rocketjob 6.0.0.rc3 → 6.0.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 6a04a33b0cd03bdf0a7cb948fc87dd6c7d7bb3b392e566a8c15df50b73e27459
-  data.tar.gz: fc62e740a0a92bae8daf1f4ffbe199af1debcb84f8859aed10ea5954dc44c7b6
+  metadata.gz: e313f192b854d066258a614ceac1131851c8df94c7e08f7cea6681fff6946d69
+  data.tar.gz: 10804682bee08715671696db4610ce4f93679398398bc385e93619f5a3aca715
 SHA512:
-  metadata.gz: 74cac01d253cf21a856e1ca4a5cf63d5e90320303bdf310cf90325c9cca242c4ed1b7a0a1c43ca00764f2f40d29822df6e6bee499c1bff56c9ddaa2401bc3862
-  data.tar.gz: 1bbc47c7d869ef28fd578a7b2575f62957aa2f83f9fc927af1d6fba7866270b15fd21cef30007b78d58847137357c75ccae2d03545560d2ffe0d674fe34c1d0e
+  metadata.gz: 158675e5ddec87a8b277708887b037746e3f1573569edd5a8959eabdf5668b144553cbe6164844f2cf69bc603798b3b5b052697506dad8a29e8477afc62cc45f
+  data.tar.gz: 680efe5603de3649b7e09340a545d2e1df1e02697d34af9451869310e9f2a87bbd05423686bb7d40d60f104553f7311721870e52ac2171bb1491b3a8decaf439

data/README.md

@@ -49,6 +49,32 @@ require "rocket_job/batch/tabular"
 
 It is important to migrate away from these plugins, since they will be removed in a future release.
 
+#### Scheduled Jobs
+
+For any scheduled jobs that include the `RocketJob::Plugins::Cron` plugin, the default behavior has changed
+so that the scheduled job instance is created immediately after the currently scheduled instance starts.
+
+To maintain the old behavior of creating the job when it fails, aborts, or completes, add the following line
+to each of the applicable jobs:
+
+~~~ruby
+self.cron_after_start = false
+~~~
+
+Additionally, scheduled jobs will now prevent a new one from being created when another scheduled instance 
+of the same job is already queued, or running with the _same_ `cron_schedule`.
+
+To maintain the old behavior of allowing multiple instances with the same cron schedule, add the following
+line to each of the applicable jobs:
+
+~~~ruby
+self.cron_singleton = false
+~~~
+
+##### Singleton
+
+Since Scheduled jobs now implement their own singleton logic, remove the singleton plugin from any scheduled jobs.
+
 #### Upgrading Batch Jobs to Rocket Job v6
 
 Rocket Job v6 replaces the array of symbol type for `input_categories` and `output_categories`

data/lib/rocket_job/batch/categories.rb

@@ -72,34 +72,38 @@ module RocketJob
       end
 
       def input_category(category_name = :main)
+        return category_name if category_name.is_a?(Category::Input)
+        raise(ArgumentError, "Cannot supply Output Category to input category") if category_name.is_a?(Category::Output)
+
         category_name = category_name.to_sym
-        category      = nil
-        # .find does not work against this association
-        input_categories.each { |catg| category = catg if catg.name == category_name }
-        unless category
-          # Auto-register main input category if missing
-          if category_name == :main
-            category              = Category::Input.new
-            self.input_categories = [category]
-          else
-            raise(ArgumentError,
-                  "Unknown Input Category: #{category_name.inspect}. Registered categories: #{input_categories.collect(&:name).join(',')}")
-          end
+        # find does not work against this association
+        input_categories.each { |category| return category if category.name == category_name }
+
+        unless category_name == :main
+          raise(
+            ArgumentError,
+            "Unknown Input Category: #{category_name.inspect}. Registered categories: #{input_categories.collect(&:name).join(',')}"
+          )
         end
+
+        # Auto-register main input category when not defined
+        category = Category::Input.new(job: self)
+        self.input_categories << category
         category
       end
 
       def output_category(category_name = :main)
+        return category_name if category_name.is_a?(Category::Output)
+        raise(ArgumentError, "Cannot supply Input Category to output category") if category_name.is_a?(Category::Input)
+
         category_name = category_name.to_sym
-        category      = nil
         # .find does not work against this association
-        output_categories.each { |catg| category = catg if catg.name == category_name }
-        unless category
-          raise(ArgumentError,
-                "Unknown Output Category: #{category_name.inspect}. Registered categories: #{output_categories.collect(&:name).join(',')}")
-        end
+        output_categories.each { |category| return category if category.name == category_name }
 
-        category
+        raise(
+          ArgumentError,
+          "Unknown Output Category: #{category_name.inspect}. Registered categories: #{output_categories.collect(&:name).join(',')}"
+        )
       end
 
       # Returns [true|false] whether the named category has already been defined
@@ -214,7 +218,7 @@ module RocketJob
         category.tabular.render(row)
       end
 
-      # Migrate existing v4 batch jobs to v5.0
+      # Migrate existing v5 batch jobs to v6
       def rocketjob_categories_migrate
         return unless attribute_present?(:input_categories) && self[:input_categories]&.first.is_a?(Symbol)
 

data/lib/rocket_job/batch/io.rb

@@ -14,11 +14,9 @@ module RocketJob
       #     Default: None ( Uses the single default input collection for this job )
       #     Validates: This value must be one of those listed in #input_categories
       def input(category = :main)
-        raise(ArgumentError, "Cannot supply Output Category to input category") if category.is_a?(Category::Output)
+        category = input_category(category)
 
-        category = input_category(category) unless category.is_a?(Category::Input)
-
-        (@inputs ||= {})[category.name] ||= RocketJob::Sliced.factory(:input, category, self)
+        (@inputs ||= {})[category.name] ||= category.data_store(self)
       end
 
       # Returns [RocketJob::Sliced::Output] output collection for holding output slices
@@ -30,11 +28,9 @@ module RocketJob
       #     Default: None ( Uses the single default output collection for this job )
       #     Validates: This value must be one of those listed in #output_categories
       def output(category = :main)
-        raise(ArgumentError, "Cannot supply Input Category to output category") if category.is_a?(Category::Input)
-
-        category = output_category(category) unless category.is_a?(Category::Output)
+        category = output_category(category)
 
-        (@outputs ||= {})[category.name] ||= RocketJob::Sliced.factory(:output, category, self)
+        (@outputs ||= {})[category.name] ||= category.data_store(self)
       end
 
       # Rapidly upload individual records in batches.
@@ -59,19 +55,19 @@ module RocketJob
       #     The category or the name of the category to access or download data from
       #     Default: None ( Uses the single default output collection for this job )
       #     Validates: This value must be one of those listed in #input_categories
-      def lookup_collection(category = :main)
-        category = input_category(category) unless category.is_a?(Category::Input)
-
-        collection = (@lookup_collections ||= {})[category.name]
-
-        unless collection
-          collection_name = "rocket_job.inputs.#{id}"
-          collection_name << ".#{category.name}" unless category.name == :main
-
-          @lookup_collections[category.name] ||=
-            LookupCollection.new(Sliced::Slice.collection.database, collection_name)
-        end
-      end
+      # def lookup_collection(category = :main)
+      #   category = input_category(category) unless category.is_a?(Category::Input)
+      #
+      #   collection = (@lookup_collections ||= {})[category.name]
+      #
+      #   unless collection
+      #     collection_name = "rocket_job.inputs.#{id}"
+      #     collection_name << ".#{category.name}" unless category.name == :main
+      #
+      #     @lookup_collections[category.name] ||=
+      #       LookupCollection.new(Sliced::Slice.collection.database, collection_name)
+      #   end
+      # end
 
       # Upload the supplied file, io, IOStreams::Path, or IOStreams::Stream.
       #
@@ -154,53 +150,7 @@ module RocketJob
       # * If an io stream is supplied, it is read until it returns nil.
       # * Only use this method for UTF-8 data, for binary data use #input_slice or #input_records.
       # * CSV parsing is slow, so it is usually left for the workers to do.
-      def upload(stream = nil, file_name: nil, category: :main, stream_mode: :line, on_first: nil, **args, &block)
-        raise(ArgumentError, "Either stream, or a block must be supplied") unless stream || block
-
-        category = input_category(category) unless category.is_a?(Category::Input)
-        stream ||= category.file_name
-        path     = nil
-
-        if stream
-          path               = IOStreams.new(stream)
-          path.file_name     = file_name if file_name
-          category.file_name = path.file_name
-
-          # Auto detect the format based on the upload file name if present.
-          if category.format == :auto
-            format = path.format
-            if format
-              # Rebuild tabular with the above file name
-              category.reset_tabular
-              category.format = format
-            end
-          end
-        end
-
-        # Tabular transformations required for upload?
-        if category.tabular?
-          # Remove non-printable characters from tabular input formats
-          # Cannot change the length of fixed width lines
-          replace = category.format == :fixed ? " " : ""
-          path&.option_or_stream(:encode, encoding: "UTF-8", cleaner: :printable, replace: replace)
-
-          # Extract the header line during the file upload when needed.
-          on_first = rocket_job_upload_header_lambda(category, on_first) if category.tabular.header?
-        end
-
-        count =
-          if block
-            input(category).upload(on_first: on_first, &block)
-          else
-            input(category).upload(on_first: on_first) do |io|
-              path.each(stream_mode, **args) { |line| io << line }
-            end
-          end
-
-        self.record_count = (record_count || 0) + count
-        count
-      end
-
+      #
       # Upload results from an Arel into RocketJob::SlicedJob.
       #
       # Params
@@ -227,18 +177,13 @@ module RocketJob
       #
       # Example: Upload user_name and zip_code
       #   arel = User.where(country_code: 'US')
-      #   job.upload_arel(arel, :user_name, :zip_code)
+      #   job.upload_arel(arel, columns: [:user_name, :zip_code])
       #
       # Notes:
       # * Only call from one thread at a time against a single instance of this job.
       # * The record_count for the job is set to the number of records returned by the arel.
       # * If an exception is raised while uploading data, the input collection is cleared out
       #   so that if a job is retried during an upload failure, data is not duplicated.
-      def upload_arel(arel, *column_names, category: :main, &block)
-        count             = input(category).upload_arel(arel, *column_names, &block)
-        self.record_count = (record_count || 0) + count
-        count
-      end
 
       # Upload the result of a MongoDB query to the input collection for processing
       # Useful when an entire MongoDB collection, or part thereof needs to be
@@ -266,24 +211,19 @@ module RocketJob
       #   criteria = User.where(state: 'FL')
       #   job.record_count = job.upload_mongo_query(criteria)
       #
-      # Example: Upload just the supplied column
+      # Example: Upload only the specified column(s)
       #   criteria = User.where(state: 'FL')
-      #   job.record_count = job.upload_mongo_query(criteria, :zip_code)
+      #   job.record_count = job.upload_mongo_query(criteria, columns: [:zip_code])
       #
       # Notes:
       # * Only call from one thread at a time against a single instance of this job.
       # * The record_count for the job is set to the number of records returned by the monqo query.
       # * If an exception is raised while uploading data, the input collection is cleared out
       #   so that if a job is retried during an upload failure, data is not duplicated.
-      def upload_mongo_query(criteria, *column_names, category: :main, &block)
-        count             = input(category).upload_mongo_query(criteria, *column_names, &block)
-        self.record_count = (record_count || 0) + count
-        count
-      end
 
       # Upload sliced range of integer requests as arrays of start and end ids.
       #
-      # Returns [Integer] last_id - start_id + 1.
+      # Returns [Integer] the number of slices uploaded.
       #
       # Uploads one range per slice so that the response can return multiple records
       # for each slice processed
@@ -302,17 +242,11 @@ module RocketJob
       # * The record_count for the job is set to: last_id - start_id + 1.
       # * If an exception is raised while uploading data, the input collection is cleared out
       #   so that if a job is retried during an upload failure, data is not duplicated.
-      def upload_integer_range(start_id, last_id, category: :main)
-        input(category).upload_integer_range(start_id, last_id)
-        count             = last_id - start_id + 1
-        self.record_count = (record_count || 0) + count
-        count
-      end
 
       # Upload sliced range of integer requests as an arrays of start and end ids
       # starting with the last range first
       #
-      # Returns [Integer] last_id - start_id + 1.
+      # Returns [Integer] the number of slices uploaded.
       #
       # Uploads one range per slice so that the response can return multiple records
       # for each slice processed.
@@ -334,14 +268,102 @@ module RocketJob
       # * The record_count for the job is set to: last_id - start_id + 1.
       # * If an exception is raised while uploading data, the input collection is cleared out
       #   so that if a job is retried during an upload failure, data is not duplicated.
-      def upload_integer_range_in_reverse_order(start_id, last_id, category: :main)
-        input(category).upload_integer_range_in_reverse_order(start_id, last_id)
-        count             = last_id - start_id + 1
+
+      def upload(object = nil, category: :main, file_name: nil, stream_mode: nil, on_first: nil, columns: nil, slice_batch_size: nil, **args, &block)
+        input_collection = input(category)
+
+        if block
+          raise(ArgumentError, "Cannot supply both an object to upload, and a block.") if object
+          if stream_mode || columns || slice_batch_size || args.size > 0
+            raise(ArgumentError, "Unknown keyword arguments when uploading a block. Only accepts :category, :file_name, or :on_first")
+          end
+
+          category           = input_category(category)
+          category.file_name = file_name if file_name
+
+          # Extract the header line during the upload when applicable.
+          extract_header = category.extract_header_callback(on_first)
+
+          count             = input_collection.upload(on_first: extract_header, slice_batch_size: slice_batch_size, &block)
+          self.record_count = (record_count || 0) + count
+          return count
+        end
+
+        count =
+          case object
+          when Range
+            if file_name || stream_mode || on_first || args.size > 0
+              raise(ArgumentError, "Unknown keyword arguments when uploading a Range. Only accepts :category, :columns, or :slice_batch_size")
+            end
+
+            first = object.first
+            last  = object.last
+            if first < last
+              input_collection.upload_integer_range(first, last, slice_batch_size: slice_batch_size || 1_000)
+            else
+              input_collection.upload_integer_range_in_reverse_order(last, first, slice_batch_size: slice_batch_size || 1_000)
+            end
+          when Mongoid::Criteria
+            if file_name || stream_mode || on_first || args.size > 0
+              raise(ArgumentError, "Unknown keyword arguments when uploading a Mongoid::Criteria. Only accepts :category, :columns, or :slice_batch_size")
+            end
+
+            input_collection.upload_mongo_query(object, columns: columns, slice_batch_size: slice_batch_size, &block)
+          when defined?(ActiveRecord::Relation) ? ActiveRecord::Relation : false
+            if file_name || stream_mode || on_first || args.size > 0
+              raise(ArgumentError, "Unknown keyword arguments when uploading an ActiveRecord::Relation. Only accepts :category, :columns, or :slice_batch_size")
+            end
+
+            input_collection.upload_arel(object, columns: columns, slice_batch_size: slice_batch_size, &block)
+
+          else
+            raise(ArgumentError, "Unknown keyword argument :columns when uploading a file") if columns
+
+            category = input_category(category)
+
+            # Extract the header line during the upload when applicable.
+            extract_header = category.extract_header_callback(on_first)
+            path = category.upload_path(object, original_file_name: file_name)
+
+            input_collection.upload(on_first: extract_header, slice_batch_size: slice_batch_size) do |io|
+              path.each(stream_mode || :line, **args) { |line| io << line }
+            end
+
+          end
+
+        self.record_count = (record_count || 0) + count
+        count
+      end
+
+      # @deprecated
+      def upload_arel(arel, *column_names, category: :main, &block)
+        count             = input(category).upload_arel(arel, columns: column_names, &block)
         self.record_count = (record_count || 0) + count
         count
       end
 
-      # Upload the supplied slices for processing by workers
+      # @deprecated
+      def upload_mongo_query(criteria, *column_names, category: :main, &block)
+        count             = input(category).upload_mongo_query(criteria, columns: column_names, &block)
+        self.record_count = (record_count || 0) + count
+        count
+      end
+
+      # @deprecated
+      def upload_integer_range(start_id, last_id, category: :main, slice_batch_size: 1_000)
+        count             = input(category).upload_integer_range(start_id, last_id, slice_batch_size: slice_batch_size)
+        self.record_count = (record_count || 0) + count
+        count
+      end
+
+      # @deprecated
+      def upload_integer_range_in_reverse_order(start_id, last_id, category: :main, slice_batch_size: 1_000)
+        count             = input(category).upload_integer_range_in_reverse_order(start_id, last_id, slice_batch_size: slice_batch_size)
+        self.record_count = (record_count || 0) + count
+        count
+      end
+
+      # Upload the supplied slice for processing by workers
       #
       # Updates the record_count after adding the records
       #
@@ -427,50 +449,28 @@ module RocketJob
         # Store the output file name in the category
         category.file_name = stream if !block && (stream.is_a?(String) || stream.is_a?(IOStreams::Path))
 
-        if output_collection.binary?
-          raise(ArgumentError, "A `header_line` is not supported with binary output collections") if header_line
-
-          return output_collection.download(&block) if block
+        header_line ||= category.render_header
 
-          IOStreams.new(stream || category.file_name).stream(:none).writer(**args) do |io|
-            output_collection.download { |record| io << record[:binary] }
-          end
-        else
-          header_line ||= category.render_header
+        return output_collection.download(header_line: header_line, &block) if block
 
-          return output_collection.download(header_line: header_line, &block) if block
+        raise(ArgumentError, "Missing mandatory `stream` or `category.file_name`") unless stream || category.file_name
 
-          raise(ArgumentError, "Missing mandatory `stream` or `category.file_name`") unless stream || category.file_name
+        if output_collection.slice_class.binary_format
+          binary_header_line = output_collection.slice_class.to_binary(header_line) if header_line
 
+          # Don't overwrite supplied stream options if any
+          stream = stream&.is_a?(IOStreams::Stream) ? stream.dup : IOStreams.new(category.file_name)
+          stream.remove_from_pipeline(output_collection.slice_class.binary_format)
+          stream.writer(**args) do |io|
+            # TODO: Binary formats should return the record count, instead of the slice count.
+            output_collection.download(header_line: binary_header_line) { |record| io.write(record) }
+          end
+        else
           IOStreams.new(stream || category.file_name).writer(:line, **args) do |io|
             output_collection.download(header_line: header_line) { |record| io << record }
           end
         end
       end
-
-      private
-
-      # Return a lambda to extract the header row from the uploaded file.
-      def rocket_job_upload_header_lambda(category, on_first)
-        case category.mode
-        when :line
-          lambda do |line|
-            category.tabular.parse_header(line)
-            category.cleanse_header!
-            category.columns = category.tabular.header.columns
-            # Call chained on_first if present
-            on_first&.call(line)
-          end
-        when :array
-          lambda do |row|
-            category.tabular.header.columns = row
-            category.cleanse_header!
-            category.columns = category.tabular.header.columns
-            # Call chained on_first if present
-            on_first&.call(line)
-          end
-        end
-      end
     end
   end
 end