Rails Ruby on Rails → 3.0.2

Comments and changes to this ticket

• You flagged this item as spam.
  

  Simone Carletti March 17th, 2009 @ 05:09 PM

  I get the same error but running functional tests.

  I've the following method

  
  def something
  user_cookie = cookies[:user_cookie] || session.session_id
  end
  

  changing the method definition to

  
  def something
  user_cookie = cookies[:user_cookie] || request.session_options[:id]
  end
  

  causes the following error.

  
  NoMethodError: You have a nil object when you didn't expect it!
  You might have expected an instance of ActiveRecord::Base.
  The error occurred while evaluating nil.[]
  app/controllers/application_controller.rb:38:in `something'
      test/functional/about_controller_test.rb:16:in `test_index'
  

• You flagged this item as spam.
  

  Pawel March 18th, 2009 @ 03:08 PM

  • Tag changed from “2.3, rails, request.session:id, session.session_id” to “2.3, rails, request.session_options:id, session.session_id”

  This is because of session lazy loading.

  You can work around this by loading the session first and then read its id.

  
  def something
    session[:foo]
    user_cookie = cookies[:user_cookie] || request.session_options[:id]
  end
  

  However I think there should be a way to read session id without loading the session itself (think of database backed sessions) -- see #2177.

• You flagged this item as spam.
  

  edbond2 June 30th, 2009 @ 06:57 AM

  • Assigned user set to “josh”
  • Tag changed from “2.3, rails, request.session_options:id, session.session_id” to “2.3, cookie, rails, request, session, test”
• 

  josh July 9th, 2009 @ 07:32 PM

  • State changed from “new” to “open”

  I suppose trying to read the session id should touch the cookie store.

  Can someone try to create a patch for this?

  Also, this should only affect the cookie store. For an example, with the AR store, we don't want to load the session because we already know the session id.

• 

  josh July 23rd, 2009 @ 03:08 AM

  • State changed from “open” to “incomplete”
• 

  josh August 19th, 2009 @ 03:55 PM

  • Assigned user cleared.

• You flagged this item as spam.
  

  John Pignata August 19th, 2009 @ 05:41 PM

  I'm not sure how to retrieve the ID without unmarshaling the cookie data and it seems reasonable to me to load this data if the id is requested so I've created a patch to do that. This should apply to 2-3-stable. Feedback welcomed - not sure what the best way is to fix this.

  • load_session_if_session_options_accessed_2_3.diff 1.7 KB
• 

  josh August 31st, 2009 @ 10:20 PM

  • Milestone cleared.

  CookieStore should probably wrap session.options with a lazy hash that triggers the unmarshaling.

• You flagged this item as spam.
  

  John Pignata September 2nd, 2009 @ 03:30 AM

  Thanks, Josh. Here's another go at it - against master this time.

  • load_session_if_session_options_id_accessed.diff 3 KB
• 

  josh September 2nd, 2009 @ 03:39 AM

  • Assigned user set to “josh”
  • State changed from “incomplete” to “open”
• 

  Repository September 3rd, 2009 @ 05:50 PM

  • State changed from “open” to “resolved”

  (from [e0f1a7dc191ffebc9f6cadb6232e567fee8aa491]) If session_options[:id] is requested when using CookieStore, unmarshal the session to access it [#2268 state:resolved]

  Signed-off-by: Joshua Peek josh@joshpeek.com
  http://github.com/rails/rails/commit/e0f1a7dc191ffebc9f6cadb6232e56...

• You flagged this item as spam.
  

  John Pignata September 3rd, 2009 @ 06:29 PM

  Looking at your commit, it seems like I made an error when testing the bugfix. CookieStore should be deriving from Hash. I've made a patch to revert this unnecessary change.

  • cookie_store_should_not_derive_from_hash.diff 1 KB
• 

  Repository September 3rd, 2009 @ 09:15 PM

  (from [bd97c3044a7b135f5b84f38c3dbdce2ccc793f70]) CookieStore should not be derived from Hash - reverting [#2268 state:resolved]

  Signed-off-by: Joshua Peek josh@joshpeek.com
  http://github.com/rails/rails/commit/bd97c3044a7b135f5b84f38c3dbdce...

• 

  Jeremy Kemper October 15th, 2010 @ 11:01 PM

  • Milestone set to “3.0.2”
  • Importance changed from “” to “”

  [bulk edit]