Sprig-Reap created also yml-files for habtm-association-tables like this example below. But Sprig requires in seeds.rb to specify a classname. So what's the benefit of having this generated files? How can I use them meaningfully?

records:

• sprig_id:
  package_id: 1
  station_id: "<%= sprig_record(PayTv::Station, 1).id %>"
• sprig_id:
  package_id: 2
  station_id: "<%= sprig_record(PayTv::Station, 1).id %>"

When using the gem pg_search spring-reap will fail with

ActiveRecord::StatementInvalid: PG::UndefinedTable: ERROR:  relation "pg_search_documents" does not exist
LINE 1: SELECT "pg_search_documents".* FROM "pg_search_documents"
                                            ^
: SELECT "pg_search_documents".* FROM "pg_search_documents"
-e:1:in `<main>'
Tasks: TOP => db:seed:reap

So, in general it would be nice if standard errors would be skipped and in summary will be shown what couldn't be serialized.

Spring creates the pg_search_documents.yml although the exception was thrown so it would be also useful if the empty file wouldn't be created or at least removed if it is empty.

Thank you for your work! 👍

When I reap simple tables with no associations it seems to work fine, for example if I reap ONLY User and Community, individually, or together.

However, when I try to reap User, Community, AND UserProfile I consistently get one of the following two errors (depending upon what subset of the data I try to reap):

no implicit conversion of nil into Stringr UserProfile:
(possible clue: that extra 'r' after the word String?)
or
undefined methodfetch' for false:FalseClasserProfile:`

the UserProfile table has 2 'parent' tables:
belongs_to :user
accepts_nested_attributes_for :user
belongs_to :community
accepts_nested_attributes_for :community
(User has_one :user_profile, and COmmunity has_many :user_profiles)

Even when I define a method self.for_seed_db which is so specific it selects a single User and UserProfile, and run

rake db:seed:reap TARGET_ENV=development MODELS=UserProfile.for_seed_db,User.for_seed_db,Community

I get the error no implicit conversion of nil into Stringr UserProfile:
but it correctly reaps the single User and the set of Communities.

Adding OMIT_EMPTY_ATTRS=true doesn't help.

Any thoughts what could prevent a specific table from being reaped and giving either of those two errors?

Rails 3.2.21, ruby 2.1.5, running on a Mac.

As far as I can tell from reviewing the code and writing some additional tests, sprig-reap supports supplying AR relations for the list of models to reap, and then only reaping models returned by those relations. Is there any reason this isn't called out in the documentation? Is this just a coincidence of implementation that my not be supported going forward, or a true feature? If the latter, perhaps some documentation can be added to the README noting as much?

It would be nice if it would be possible to create a white-/blacklisting configuration of attributes and models.

Hi, I was previously running both sprig and sprig-reap OK but lately I am getting errors even when trying to do the simplest reap task. The only thing I can think of is that perhaps before I was using a different version of ruby and/or rails. My current version of ruby is 2.2.0 and rails 4.2.0.

Right now if I run rake db:seed:reap it quickly shows "Finished reaping!" but doesn't create any files although there's definitely data in the database. And if I try and specify options and say limit to just my User model I get the following nil error. Other models throw errors as well and it almost feels like the errors occur because I require many of my models to be linked to a user record (i.e. user_id cannot be null).

`rake db:seed:reap TARGET_ENV=development MODELS=User

D, [2015-03-18T06:10:57.471191 #86792] DEBUG -- : Reaping records from the database...
rake aborted!
NoMethodError: undefined method -' for nil:NilClass /Users/dude/.rvm/gems/ruby-2.2.0@tigerwod/gems/sprig-reap-0.2.0/lib/sprig/reap/tsortable_hash.rb:20:inblock (2 levels) in resolve_circular_habtm_dependencies!'`

It would be phantastic if reap would be able to automatically set relations (ie ancestry: '<%= sprig_record(Qbrick::Page, 1).id %>') which are already in other yaml files.

While active records supports association reflections which enable to aggregate information about the relation attributes this could be realistic and would be a really great feature.

EDIT:
Hu? This seems to work sometimes? Is it dependent from the relation?

EDIT 2:
But in my cases the used sprig class was the wrong one. 😿

I already hat a sprig definition with the structure

options:
  find_existing_by: ['identifier']

records:
  - sprig_id: 1
    identifier: lorem
  - sprig_id: 2
    identifier: ipsum

After using sprig-reap the structure was

options:
  find_existing_by: ['identifier']

records:
  - sprig_id: 1
    identifier: lorem
  - sprig_id: 2
    identifier: ipsum

- sprig_id: 3
  identifier: dolor
- sprig_id: 3
  identifier: sit

So there was wrong indention (the duplicated id is already mentioned in another issue)

If there are many empty attributes, the yaml files will be large without any benefit 😉

I am pretty sure that this isn't intended 😉

I have a dummy development app inside which there are bunch of engines with isolated namespaces.
All my models are using this namespace. Let's say "Dummy".

Reap, whilst not reporting any issues, is not generating any files as it doesn't find any models.
If i specifically say something like
rake db:seed:reap MODELS=Dummy::Model
it generates yml just fine.

Is there any way to define a namespace globally?
Or at list to say something like MODELS=Dummy::*

Haven't tried yet Sprig itself, but i suppose i'll have problems there too?

OS X 10.7 - Rails 4.0.3, ruby 2.1.1p76 using PostgreSQL.
My Rails app. uses either ActsAsTaggableOn and Carrierwave. Every user has a Post and an Avatar uploaded via carrierwave on S3. Dumping all the user in a users.yml seed file works fine but the script explodes in my face with:

ArgumentError: undefined class/module Struct::AvatarUploader

The strange thing is that when i run for the first time

bundle exec rake db:seed:reap TARGET_ENV=development MODELS=User,Post IGNORED_ATTRS=created_at,updated_at

I don't get any error. I start receiving the error above from the second time i launch that command onward.

Here is a snatch of the users.yml file:

records:
- sprig_id: 2
  username: user2
  email: [email protected]
  encrypted_password: "$2a$10$GVFZPxh3tu2uO0mNs/D9SuihSzzsAOANZ.1J4MNGu73sUoGAYYQua"
  admin: false
  locked: false
  slug: user2
  reset_password_token: 
  reset_password_sent_at: 
  remember_created_at: 
  sign_in_count: 0
  current_sign_in_at: 
  last_sign_in_at: 
  current_sign_in_ip: 
  last_sign_in_ip: 
  confirmation_token: 
  confirmed_at: &1 2014-06-27 19:35:14.399730000 Z
  confirmation_sent_at: 
  unconfirmed_email: 
  avatar: &2 !ruby/object:AvatarUploader
    model: !ruby/object:User
      attributes:
        id: 2
        username: user2
        email: [email protected]
        encrypted_password: "$2a$10$GVFZPxh3tu2uO0mNs/D9SuihSzzsAOANZ.1J4MNGu73sUoGAYYQua"
        admin: false
        locked: false
        slug: user2
        reset_password_token: 
        reset_password_sent_at: 
        remember_created_at: 
        sign_in_count: 0
        current_sign_in_at: 
        last_sign_in_at: 
        current_sign_in_ip: 
        last_sign_in_ip: 
        confirmation_token: 
        confirmed_at: *1
        confirmation_sent_at: 
        unconfirmed_email: 
        avatar: u2.jpg
        created_at: &3 2014-06-27 19:35:14.410751000 Z
        updated_at: &4 2014-06-27 19:35:14.410751000 Z
        on_private_message: '1'
        on_post_voted: '1'
        on_post_commented: '1'
        on_post_linked: '1'
        on_post_smart_filtered: '1'
        incomplete_reminder_timing: '1'
        digest_timing: '1'
        job_title: Corporate Metrics Manager
        summary: repurpose bleeding-edge channels
    mounted_as: :avatar
    storage: !ruby/object:CarrierWave::Storage::File
      uploader: *2
    file: !ruby/object:CarrierWave::SanitizedFile
      file: "...[edited].../public/uploads/user/avatar/2/u2.jpg"
      original_filename: 
      content_type: 
    versions:
      :thumb: &5 !ruby/object:AvatarUploader::Uploader70229152586100
        model: !ruby/object:User
          attributes:
            id: 2
            username: user2
            email: [email protected]
            encrypted_password: "$2a$10$GVFZPxh3tu2uO0mNs/D9SuihSzzsAOANZ.1J4MNGu73sUoGAYYQua"
            admin: false
            locked: false
            slug: user2
            reset_password_token: 
            reset_password_sent_at: 
            remember_created_at: 
            sign_in_count: 0
            current_sign_in_at: 
            last_sign_in_at: 
            current_sign_in_ip: 
            last_sign_in_ip: 
            confirmation_token: 
            confirmed_at: *1
            confirmation_sent_at: 
            unconfirmed_email: 
            avatar: u2.jpg
            created_at: *3
            updated_at: *4
            on_private_message: '1'
            on_post_voted: '1'
            on_post_commented: '1'
            on_post_linked: '1'
            on_post_smart_filtered: '1'
            incomplete_reminder_timing: '1'
            digest_timing: '1'
            job_title: Corporate Metrics Manager
            summary: repurpose bleeding-edge channels
        mounted_as: :avatar
        storage: !ruby/object:CarrierWave::Storage::File
          uploader: *5
        file: !ruby/object:CarrierWave::SanitizedFile
          file: "...[edited].../public/uploads/user/avatar/2/thumb_u2.jpg"
          original_filename: 
          content_type: 
        versions: {}
  on_private_message: '1'
  on_post_voted: '1'
  on_post_commented: '1'
  on_post_linked: '1'
  on_post_smart_filtered: '1'
  incomplete_reminder_timing: '1'
  digest_timing: '1'
  job_title: Corporate Metrics Manager
  summary: repurpose bleeding-edge channels

Any clue, hint help is appreciated.

Working on #16 I found an issue using ruby 2.1.2. Yaml files are parsed differently from 1.9.x so loading some fixtures has quoting issues.