seed_dump 3.3.1 → 3.4.1

data/lib/seed_dump/environment.rb

@@ -6,9 +6,20 @@ class SeedDump
 
       models = retrieve_models(env) - retrieve_models_exclude(env)
 
-      limit = retrieve_limit_value(env)
+      # Sort models by foreign key dependencies (issues #78, #83)
+      # This ensures models are dumped in the correct order so that
+      # seeds can be imported without foreign key violations.
+      models = sort_models_by_dependencies(models)
+
+      global_limit = retrieve_limit_value(env)
+      model_limits = retrieve_model_limits_value(env)
       append = retrieve_append_value(env)
       models.each do |model|
+        # Determine the limit to apply for this model:
+        # 1. Check MODEL_LIMITS for a per-model override
+        # 2. Fall back to global LIMIT
+        # 3. If neither, no limit is applied
+        limit = limit_for_model(model, model_limits, global_limit)
         model = model.limit(limit) if limit.present?
 
         SeedDump.dump(model,
@@ -16,7 +27,11 @@ class SeedDump
                       batch_size: retrieve_batch_size_value(env),
                       exclude: retrieve_exclude_value(env),
                       file: retrieve_file_value(env),
-                      import: retrieve_import_value(env))
+                      group_sti_by_class: retrieve_group_sti_by_class_value(env),
+                      header: retrieve_header_value(env),
+                      import: retrieve_import_value(env),
+                      insert_all: retrieve_insert_all_value(env),
+                      upsert_all: retrieve_upsert_all_value(env))
 
         append = true # Always append for every model after the first
                       # (append for the first model is determined by
@@ -60,7 +75,7 @@ class SeedDump
       # model classes in the project.
       models = if models_env
                  models_env.split(',')
-                           .collect {|x| x.strip.underscore.singularize.camelize.constantize }
+                           .collect {|x| model_name_to_constant(x.strip) }
                else
                  ActiveRecord::Base.descendants
                end
@@ -69,13 +84,171 @@ class SeedDump
       # Filter the set of models to exclude:
       #   - The ActiveRecord::SchemaMigration model which is internal to Rails
       #     and should not be part of the dumped data.
+      #   - Classes that don't respond to table_exists? or exists? (e.g., abstract
+      #     classes or non-model descendants of ActiveRecord::Base).
       #   - Models that don't have a corresponding table in the database.
       #   - Models whose corresponding database tables are empty.
       filtered_models = models.select do |model|
-                          !ACTIVE_RECORD_INTERNAL_MODELS.include?(model.to_s) && \
-                          model.table_exists? && \
-                          model.exists?
+                          begin
+                            !ACTIVE_RECORD_INTERNAL_MODELS.include?(model.to_s) && \
+                            model.table_exists? && \
+                            model.exists?
+                          rescue NoMethodError
+                            # Skip classes that don't properly respond to table_exists? or exists?
+                            # This can happen with abstract classes or other non-model descendants
+                            false
+                          end
                         end
+
+      # Deduplicate HABTM models that share the same table (issues #26, #114).
+      # Rails creates two auto-generated models for each HABTM association
+      # (e.g., User::HABTM_Roles and Role::HABTM_Users) that both point to
+      # the same join table. We only want to dump one of them.
+      deduped_habtm = deduplicate_habtm_models(filtered_models)
+
+      # Deduplicate STI models that share the same table (issue #120).
+      # With STI, subclasses (e.g., AdminUser < User) share the same table as
+      # their parent. We only want to dump the base class, which will include
+      # all records including subclass records with proper type discrimination.
+      deduplicate_sti_models(deduped_habtm)
+    end
+
+    # Internal: Deduplicates HABTM models that share the same table.
+    #
+    # When using has_and_belongs_to_many, Rails creates auto-generated models
+    # like User::HABTM_Roles and Role::HABTM_Users that both reference the same
+    # join table. Without deduplication, the join table data would be dumped twice.
+    #
+    # models - Array of ActiveRecord model classes.
+    #
+    # Returns the Array with duplicate HABTM models removed.
+    def deduplicate_habtm_models(models)
+      habtm, non_habtm = models.partition { |m| m.to_s.include?('HABTM_') }
+      non_habtm + habtm.uniq(&:table_name)
+    end
+
+    # Internal: Deduplicates STI models that share the same table.
+    #
+    # With Single Table Inheritance, subclasses like AdminUser < User share the
+    # same database table as their parent. Without deduplication, each STI class
+    # would be dumped separately, creating duplicate records.
+    #
+    # The solution is to keep only the base class for each STI hierarchy, which
+    # will include all records (base and subclass) with proper type discrimination.
+    #
+    # models - Array of ActiveRecord model classes.
+    #
+    # Returns the Array with STI subclasses removed (only base classes kept).
+    def deduplicate_sti_models(models)
+      models.select do |model|
+        # Keep the model only if it IS its own base class
+        # For STI subclasses, base_class returns the parent (e.g., AdminUser.base_class => User)
+        # For non-STI models, base_class returns self
+        model.base_class == model
+      end
+    end
+
+    # Internal: Sorts models by foreign key dependencies using topological sort.
+    #
+    # Models with foreign keys (belongs_to associations) depend on the models
+    # they reference. This method ensures that referenced models are dumped
+    # before the models that depend on them, preventing foreign key violations
+    # when importing seeds (issues #78, #83).
+    #
+    # For example, if Book belongs_to Author, Author will be sorted before Book.
+    #
+    # Uses Kahn's algorithm for topological sorting. If there are circular
+    # dependencies, the remaining models are appended in their original order.
+    #
+    # models - Array of ActiveRecord model classes to sort.
+    #
+    # Returns a new Array with models sorted by dependencies (dependencies first).
+    def sort_models_by_dependencies(models)
+      return models if models.empty?
+
+      # Build a lookup for models by table name for faster dependency resolution
+      model_by_table = models.each_with_object({}) do |model, hash|
+        hash[model.table_name] = model
+      end
+
+      # Build dependency graph: model -> models it depends on (via belongs_to)
+      dependencies = {}
+      models.each do |model|
+        dependencies[model] = find_model_dependencies(model, model_by_table)
+      end
+
+      # Topological sort using Kahn's algorithm
+      topological_sort(models, dependencies)
+    end
+
+    # Internal: Finds the models that a given model depends on via belongs_to.
+    #
+    # model - The ActiveRecord model class to find dependencies for.
+    # model_by_table - Hash mapping table names to model classes.
+    #
+    # Returns an Array of model classes that this model depends on.
+    def find_model_dependencies(model, model_by_table)
+      deps = []
+
+      # Check belongs_to associations for foreign key dependencies
+      model.reflect_on_all_associations(:belongs_to).each do |assoc|
+        # Get the table name this association points to
+        # Use the association's class_name if available, otherwise infer from name
+        begin
+          referenced_class = assoc.klass
+          referenced_table = referenced_class.table_name
+
+          # Only add as dependency if it's in our set of models to dump
+          if model_by_table.key?(referenced_table)
+            dep_model = model_by_table[referenced_table]
+            deps << dep_model unless dep_model == model
+          end
+        rescue NameError, ArgumentError
+          # Skip if we can't resolve the class (e.g., polymorphic without type)
+          next
+        end
+      end
+
+      deps.uniq
+    end
+
+    # Internal: Performs topological sort on models based on their dependencies.
+    #
+    # Uses Kahn's algorithm:
+    # 1. Find all models with no dependencies (no incoming edges)
+    # 2. Add them to the result and remove them from the graph
+    # 3. Repeat until all models are sorted or a cycle is detected
+    #
+    # models - Array of model classes.
+    # dependencies - Hash mapping each model to its dependencies.
+    #
+    # Returns an Array of models in topologically sorted order.
+    def topological_sort(models, dependencies)
+      result = []
+      remaining = models.dup
+
+      # Calculate in-degree (number of models depending on each model)
+      # We need to track which models are "ready" (all their dependencies satisfied)
+      while remaining.any?
+        # Find models whose dependencies have all been processed
+        ready = remaining.select do |model|
+          dependencies[model].all? { |dep| result.include?(dep) }
+        end
+
+        if ready.empty?
+          # Circular dependency detected - add remaining in original order
+          result.concat(remaining)
+          break
+        end
+
+        # Add ready models to result (maintain relative order for stability)
+        ready.each do |model|
+          result << model
+          remaining.delete(model)
+        end
+      end
+
+      result
     end
 
     # Internal: Returns a Boolean indicating whether the value for the "APPEND"
@@ -92,13 +265,74 @@ class SeedDump
       parse_boolean_value(env['IMPORT'])
     end
 
+    # Internal: Returns a Boolean indicating whether the value for the "INSERT_ALL"
+    # key in the given Hash is equal to the String "true" (ignoring case),
+    # false if no value exists. INSERT_ALL uses Rails 6+ insert_all for faster
+    # bulk inserts that bypass validations and callbacks.
+    def retrieve_insert_all_value(env)
+      parse_boolean_value(env['INSERT_ALL'])
+    end
+
+    # Internal: Returns a Boolean indicating whether the value for the "UPSERT_ALL"
+    # key in the given Hash is equal to the String "true" (ignoring case),
+    # false if no value exists. UPSERT_ALL uses Rails 6+ upsert_all to preserve
+    # original record IDs, which fixes foreign key reference issues when parent
+    # records have been deleted (issue #104).
+    def retrieve_upsert_all_value(env)
+      parse_boolean_value(env['UPSERT_ALL'])
+    end
+
+    # Internal: Returns a Boolean indicating whether the value for the "HEADER"
+    # key in the given Hash is equal to the String "true" (ignoring case),
+    # false if no value exists. HEADER adds a comment at the top of the seed file
+    # showing when and how it was generated for traceability (issue #126).
+    def retrieve_header_value(env)
+      parse_boolean_value(env['HEADER'])
+    end
+
+    # Internal: Returns a Boolean indicating whether the value for the "GROUP_STI_BY_CLASS"
+    # key in the given Hash is equal to the String "true" (ignoring case),
+    # false if no value exists. GROUP_STI_BY_CLASS groups STI records by their actual
+    # class instead of base_class to fix enum issues (issue #170).
+    def retrieve_group_sti_by_class_value(env)
+      parse_boolean_value(env['GROUP_STI_BY_CLASS'])
+    end
+
     # Internal: Retrieves an Array of Class constants parsed from the value for
     # the "MODELS_EXCLUDE" key in the given Hash, and an empty Array if such
     # key exists.
     def retrieve_models_exclude(env)
       env['MODELS_EXCLUDE'].to_s
                            .split(',')
-                           .collect { |x| x.strip.underscore.singularize.camelize.constantize }
+                           .collect { |x| model_name_to_constant(x.strip) }
+    end
+
+    # Internal: Converts a model name string to a constant.
+    #
+    # This method handles the issue where model names ending in 's' (like "Boss")
+    # were incorrectly singularized to "Bos" by older Rails versions (issue #121).
+    #
+    # The strategy is:
+    # 1. Try camelized form first (handles "Boss", "boss", "user_profile")
+    # 2. Fall back to underscore.singularize.camelize for plural table names
+    #
+    # model_name - String name of the model (e.g., "Boss", "boss", "users")
+    #
+    # Returns the Class constant for the model.
+    # Raises NameError if the model cannot be found.
+    def model_name_to_constant(model_name)
+      # First, try the camelized version directly
+      # This handles: "Boss" -> Boss, "boss" -> Boss, "user_profile" -> UserProfile
+      camelized = model_name.camelize
+      begin
+        return camelized.constantize
+      rescue NameError
+        # Fall through to try singularized version
+      end
+
+      # Fall back to traditional approach for plural names
+      # This handles: "users" -> User, "bosses" -> Boss
+      model_name.underscore.singularize.camelize.constantize
     end
 
     # Internal: Retrieves an Integer from the value for the "LIMIT" key in the
@@ -107,10 +341,71 @@ class SeedDump
       retrieve_integer_value('LIMIT', env)
     end
 
+    # Internal: Parses the MODEL_LIMITS environment variable into a Hash.
+    #
+    # MODEL_LIMITS allows per-model limit overrides to prevent LIMIT from
+    # breaking associations (issue #142). Format: "Model1:limit1,Model2:limit2"
+    #
+    # A limit of 0 means "unlimited" (dump all records for that model).
+    #
+    # Example: MODEL_LIMITS="Teacher:0,Student:50"
+    #   - Teacher: dumps all records (0 = unlimited)
+    #   - Student: dumps 50 records
+    #   - Other models: fall back to global LIMIT or dump all if no LIMIT set
+    #
+    # env - Hash of environment variables.
+    #
+    # Returns a Hash mapping model names (String) to limits (Integer), or
+    # empty Hash if MODEL_LIMITS is not set.
+    def retrieve_model_limits_value(env)
+      return {} unless env['MODEL_LIMITS']
+
+      env['MODEL_LIMITS'].split(',').each_with_object({}) do |pair, hash|
+        model_name, limit = pair.split(':').map(&:strip)
+        hash[model_name] = limit.to_i if model_name && limit
+      end
+    end
+
+    # Internal: Determines the limit to apply for a given model.
+    #
+    # Precedence:
+    # 1. Per-model limit from MODEL_LIMITS (0 means unlimited)
+    # 2. Global LIMIT
+    # 3. nil (no limit, dump all records)
+    #
+    # model - The ActiveRecord model class.
+    # model_limits - Hash of per-model limits from MODEL_LIMITS.
+    # global_limit - The global LIMIT value (Integer or nil).
+    #
+    # Returns an Integer limit or nil if no limit should be applied.
+    def limit_for_model(model, model_limits, global_limit)
+      model_name = model.to_s
+
+      if model_limits.key?(model_name)
+        limit = model_limits[model_name]
+        # 0 means unlimited - return nil to skip applying limit
+        limit == 0 ? nil : limit
+      else
+        global_limit
+      end
+    end
+
     # Internal: Retrieves an Array of Symbols from the value for the "EXCLUDE"
     # key from the given Hash, and nil if no such key exists.
+    #
+    # If INCLUDE_ALL is set to 'true', returns an empty array to disable
+    # the default exclusion of id, created_at, updated_at columns. This provides
+    # a cleaner alternative to EXCLUDE="" (issue #147).
+    #
+    # Note that explicit EXCLUDE values take precedence over INCLUDE_ALL.
     def retrieve_exclude_value(env)
-      env['EXCLUDE'] ? env['EXCLUDE'].split(',').map {|e| e.strip.to_sym} : nil
+      if env['EXCLUDE']
+        env['EXCLUDE'].split(',').map { |e| e.strip.to_sym }
+      elsif parse_boolean_value(env['INCLUDE_ALL'])
+        []
+      else
+        nil
+      end
     end
 
     # Internal: Retrieves the value for the "FILE" key from the given Hash, and

metadata

@@ -1,18 +1,18 @@
 --- !ruby/object:Gem::Specification
 name: seed_dump
 version: !ruby/object:Gem::Version
-  version: 3.3.1
+  version: 3.4.1
 platform: ruby
 authors:
 - Rob Halff
 - Ryan Oblak
-autorequire: 
+autorequire:
 bindir: bin
 cert_chain: []
-date: 2018-05-08 00:00:00.000000000 Z
+date: 2025-12-24 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
-  name: activesupport
+  name: activerecord
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - ">="
@@ -26,7 +26,7 @@ dependencies:
       - !ruby/object:Gem::Version
         version: '4'
 - !ruby/object:Gem::Dependency
-  name: activerecord
+  name: activesupport
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - ">="
@@ -45,44 +45,58 @@ dependencies:
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '2.0'
+        version: '11.1'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '2.0'
+        version: '11.1'
 - !ruby/object:Gem::Dependency
   name: factory_bot
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: 4.8.2
+        version: '6.1'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: 4.8.2
+        version: '6.1'
 - !ruby/object:Gem::Dependency
   name: activerecord-import
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '0.4'
+        version: '0.28'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '0.28'
+- !ruby/object:Gem::Dependency
+  name: rspec
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '3.13'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '0.4'
+        version: '3.13'
 - !ruby/object:Gem::Dependency
-  name: jeweler
+  name: database_cleaner-active_record
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
@@ -95,39 +109,124 @@ dependencies:
     - - "~>"
       - !ruby/object:Gem::Version
         version: '2.0'
+- !ruby/object:Gem::Dependency
+  name: appraisal
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '2.4'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '2.4'
+- !ruby/object:Gem::Dependency
+  name: rake
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+- !ruby/object:Gem::Dependency
+  name: sqlite3
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '1.3'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '1.3'
+- !ruby/object:Gem::Dependency
+  name: mutex_m
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+- !ruby/object:Gem::Dependency
+  name: logger
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+- !ruby/object:Gem::Dependency
+  name: benchmark
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+- !ruby/object:Gem::Dependency
+  name: base64
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
 description: Dump (parts) of your database to db/seeds.rb to get a headstart creating
   a meaningful seeds.rb file
 email: rroblak@gmail.com
 executables: []
 extensions: []
-extra_rdoc_files:
-- README.md
+extra_rdoc_files: []
 files:
-- ".rspec"
-- Gemfile
 - MIT-LICENSE
 - README.md
-- Rakefile
-- VERSION
 - lib/seed_dump.rb
 - lib/seed_dump/dump_methods.rb
 - lib/seed_dump/dump_methods/enumeration.rb
 - lib/seed_dump/environment.rb
 - lib/seed_dump/railtie.rb
 - lib/tasks/seed_dump.rake
-- seed_dump.gemspec
-- spec/dump_methods_spec.rb
-- spec/environment_spec.rb
-- spec/factories/another_samples.rb
-- spec/factories/samples.rb
-- spec/factories/yet_another_samples.rb
-- spec/helpers.rb
-- spec/spec_helper.rb
 homepage: https://github.com/rroblak/seed_dump
 licenses:
 - MIT
 metadata: {}
-post_install_message: 
+post_install_message:
 rdoc_options: []
 require_paths:
 - lib
@@ -142,9 +241,8 @@ required_rubygems_version: !ruby/object:Gem::Requirement
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubyforge_project: 
-rubygems_version: 2.7.6
-signing_key: 
+rubygems_version: 3.5.22
+signing_key:
 specification_version: 4
-summary: "{Seed Dumper for Rails}"
+summary: Seed Dumper for Rails
 test_files: []

data/.rspec

@@ -1,2 +0,0 @@
---color
---format progress

data/Gemfile

@@ -1,20 +0,0 @@
-source 'https://rubygems.org'
-
-gem 'activesupport', '>= 4'
-gem 'activerecord', '>= 4'
-
-group :development, :test do
-  gem 'byebug', '~> 2.0'
-  gem 'factory_bot', '~> 4.8.2'
-  gem 'activerecord-import', '~> 0.4'
-end
-
-group :development do
-  gem 'jeweler', '~> 2.0'
-end
-
-group :test do
-  gem 'rspec', '~> 3.7.0'
-  gem 'sqlite3', '~> 1.0'
-  gem 'database_cleaner', '~> 1.0'
-end

data/Rakefile

@@ -1,32 +0,0 @@
-require 'rubygems'
-require 'rake'
-
-begin
-  require 'jeweler'
-  Jeweler::Tasks.new do |gem|
-    gem.name = "seed_dump"
-    gem.summary = "{Seed Dumper for Rails}"
-    gem.description = %Q{Dump (parts) of your database to db/seeds.rb to get a headstart creating a meaningful seeds.rb file}
-    gem.email = 'rroblak@gmail.com'
-    gem.homepage = 'https://github.com/rroblak/seed_dump'
-    gem.authors = ['Rob Halff', 'Ryan Oblak']
-    gem.license = 'MIT'
-  end
-  Jeweler::GemcutterTasks.new
-rescue LoadError
-  puts "Jeweler (or a dependency) not available. Install it with: gem install jeweler"
-end
-
-require 'rdoc/task'
-Rake::RDocTask.new do |rdoc|
-  version = File.exist?('VERSION') ? File.read('VERSION') : ""
-
-  rdoc.rdoc_dir = 'rdoc'
-  rdoc.title = "seed_dump #{version}"
-  rdoc.rdoc_files.include('README*')
-  rdoc.rdoc_files.include('lib/**/*.rb')
-end
-
-require 'rspec/core/rake_task'
-RSpec::Core::RakeTask.new(:spec)
-task :default => :spec

data/VERSION

@@ -1 +0,0 @@
-3.3.1