_search advanced_search fuzzy_search

Usage: Basic_search: It searches based on the input text.

User.basic_search(‘abc’) # Searches on all the model column attributes

User.basic_search(last_name: 'abc', first_name: 'xyz')

Advanced_search: Here, we can use postgres syntaxes like !, & and | (and, or and, not) and then, some others based on the requirement. It converts user’s search DSL into Pg syntax. For this, we need to make sure that the necessary exceptions should be used to handle the syntax errors.

User.advanced_search(last_name: 'text1|text2’) - It  searches with the text1 or text2 on last_name on User  model.

User.advanced_search(last_name: '!text2’) - It searches for the records whose last_name is not text2.

These searches can be chainable as shown below:

User.advanced_search(last_name: 'text1|text2’).basic_search(last_name: 'abc', first_name: 'xyz')

Fuzzy_search: We need to install pg_trgm module to work with fuzzy_search. Run the command below to install this module. It searches for partial appearance of your text in the DB.

rake textacular:create_trigram_migration  
rake db:migrate

Now, we are ready to use fuzzy_search.

User.fuzzy_search('Lorem')

By default, fuzzy search, searches for the records which are 30% of the search text matches with respect to the entire content. We can set this threshold limit by using the command below.

ActiveRecord::Base.connection.execute("SELECT set_limit(0.6);")

So, it expects 60% of search text to match with the original content. We can use OR condition to search on multiple columns. Need to pass a hash with columns with input text as param one and pass second param as a false. It takes AND, if you miss second param or if it True.

User.fuzzy_search({first_name: 'user', last_name: 'test'}, false)

User.fuzzy_search(first_name: ‘user’, last_name: 'test') - It takes AND condition.

By default, the Textacular searches in all text and string columns. We can alter its default behaviour by overriding the searchable_columns method in the model.

def self.searchable_columns
   [:title, :body]
end

We can override self.searchable_language in the model to set proper dictionary settings.

def self.searchable_language 
  'arabic' 
end

Read More : http://www.railscarma.com/blog/technical-articles/how-to-use-textacular-gem-to-search-data-in-your-rails-application/

multiple select has a hidden

It’s unclear if “select_tag” does this but the normal select method also generates a hidden variable if the “multiple” option is set, see http://apidock.com/rails/ActionView/Helpers/FormOptionsHelper/select

not submitted even if non-multiple in certain cases.

As a note, just for followers, even if a select is not “multiple” if it is “unselected” (ex: in javascript you can set its value like document.getElementById(‘select_id’).value = ‘a non option’; ). And if that form is submitted, the browser also seems to not send anything about it to the server. So it’s for non-multiples as well, just this case is rare since typically the select will default to its “first value” and then users can only change it another known value, so typically you won’t run into that. This isn’t related to rails but thought I’d mention it.

Converting Hash to Struct in Ruby

With the fact we know about the two, many will perhaps prefer using struct rather than hash. But what we want to focus here is that behind the low performance given by hash, they still can become advantageous. If you can’t use it on its own little way, then convert it into struct so better usability can be achieved.

If you have already defined the struct and you wanted to initiate converting hash to struct, we can help you by using the following methods in a given example below. If you wanted to convert has to a struct in Ruby, let us say for example we have a given of:

h = { :a => 1, :b => 2 }

and want to have a struct such as:

s.a == 1
s.b == 2

To convert, you can do any of these methods: Conversion Method 1:

On this method, the result will appear to be OpenStruct and not specifically as struct:

pry(main)> require 'ostruct'
pry(main)> s = OpenStruct.new(h)
=> #<OpenStruct a=1, b=2>
pry(main)> puts s.a, s.b

Conversion Method 2:

If you have struct defined already and want to start something with a hash, you can follow this:

Person = Struct.new(:first_name, :last_name, :age)

person_hash = { first_name: "Foo", last_name: "Bar", age: 29 }

person =  Person.new(*person_hash.values_at(*Person.members))

=> #<struct Person first_name="Foo", last_name="Bar", age=29>

Conversion Method 3:

Since the hash key order was guaranteed in the Ruby 1.9+, you can follow this:

s = Struct.new(*(k = h.keys)).new(*h.values_at(*k))

The hash to struct conversion example we provided can help, but if you want a more extensive idea, come to professionals for formal assistance. Read More From Here http://www.railscarma.com/blog/technical-articles/guide-converting-hash-struct-ruby/

association our data

we can load our data whenever we want.

ActiveRecord::Associations::Preloader.new.preload(@users, :company)

Preload our own with out incude

we can preload our data whenever we want.

ActiveRecord::Associations::Preloader.new.preload(@users, :address)

How to use belongs_to in Ruby on Rails

Rails 4: Let’s assume we have two models Movie and Actor. Actor has many movies and Movie belongs to Actor. In most of the cases, the developers used to validate the foreign key actor_id is present or not. In this case, it will not validate actor model as to whether the entered actor id exists or not. It is just like attribute validation which always validates that the fields should be non-empty when submitting the form.

Validation with foreign key:
class Movie < ApplicationRecord
belongs_to :actor
validates :actor_id, presence: true
end

class Movie < ApplicationRecord
belongs_to :actor
validates :actor_id, presence: true
end

class Actor < ApplicationRecord
has_many :movies, dependent: :destroy
end

class Actor < ApplicationRecord
has_many :movies, dependent: :destroy
end

The scenario above, validates actor id presence or not like normal attribute presence validator.

Ex: Actor.create(title: “abc”).
=> {id: 1, title: 'abc'}
m = Movie.new(title: “ok ok”, actor_id: 111)
=> m.valid? => true
=> Actor.find(111)
ActiveRecord::RecordNotFound: Couldn't find Actor with 'id'=111

Ex: Actor.create(title: “abc”).
=> {id: 1, title: 'abc'}
m = Movie.new(title: “ok ok”, actor_id: 111)
=> m.valid? => true
=> Actor.find(111)
ActiveRecord::RecordNotFound: Couldn't find Actor with  'id'=111

We can still save the movie record without valid actor.

With associations

class Movie < ApplicationRecord
belongs_to :actor
validates :actor, presence: true
end

class Movie < ApplicationRecord
belongs_to :actor
validates :actor, presence: true
end

(or)

class Movie < ApplicationRecord
belongs_to :actor, required: true
end

class Movie < ApplicationRecord
belongs_to :actor, required: true
end

class Actor < ApplicationRecord
has_many :movies, dependent: :destroy
end

class Actor < ApplicationRecord
has_many :movies, dependent: :destroy
end

Ex: Actor.create(title: “abc”).
==> {id: 1, title: 'abc'}
m = Movie.new(title: “ok ok”, actor_id: 111)
==> m.valid? => false
==> m.errors.full_messages, ['Actor can't be blank']

Ex: Actor.create(title: “abc”).
==> {id: 1, title: 'abc'}
m = Movie.new(title: “ok ok”, actor_id: 111)
==> m.valid? => false
==> m.errors.full_messages, ['Actor can't be blank']

In this case, it will always validate whether the entered actor exists or not. In case of an invalid actor it throws error to you. This is the best practise to make your associations. It always checks for the associated object exists or not.

Rails5 From rails5 these validations were added by default. It validates association object should be present when you define belongs_to associations.

Release notes http://guides.rubyonrails.org/5_0_release_notes.html(belongs_to will now trigger a validation error by default if the association is not present.)

We can opt out of this feature completely from the application by setting config opton in initializer file.

config/initializers/new_framework_defaults.rb

config/initializers/new_framework_defaults.rb

Rails.application.config.active_record.belongs_to_required_by_default = false Rails.application.config.active_record.belongs_to_required_by_default = false

This initializer file is present in rails5 application only. Need to add this initializer file manually when you migrate from older version of your rails application and make necessary changes.

class Post < ApplicationRecord
has_many :comments, dependent: :destroy	
end

class Post < ApplicationRecord
has_many :comments, dependent: :destroy	
end

class Comment < ApplicationRecord
belongs_to :post
end

class Comment < ApplicationRecord
belongs_to :post
end

c = Comment.create(title: “awesome post”)
c.errors.full_messages.to_sentence
=> “Post must exist”

c = Comment.create(title: “awesome post”)
c.errors.full_messages.to_sentence
=> “Post must exist”

We can not create any comment record without an associated record.

We can opt out this default behaviour by setting

belongs_to :post, optional: true
belongs_to :post, optional: true
c = Comment.create(title: “awesome post”)
c = Comment.create(title: “awesome post”)
=> <Comment id: 1, title: “awesome post”, post_id: nil>

http://www.railscarma.com/blog/technical-articles/use-belongs_to-ruby-rails/

Used within collection_check_boxes

Note the checked: option.

%table
  = collection_check_boxes(:ahj_types, :ids, AhjType.order(:TypeName), :AHJTypeID, :TypeName) do |b|
    %tr
      %td{style: 'padding: 0 1em;'}
        = b.label(class: "check_box")
      %td{style: 'padding: 0 1em;'}
        = b.check_box(class: "check_box", checked: (params[:ahj_types][:ids].member?(b.value.to_s)))

dealing with semicolon

Use tag! method if you have semicolon, for example:

xml.tag!(“atom:link”, “href”=>“http://rubyplus.com/episodes.rss”, “rel”=>“self”, “type”=>“application/rss+xml”)

rails date_select examples

I have combined some of the tips here: http://www.rubyplus.net/2016/11/rails-dateselect-examples.html

Track Changes To Your Model’s Data with Paper Trail Gem

Paper trail lets us track all the changes in the models data for the purpose of its editing and versioning. By using this gem, we can see how a model looks, at every stage of its life-cycle and we can take it back to any version of the model data and we can even undo all the changes after a record has been destroyed so as to restore it completely.

gem ‘paper_trail’ run bundle install to install it

After bundle install you can run the following command for adding versions table to your database

bundle exec rails generate paper_trail:install

bundle exec rake db:migrate

Add has_paper_trail method to your models for tracking.

class Product < ActiveRecord::Base
 	has_paper_trail

end

If you are using a current_user method then you can track who is responsible for particular change by the below callback

class ApplicationController

before_action :set_paper_trail_whodunnit end

Features of paper trail gem

It stores each and every change happened in the model  like create, update and destroy
It does not store any updates unless if any modifications
It allows you to get every version including the actual and even once destroyed

Basic Usage

Now you have a versions method which returns all the changes in particular model

product = Product.find 2

product.versions

# [<PaperTrail::Version>, <PaperTrail::Version>, ...]

Based on the vesrion you can find the changes that happened in a model

v = product.versions.last

   v.event # it returns update / create / destroy
   v.whodunnit #it returns a user id who did this
   v.reify the poduct as it was before the update

you can navigate versions by using previous_version and next_version methods

product = product.previous_version

   product = product.next_version #it returns nil if there is no object

   product.versions.last.whodunnit #it returns the user who did the particular change

For More Read From Here : http://www.railscarma.com/blog/technical-articles/track-changes-models-data-paper-trail-gem/