Datamapper DataMapper → 1.0.0

Comments and changes to this ticket

• 

  Balint Erdi December 12th, 2008 @ 09:36 AM

  • Tag set to “dm-validations, validations”

  Here goes the failing spec. If anyone could take a quick look at it and confirm that all_valid? was meant to be used this way and this is actually a bug it would be fine.

  Otherwise I'll proceed and try to implement the actual fix.

  • 0001-failing-spec-for-ticket-706-all_valid-goes-into-inf.patch 2.5 KB
• 

  Balint Erdi December 12th, 2008 @ 10:15 AM

  • Title changed from “all_valid? goes into infinite loop on self-referencing many-to-many associations” to “[PATCH] all_valid? goes into infinite loop on self-referencing many-to-many associations”

  See my patch that fixes the issue attached.

  • 0002-recursive_valid-function-does-not-go-into-infinite.patch 5 KB
• 

  Dan Kubb (dkubb) January 8th, 2009 @ 05:13 AM

  • State changed from “unconfirmed” to “accepted”
  • Assigned user set to “Dan Kubb (dkubb)”
• 

  Dan Kubb (dkubb) June 7th, 2009 @ 04:01 AM

  • State changed from “accepted” to “unconfirmed”
  • Assigned user changed from “Dan Kubb (dkubb)” to “Michael Klishin (antares)”
• 

  Martin Gamsjaeger (snusnu) July 9th, 2009 @ 04:39 PM

  • State changed from “unconfirmed” to “confirmed”

  I extracted a standalone script from the pastie and adapted it to work with current next branch (I also had to put the validates_with_block call into Recommendation instead of User as shown in the pastie). As can be seen below, I can confirm that this is still an issue with 0.10.

  require "rubygems"
  require "dm-core"
  require "dm-validations"
  
  DataMapper::Logger.new(STDOUT, :debug)
  DataMapper.setup(:default, "sqlite3::memory:")
  
  
  class User
    include DataMapper::Resource
  
    property :id, Serial
  
    has 0..n, :recommendations
    has 0..n, :recommendees,   :through => :recommendations, :model => "User", :target_key => [:user_id]
    has 0..n, :recommended_by, :through => :recommendations, :model => "User", :target_key => [:recommendee_id], :via => :user
  
  end
  
  class Recommendation
    include DataMapper::Resource
  
    property  :user_id,        Integer, :key => true
    property  :recommendee_id, Integer, :key => true
  
    belongs_to :user
    belongs_to :recommendee, 'User', :source_key => [:recommendee_id]
    
    validates_with_block :recommendee_id do
      if @recommendee_id == @user_id
        [false, "A user can not recommend himself"]
      else
        true
      end
    end
  end
  
  DataMapper.auto_migrate!
  
  u = User.create
  
  u = User.get(u.id)
  u.recommendations
  u.recommendations.create(:recommendee => u)
  u.valid?
  u.all_valid?
  rec = u.recommendations.first
  
  puts rec.instance_variables.map { |ivar| "ivar name: #{ivar} ivar: #{rec.t(ivar).inspect}" }.join("\n")
  
  
  # mungo:snippets snusnu$ ruby balint.rb 
  #  ~ (0.000349) SELECT sqlite_version(*)
  #  ~ (0.000393) DROP TABLE IF EXISTS "users"
  #  ~ (0.000066) DROP TABLE IF EXISTS "recommendations"
  #  ~ (0.000061) PRAGMA table_info("users")
  #  ~ (0.001083) CREATE TABLE "users" ("id" INTEGER NOT NULL PRIMARY KEY AUTOINCREMENT)
  #  ~ (0.000027) PRAGMA table_info("recommendations")
  #  ~ (0.000601) CREATE TABLE "recommendations" ("user_id" INTEGER NOT NULL, "recommendee_id" INTEGER NOT NULL, PRIMARY KEY("user_id", "recommendee_id"))
  #  ~ (0.000096) INSERT INTO "users" DEFAULT VALUES
  #  ~ (0.000065) SELECT "id" FROM "users" WHERE "id" = 1
  # /opt/local/lib/ruby/gems/1.8/gems/extlib-0.9.13/lib/extlib/lazy_array.rb:293:in `respond_to?': stack level too deep (SystemStackError)
  #   from /opt/local/lib/ruby/gems/1.8/gems/dm-core-0.10.0/lib/dm-core/collection.rb:908:in `respond_to?'
  #   from /opt/local/lib/ruby/gems/1.8/gems/dm-validations-0.10.0/lib/dm-validations.rb:108:in `recursive_valid?'
  #   from /opt/local/lib/ruby/gems/1.8/gems/dm-validations-0.10.0/lib/dm-validations.rb:104:in `each'
  #   from /opt/local/lib/ruby/gems/1.8/gems/dm-validations-0.10.0/lib/dm-validations.rb:104:in `recursive_valid?'
  #   from /opt/local/lib/ruby/gems/1.8/gems/dm-validations-0.10.0/lib/dm-validations.rb:111:in `recursive_valid?'
  #   from /opt/local/lib/ruby/gems/1.8/gems/extlib-0.9.13/lib/extlib/lazy_array.rb:458:in `each'
  #   from /opt/local/lib/ruby/gems/1.8/gems/extlib-0.9.13/lib/extlib/lazy_array.rb:458:in `each'
  #   from /opt/local/lib/ruby/gems/1.8/gems/dm-validations-0.10.0/lib/dm-validations.rb:109:in `recursive_valid?'
  #    ... 4980 levels...
  #   from /opt/local/lib/ruby/gems/1.8/gems/dm-validations-0.10.0/lib/dm-validations.rb:104:in `each'
  #   from /opt/local/lib/ruby/gems/1.8/gems/dm-validations-0.10.0/lib/dm-validations.rb:104:in `recursive_valid?'
  #   from /opt/local/lib/ruby/gems/1.8/gems/dm-validations-0.10.0/lib/dm-validations.rb:97:in `all_valid?'
  #   from balint.rb:46
  # mungo:snippets snusnu$
  

• 

  Martin Gamsjaeger (snusnu) July 9th, 2009 @ 07:45 PM

  • Assigned user changed from “Michael Klishin (antares)” to “Martin Gamsjaeger (snusnu)”
  • State changed from “confirmed” to “accepted”
  • Milestone set to “0.10.0”
• 

  Martin Gamsjaeger (snusnu) July 11th, 2009 @ 12:04 PM

  I think we should look into this once the next branch supports self referential m:m. Since that is currently not possible, it makes more sense to wait till that's known to work?

• 

  Dan Kubb (dkubb) October 4th, 2009 @ 09:33 PM

  • Milestone changed from “0.10.0” to “0.10.2”

  [project:id#20609 not-tagged:"0.10.0" milestone:id#51895 bulk edit command]

• 

  Dan Kubb (dkubb) October 4th, 2009 @ 09:39 PM

  Resource#all_valid? is no longer part of dm-validations, since it duplicates functionality in Resource#valid? now, and the implementation is not as good.

  I'm going to leave this ticket open so that we can verify that Resource#valid? does not have a similar issue with m:m.

• 

  Dan Kubb (dkubb) February 1st, 2010 @ 04:33 PM

  • Milestone changed from “0.10.2” to “1.0.0”

  [bulk edit]

• 

  Martin Gamsjaeger (snusnu) March 30th, 2010 @ 05:35 PM

  • State changed from “accepted” to “resolved”
  • Assigned user cleared.

  I think it's confusing to keep this ticket around while we don't even have the #all_valid? method anymore. I'm marking this resolved for now. If a problem crops up again, I'm almost sure that the potential reporter wouldn't "find" this ticket and instead would open up a new one anyway.