TypeError: no implicit conversion of nil into String about crypt_keeper HOT 5 CLOSED

Are you using master or the latest gem release? The examples below represent master.

Sent from my phone

On Dec 27, 2013, at 5:41 AM, Claudio Poli [email protected] wrote:

I'm giving this gem a quick try, I applied this into my model:

crypt_keeper :token, :secret, encryptor: :aes, key: 'super_good_password', salt: 'salt'
2.0.0p353 :001 > AuthToken.first
TypeError: no implicit conversion of nil into String
gems/aes-0.5.0/lib/aes/aes.rb:76:in update' gems/aes-0.5.0/lib/aes/aes.rb:76:indecrypt'
gems/aes-0.5.0/lib/aes/aes.rb:9:in decrypt' bundler/gems/crypt_keeper-1c9050c4c9db/lib/crypt_keeper/provider/aes.rb:34:indecrypt'
bundler/gems/crypt_keeper-1c9050c4c9db/lib/crypt_keeper/helper.rb:34:in load' bundler/gems/rails-b23d3ffcd157/activerecord/lib/active_record/attribute_methods/serialization.rb:89:inunserialize'
bundler/gems/rails-b23d3ffcd157/activerecord/lib/active_record/attribute_methods/serialization.rb:80:in unserialized_value' bundler/gems/rails-b23d3ffcd157/activerecord/lib/active_record/attribute_methods/serialization.rb:67:intype_cast'
bundler/gems/rails-b23d3ffcd157/activerecord/lib/active_record/attribute_methods/read.rb:101:in block in read_attribute' bundler/gems/rails-b23d3ffcd157/activerecord/lib/active_record/attribute_methods/read.rb:84:infetch'
bundler/gems/rails-b23d3ffcd157/activerecord/lib/active_record/attribute_methods/read.rb:84:in read_attribute' bundler/gems/rails-b23d3ffcd157/activerecord/lib/active_record/attribute_methods.rb:272:inattribute_for_inspect'
bundler/gems/rails-b23d3ffcd157/activerecord/lib/active_record/core.rb:354:in block in inspect' bundler/gems/rails-b23d3ffcd157/activerecord/lib/active_record/core.rb:352:incollect'
bundler/gems/rails-b23d3ffcd157/activerecord/lib/active_record/core.rb:352:in inspect' bundler/gems/rails-b23d3ffcd157/railties/lib/rails/commands/console.rb:90:instart'
In the fetched row secret is nil, however it was my understanding that this should just return the original value instead of raising:
https://github.com/jmazzi/crypt_keeper/blob/master/lib/crypt_keeper/provider/aes.rb#L28

—
Reply to this email directly or view it on GitHub.

from crypt_keeper.

I tried both, I'm afraid the result is the same.

from crypt_keeper.

@masterkain hmm, that's odd. The backtrace mentions line 34 in helper.rb which you can see checks for value.blank?. I'm wondering if your version of blank? has a different behavior than what it normally does.

Which version of rails are you using? Can you comment out the crypt_keeper line in your model and try AuthToken.first.token.blank? and AuthToken.first.secret.blank?? Do they return true?

from crypt_keeper.

uhm, the backtrace does mention serialization.

I have now converted the table values to use with postgres_pgp, but later I'm going to try AES again with a clean set because perhaps it has something to do with my store_accessor :extra hstore attribute. Rails 4.0.1 (4-0-stable)
token and secret however are text columns, not "hstore'd" ones.

from crypt_keeper.

Did you try the examples I posted?

Sent from my phone

On Dec 27, 2013, at 6:02 PM, Claudio Poli [email protected] wrote:

uhm, the backtrace does mention serialization.

I have now converted the table values to use with postgres_pgp, but later I'm going to try again because perhaps it has something to do with my store_accessor :extra hstore attribute. Rails 4.0.1.

—
Reply to this email directly or view it on GitHub.

from crypt_keeper.