Redirect_to in Rails 4 - ruby-on-rails

Question

For example I have a controller PostController in RoR 4. and actions in it:
def index
#posts = Post.all.order('created_at DESC')
end
def new
#post = Post.new #second question
end
def create
#post = Post.new(post_params)
post.save
redirect_to #post #first question
end
def show
#post = Post.find(params[:id])
end
private
def post_params
params.require(:post).permit(:title, :body)
end
so, the first question is: Why when we write redirect_to #post it automatically redirect to show/id? How rails know, that it must go to show view?
second question: why I must write #post = Post.new in new action? when I comment it, I also can create a Post. What is the deference write it in new action or missing it?
I'm new in rails and there are many magic in it

Answer 1

Q1). This is part of one of main RoR's mantras - convention over configuration. It just assumes, when performing redirect_to #post, you actually want to see the #post resource. The standard REST action of "reading" the resource is show.
But, if - for example - in your view you will do something like
<%= link_to #post.title, #post, method: :delete %>
What will be generated by that will take you to destroy action of #post's resource.
Having conventions like this make you write less code, and - make you happy!
Q2). I'm surprised if what you mentioned worked. It shouldn't, because similar magic is performed in your form_for #post tag in your views. If #post is empty, form_for shouldn't be able to resolve what url to generate, or how to behave in general - depending on Post.new stored in #post it assumes you want to create the resource, and prepare your <form> tag to submit with POST method, but if in #post there is stored already persisted resource, say Post.find(1), it will create the <form> to be submitted with PATCH method.
Again - based on convention, you can write less code, to achieve more.
I strongly advise against omitting something like #post = Post.new. It may lead to hard to identify bugs, even if it worked for you at the moment
Hope it clarifies things!
Good luck!

Answer 2

First questions: it redirects you to the show page using the url_for method (you can read its source if you are curious), remember that #book has an id so it takes that id and includes it in the url of the /books/:id
Second question : you need #post for when you submit an invalid form, then #post will be populated with the errors and displayed on the new page. try to submit an invalid post, you will get complains about a not defined method.
to display the errors of an invalid submission :
<%= form_for :post posts_path do |f| %>
<% if #post.errors.any? %>
See how we used #post above? just to get the errors in it.

Answer 3

When passing a record to redirect_to the url will be generated by calling url_for, which will return a named url for that record. For this it uses the class name for lookup. So passing the Post record will attempt to use the post_path route with the original record as parameter which will resolve in /posts/:id.

Answer 4

First question
Rails magic.
As you pass an ActiveRecord object to redirect, the restful routing interprets this as you wanting to go to the show page of that object.
Plenty of different ways to use redirect_to here.
Second question
Usually #post = Post.new is required as you reference #post in the new action as your forms when constructed with form_for .. useful for validation and dealing with errors.
Technically you could use a form_tag and define Post.new right there in the form. This would mean the #post = Post.new is redundant and works when you comment it out. This way would mean you have no way to pass your objects errors back to the user on the new page though.