mindee 4.3.0 → 4.4.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 05c6fc0946b9d844042798b1b01160d5fbbdfa69f7b4a4e02ffb79e062680671
-  data.tar.gz: 23bfb847706e7c3d9bb15cd0acae3aa25d33c8fbbd126d8a6db0d55a00cb268c
+  metadata.gz: 3e31fcdde3f020eb4f16f99878bd4dfa65c417212b55938ed47acd0dfbab4eb5
+  data.tar.gz: 26d113de4272d259361167058cdb2c2a0cf25f8c15d4ebe4765a8b07ef50b751
 SHA512:
-  metadata.gz: d693ccee678f81b71efaf221b6a5342e55e826458d345a9d5ae4f2d8644a729be6cd58b50e43481d12b5d8ed46afe02772acdc3c4b7f2fb9442de35e3d971b15
-  data.tar.gz: ee173692eff34f62deed6930e703d25996cdaf5910e5ff8a6ec6f771ad366af397122cb45f99a96755f0014a517b226c13f1c5a2af83588d1697f0599425a13f
+  metadata.gz: f943da727ff6efb3c93e49e8aef488ed0167473d111544d28b959e4e6b45a90af4fa7c330fa261ed886012a4ad53a218ad593666bebf22595c254215bab66563
+  data.tar.gz: ca2bbf7e537b50fc04f8a8503f5e8d97362efd4e8e4d3d2ffb9f3c4c4a37fb6b919ec6686c02eb4204a40129fe05938a1201efb3a5221fc6b536cf47941a2827

data/.rubocop.yml

@@ -10,7 +10,7 @@ AllCops:
     - local_test/*
     - Steepfile
 
-  TargetRubyVersion: 3.0
+  TargetRubyVersion: 3.0.0
   SuggestExtensions: false
 
 Gemspec/DevelopmentDependencies:

data/CHANGELOG.md

@@ -1,5 +1,21 @@
 # Mindee Ruby API Library Changelog
 
+
+## v4.4.1 - 2025-05-13
+### Fixes
+* :bug: fix bad option in CLI
+* :bug: don't open the PDF unless needed
+
+
+## v4.4.0 - 2025-04-23
+### Changes
+* :sparkles: add support for workflow polling
+* :sparkles: add extras accessor from inference
+### Fixes
+* :recycle: fix misc typing issues
+* :bug: fix improper return format for `raw_http`
+
+
 ## v4.3.0 - 2025-04-08
 ### Changes
 * :sparkles: add support for Financial Document V1.12

data/bin/mindee.rb

@@ -99,7 +99,7 @@ else
 end
 
 if options[:cut_pages].nil? || !options[:cut_pages].is_a?(Integer) || options[:cut_pages] < 0
-  page_options = options[:cut_pages].nil?
+  page_options = nil
 else
   page_options = Mindee::PageOptions.new(params: {
     page_indexes: (0..options[:cut_pages].to_i).to_a,

data/docs/code_samples/workflow_execution.txt

@@ -3,7 +3,7 @@
 # gem install mindee
 #
 
-require_relative 'mindee'
+require 'mindee'
 
 workflow_id = 'workflow-id'
 

data/docs/code_samples/workflow_polling.txt

@@ -0,0 +1,36 @@
+#
+# Install the Ruby client library by running:
+# gem install mindee
+#
+
+require 'mindee'
+
+workflow_id = 'workflow-id'
+
+# Init a new client
+mindee_client = Mindee::Client.new
+
+# Load a file from disk
+input_source = mindee_client.source_from_path('path/to/my/file.ext')
+
+# Initialize a custom endpoint for this product
+custom_endpoint = mindee_client.create_endpoint(
+  account_name: 'my-account',
+  endpoint_name: 'my-endpoint',
+  version: 'my-version'
+)
+
+# Parse the file
+result = mindee_client.parse(
+  input_source,
+  Mindee::Product::Universal::Universal,
+  endpoint: custom_endpoint,
+  options: {
+    rag: true,
+    workflow_id: workflow_id
+  }
+)
+
+# Print a full summary of the parsed data in RST format
+puts result.document
+

data/lib/mindee/client.rb

@@ -54,8 +54,8 @@ module Mindee
   # @!attribute delay_sec [Numeric] Delay between polling attempts. Defaults to 1.5.
   # @!attribute max_retries [Integer] Maximum number of retries. Defaults to 80.
   class ParseOptions
-    attr_accessor :all_words, :full_text, :close_file, :page_options, :cropper,
-                  :initial_delay_sec, :delay_sec, :max_retries
+    attr_accessor :all_words, :full_text, :close_file, :page_options, :cropper, :rag,
+                  :workflow_id, :initial_delay_sec, :delay_sec, :max_retries
 
     def initialize(params: {})
       params = params.transform_keys(&:to_sym)
@@ -66,6 +66,8 @@ module Mindee
       raw_page_options = PageOptions.new(params: raw_page_options) unless raw_page_options.is_a?(PageOptions)
       @page_options = raw_page_options
       @cropper = params.fetch(:cropper, false)
+      @rag = params.fetch(:rag, false)
+      @workflow_id = params.fetch(:workflow_id, nil)
       @initial_delay_sec = params.fetch(:initial_delay_sec, 2)
       @delay_sec = params.fetch(:delay_sec, 1.5)
       @max_retries = params.fetch(:max_retries, 80)
@@ -176,13 +178,10 @@ module Mindee
 
       prediction, raw_http = endpoint.predict(
         input_source,
-        options.all_words,
-        options.full_text,
-        options.close_file,
-        options.cropper
+        options
       )
 
-      Mindee::Parsing::Common::ApiResponse.new(product_class, prediction, raw_http)
+      Mindee::Parsing::Common::ApiResponse.new(product_class, prediction, raw_http.to_s)
     end
 
     # Enqueue a document for async parsing
@@ -207,6 +206,8 @@ module Mindee
     #       - `:on_min_pages` [Integer] Apply the operation only if the document has at least this many pages.
     #   * `:cropper` [bool] Whether to include cropper results for each page.
     #       This performs a cropping operation on the server and will increase response time.
+    #   * `:rag` [bool] Whether to enable Retrieval-Augmented Generation. Only works if a Workflow ID is provided.
+    #   * `:workflow_id` [String, nil] ID of the workflow to use.
     # @param endpoint [Mindee::HTTP::Endpoint] Endpoint of the API.
     # @return [Mindee::Parsing::Common::ApiResponse]
     def enqueue(input_source, product_class, endpoint: nil, options: {})
@@ -216,12 +217,9 @@ module Mindee
 
       prediction, raw_http = endpoint.predict_async(
         input_source,
-        opts.all_words,
-        opts.full_text,
-        opts.close_file,
-        opts.cropper
+        opts
       )
-      Mindee::Parsing::Common::ApiResponse.new(product_class, prediction, raw_http)
+      Mindee::Parsing::Common::ApiResponse.new(product_class, prediction, raw_http.to_json)
     end
 
     # Parses a queued document
@@ -236,7 +234,7 @@ module Mindee
       endpoint = initialize_endpoint(product_class) if endpoint.nil?
       logger.debug("Fetching queued document as '#{endpoint.url_root}'")
       prediction, raw_http = endpoint.parse_async(job_id)
-      Mindee::Parsing::Common::ApiResponse.new(product_class, prediction, raw_http)
+      Mindee::Parsing::Common::ApiResponse.new(product_class, prediction, raw_http.to_json)
     end
 
     # Enqueue a document for async parsing and automatically try to retrieve it
@@ -261,6 +259,8 @@ module Mindee
     #       - `:on_min_pages` [Integer] Apply the operation only if the document has at least this many pages.
     #   * `:cropper` [bool, nil] Whether to include cropper results for each page.
     #       This performs a cropping operation on the server and will increase response time.
+    #   * `:rag` [bool] Whether to enable Retrieval-Augmented Generation. Only works if a Workflow ID is provided.
+    #   * `:workflow_id` [String, nil] ID of the workflow to use.
     #   * `:initial_delay_sec` [Numeric] Initial delay before polling. Defaults to 2.
     #   * `:delay_sec` [Numeric] Delay between polling attempts. Defaults to 1.5.
     #   * `:max_retries` [Integer] Maximum number of retries. Defaults to 80.
@@ -323,8 +323,7 @@ module Mindee
     def execute_workflow(input_source, workflow_id, options: {})
       opts = options.is_a?(WorkflowOptions) ? options : WorkflowOptions.new(params: options)
       if opts.respond_to?(:page_options) && input_source.is_a?(Input::Source::LocalInputSource)
-        process_pdf_if_required(input_source,
-                                opts)
+        process_pdf_if_required(input_source, opts)
       end
 
       workflow_endpoint = Mindee::HTTP::WorkflowEndpoint.new(workflow_id, api_key: @api_key)
@@ -496,7 +495,7 @@ module Mindee
     # @param opts [ParseOptions]
     def process_pdf_if_required(input_source, opts)
       return unless input_source.is_a?(Mindee::Input::Source::LocalInputSource) &&
-                    opts.page_options &&
+                    opts.page_options.on_min_pages &&
                     input_source.pdf?
 
       input_source.process_pdf(opts.page_options)

data/lib/mindee/http/endpoint.rb

@@ -34,6 +34,8 @@ module Mindee
       attr_reader :request_timeout
       # @return [String]
       attr_reader :url_root
+      # @return [String]
+      attr_reader :base_url
 
       def initialize(owner, url_name, version, api_key: '')
         @owner = owner
@@ -44,25 +46,19 @@ module Mindee
           logger.debug('API key set from environment')
         end
         @api_key = api_key.nil? || api_key.empty? ? ENV.fetch(API_KEY_ENV_NAME, API_KEY_DEFAULT) : api_key
-        base_url = ENV.fetch(BASE_URL_ENV_NAME, BASE_URL_DEFAULT)
-        @url_root = "#{base_url.chomp('/')}/products/#{@owner}/#{@url_name}/v#{@version}"
+        @base_url = ENV.fetch(BASE_URL_ENV_NAME, BASE_URL_DEFAULT).chomp('/')
+        @url_root = "#{@base_url}/products/#{@owner}/#{@url_name}/v#{@version}"
       end
 
       # Call the prediction API.
       # @param input_source [Mindee::Input::Source::LocalInputSource, Mindee::Input::Source::URLInputSource]
-      # @param all_words [bool] Whether the full word extraction needs to be performed
-      # @param full_text [bool] Whether to include the full OCR text response in compatible APIs
-      # @param close_file [bool] Whether the file will be closed after reading
-      # @param cropper [bool] Whether a cropping operation will be applied
+      # @param opts [ParseOptions] Parse options.
       # @return [Array]
-      def predict(input_source, all_words, full_text, close_file, cropper)
+      def predict(input_source, opts)
         check_api_key
         response = predict_req_post(
           input_source,
-          all_words: all_words,
-          full_text: full_text,
-          close_file: close_file,
-          cropper: cropper
+          opts
         )
         if !response.nil? && response.respond_to?(:body)
           hashed_response = JSON.parse(response.body, object_class: Hash)
@@ -76,14 +72,11 @@ module Mindee
 
       # Call the prediction API.
       # @param input_source [Mindee::Input::Source::LocalInputSource, Mindee::Input::Source::URLInputSource]
-      # @param all_words [bool] Whether the full word extraction needs to be performed
-      # @param full_text [bool] Whether to include the full OCR text response in compatible APIs.
-      # @param close_file [bool] Whether the file will be closed after reading
-      # @param cropper [bool] Whether a cropping operation will be applied
+      # @param opts [ParseOptions, Hash] Parse options.
       # @return [Array]
-      def predict_async(input_source, all_words, full_text, close_file, cropper)
+      def predict_async(input_source, opts)
         check_api_key
-        response = document_queue_req_get(input_source, all_words, full_text, close_file, cropper)
+        response = document_queue_req_post(input_source, opts)
         if !response.nil? && response.respond_to?(:body)
           hashed_response = JSON.parse(response.body, object_class: Hash)
           return [hashed_response, response.body] if ResponseValidation.valid_async_response?(response)
@@ -100,7 +93,7 @@ module Mindee
       # @return [Array]
       def parse_async(job_id)
         check_api_key
-        response = document_queue_req(job_id)
+        response = document_queue_req_get(job_id)
         hashed_response = JSON.parse(response.body, object_class: Hash)
         return [hashed_response, response.body] if ResponseValidation.valid_async_response?(response)
 
@@ -112,17 +105,14 @@ module Mindee
       private
 
       # @param input_source [Mindee::Input::Source::LocalInputSource, Mindee::Input::Source::URLInputSource]
-      # @param all_words [bool] Whether the full word extraction needs to be performed
-      # @param full_text [bool] Whether to include the full OCR text response in compatible APIs.
-      # @param close_file [bool] Whether the file will be closed after reading
-      # @param cropper [bool] Whether a cropping operation will be applied
+      # @param opts [ParseOptions] Parse options.
       # @return [Net::HTTPResponse, nil]
-      def predict_req_post(input_source, all_words: false, full_text: false, close_file: true, cropper: false)
+      def predict_req_post(input_source, opts)
         uri = URI("#{@url_root}/predict")
 
         params = {} # : Hash[Symbol | String, untyped]
-        params[:cropper] = 'true' if cropper
-        params[:full_text_ocr] = 'true' if full_text
+        params[:cropper] = 'true' if opts.cropper
+        params[:full_text_ocr] = 'true' if opts.full_text
         uri.query = URI.encode_www_form(params)
 
         headers = {
@@ -131,32 +121,33 @@ module Mindee
         }
         req = Net::HTTP::Post.new(uri, headers)
         form_data = if input_source.is_a?(Mindee::Input::Source::URLInputSource)
-                      [['document', input_source.url]]
+                      [['document', input_source.url]] # : Array[untyped]
                     else
-                      [input_source.read_contents(close: close_file)]
+                      [input_source.read_contents(close: opts.close_file)] # : Array[untyped]
                     end
-        form_data.push ['include_mvision', 'true'] if all_words
+        form_data.push ['include_mvision', 'true'] if opts.all_words
 
         req.set_form(form_data, 'multipart/form-data')
-        response = nil
         Net::HTTP.start(uri.hostname, uri.port, use_ssl: true, read_timeout: @request_timeout) do |http|
-          response = http.request(req)
+          return http.request(req)
         end
-        response
+        raise Mindee::Errors::MindeeError, 'Could not resolve server response.'
       end
 
       # @param input_source [Mindee::Input::Source::LocalInputSource, Mindee::Input::Source::URLInputSource]
-      # @param all_words [bool] Whether the full word extraction needs to be performed
-      # @param full_text [bool] Whether to include the full OCR text response in compatible APIs.
-      # @param close_file [bool] Whether the file will be closed after reading
-      # @param cropper [bool] Whether a cropping operation will be applied
-      # @return [Net::HTTPResponse, nil]
-      def document_queue_req_get(input_source, all_words, full_text, close_file, cropper)
-        uri = URI("#{@url_root}/predict_async")
+      # @param opts [ParseOptions] Parse options.
+      # @return [Net::HTTPResponse]
+      def document_queue_req_post(input_source, opts)
+        uri = if opts.workflow_id
+                URI("#{@base_url}/workflows/#{opts.workflow_id}/predict_async")
+              else
+                URI("#{@url_root}/predict_async")
+              end
 
         params = {} # : Hash[Symbol | String, untyped]
-        params[:cropper] = 'true' if cropper
-        params[:full_text_ocr] = 'true' if full_text
+        params[:cropper] = 'true' if opts.cropper
+        params[:full_text_ocr] = 'true' if opts.full_text
+        params[:rag] = 'true' if opts.rag
         uri.query = URI.encode_www_form(params)
 
         headers = {
@@ -165,24 +156,23 @@ module Mindee
         }
         req = Net::HTTP::Post.new(uri, headers)
         form_data = if input_source.is_a?(Mindee::Input::Source::URLInputSource)
-                      [['document', input_source.url]]
+                      [['document', input_source.url]] # : Array[untyped]
                     else
-                      [input_source.read_contents(close: close_file)]
+                      [input_source.read_contents(close: opts.close_file)] # : Array[untyped]
                     end
-        form_data.push ['include_mvision', 'true'] if all_words
+        form_data.push ['include_mvision', 'true'] if opts.all_words
 
         req.set_form(form_data, 'multipart/form-data')
 
-        response = nil
         Net::HTTP.start(uri.hostname, uri.port, use_ssl: true, read_timeout: @request_timeout) do |http|
-          response = http.request(req)
+          return http.request(req)
         end
-        response
+        raise Mindee::Errors::MindeeError, 'Could not resolve server response.'
       end
 
       # @param job_id [String]
       # @return [Net::HTTPResponse, nil]
-      def document_queue_req(job_id)
+      def document_queue_req_get(job_id)
         uri = URI("#{@url_root}/documents/queue/#{job_id}")
 
         headers = {

data/lib/mindee/http/response_validation.rb

@@ -42,7 +42,7 @@ module Mindee
       # Checks and correct the response object depending on the possible kinds of returns.
       # @param response [Net::HTTPResponse]
       def self.clean_request!(response)
-        return response if (response.code.to_i < 200) || (response.code.to_i > 302)
+        return response if (response.code.to_i < 200) || (response.code.to_i > 302) # : Net::HTTPResponse
 
         return response if response.body.empty?
 

data/lib/mindee/parsing/common/api_response.rb

@@ -35,7 +35,7 @@ module Mindee
 
         # @param product_class [Mindee::Inference]
         # @param http_response [Hash]
-        # @param raw_http [String]
+        # @param raw_http [Hash]
         def initialize(product_class, http_response, raw_http)
           logger.debug('Handling API response')
           @raw_http = raw_http.to_s

data/lib/mindee/parsing/common/document.rb

@@ -69,7 +69,7 @@ module Mindee
 
           full_text_ocr = String.new
           raw_prediction.dig('inference', 'pages').each do |page|
-            full_text_ocr << (page['extras']['full_text_ocr']['content'])
+            full_text_ocr << page['extras']['full_text_ocr']['content']
           end
           artificial_text_obj = { 'content' => full_text_ocr }
           if @extras.nil? || @extras.empty?

data/lib/mindee/parsing/common/extras/extras.rb

@@ -13,6 +13,8 @@ module Mindee
           attr_reader :cropper
           # @return [Mindee::Parsing::Common::Extras::FullTextOCRExtra, nil]
           attr_reader :full_text_ocr
+          # @return [RAGExtra, nil]
+          attr_reader :rag
 
           def initialize(raw_prediction)
             if raw_prediction['cropper']
@@ -21,9 +23,10 @@ module Mindee
             if raw_prediction['full_text_ocr']
               @full_text_ocr = Mindee::Parsing::Common::Extras::FullTextOCRExtra.new(raw_prediction['full_text_ocr'])
             end
+            @rag = Mindee::Parsing::Common::Extras::RAGExtra.new(raw_prediction['rag']) if raw_prediction['rag']
 
             raw_prediction.each do |key, value|
-              instance_variable_set("@#{key}", value) unless ['cropper', 'full_text_ocr'].include?(key)
+              instance_variable_set("@#{key}", value) unless ['cropper', 'full_text_ocr', 'rag'].include?(key)
             end
           end
 

data/lib/mindee/parsing/common/extras/rag_extra.rb

@@ -0,0 +1,24 @@
+# frozen_string_literal: true
+
+module Mindee
+  module Parsing
+    module Common
+      module Extras
+        # Retrieval-Augmented Generation extra.
+        class RAGExtra
+          # ID of the matching document
+          # @return [String, nil]
+          attr_reader :matching_document_id
+
+          def initialize(raw_prediction)
+            @matching_document_id = raw_prediction['matching_document_id'] if raw_prediction['matching_document_id']
+          end
+
+          def to_s
+            @matching_document_id || ''
+          end
+        end
+      end
+    end
+  end
+end

data/lib/mindee/parsing/common/extras.rb

@@ -3,3 +3,4 @@
 require_relative 'extras/extras'
 require_relative 'extras/cropper_extra'
 require_relative 'extras/full_text_ocr_extra'
+require_relative 'extras/rag_extra'

data/lib/mindee/parsing/common/inference.rb

@@ -29,6 +29,8 @@ module Mindee
         # Whether this product has access to synchronous endpoint.
         # @return [bool]
         attr_reader :has_sync
+        # @return [Mindee::Parsing::Common::Extras::Extras] Potential Extras fields sent back along the prediction.
+        attr_reader :extras
 
         @endpoint_name = nil
         @endpoint_version = nil
@@ -40,6 +42,7 @@ module Mindee
           @is_rotation_applied = raw_prediction['is_rotation_applied']
           @product = Product.new(raw_prediction['product'])
           @pages = [] # : Array[Page]
+          @extras = Extras::Extras.new(raw_prediction['extras'])
         end
 
         # @return [String]

data/lib/mindee/parsing/standard/tax_field.rb

@@ -39,10 +39,10 @@ module Mindee
         def to_s
           printable = printable_values
           out_str = String.new
-          out_str << ("Base: #{printable[:base]}")
-          out_str << (", Code: #{printable[:code]}")
-          out_str << (", Rate (%): #{printable[:rate]}")
-          out_str << (", Amount: #{printable[:value]}")
+          out_str << "Base: #{printable[:base]}"
+          out_str << ", Code: #{printable[:code]}"
+          out_str << ", Rate (%): #{printable[:rate]}"
+          out_str << ", Amount: #{printable[:value]}"
           out_str.strip
         end
 
@@ -60,10 +60,10 @@ module Mindee
         def to_table_line
           printable = printable_values
           out_str = String.new
-          out_str << ("| #{printable[:base].ljust(13, ' ')}")
-          out_str << (" | #{printable[:code].ljust(6, ' ')}")
-          out_str << (" | #{printable[:rate].ljust(8, ' ')}")
-          out_str << (" | #{printable[:value].ljust(13, ' ')} |")
+          out_str << "| #{printable[:base].ljust(13, ' ')}"
+          out_str << " | #{printable[:code].ljust(6, ' ')}"
+          out_str << " | #{printable[:rate].ljust(8, ' ')}"
+          out_str << " | #{printable[:value].ljust(13, ' ')} |"
           out_str.strip
         end
       end
@@ -94,7 +94,7 @@ module Mindee
           return '' if nil? || empty?
 
           out_str = String.new
-          out_str << ("\n#{line_separator('-')}")
+          out_str << "\n#{line_separator('-')}"
           out_str << "\n  | Base          | Code   | Rate (%) | Amount        |"
           out_str << "\n#{line_separator('=')}"
           each do |entry|

data/lib/mindee/pdf/pdf_extractor.rb

@@ -58,7 +58,7 @@ module Mindee
             end
             formatted_max_index = format('%03d', page_index_list[page_index_list.length - 1] + 1).to_s
             field_filename = "#{basename}_#{format('%03d',
-                                                   (page_index_list[0] + 1))}-#{formatted_max_index}#{extension}"
+                                                   page_index_list[0] + 1)}-#{formatted_max_index}#{extension}"
             extracted_pdf = Mindee::PDF::PDFExtractor::ExtractedPDF.new(cut_pages(page_index_list),
                                                                         field_filename)
             extracted_pdfs << extracted_pdf

data/lib/mindee/product/bill_of_lading/bill_of_lading_v1_carrier_items.rb

@@ -46,14 +46,14 @@ module Mindee
             "\n  #{entry.to_table_line}\n#{self.class.line_items_separator('-')}"
           end.join
           out_str = String.new
-          out_str << ("\n#{self.class.line_items_separator('-')}\n ")
+          out_str << "\n#{self.class.line_items_separator('-')}\n "
           out_str << ' | Description                         '
           out_str << ' | Gross Weight'
           out_str << ' | Measurement'
           out_str << ' | Measurement Unit'
           out_str << ' | Quantity'
           out_str << ' | Weight Unit'
-          out_str << (" |\n#{self.class.line_items_separator('=')}")
+          out_str << " |\n#{self.class.line_items_separator('=')}"
           out_str + lines
         end
       end

data/lib/mindee/product/financial_document/financial_document_v1_line_items.rb

@@ -44,7 +44,7 @@ module Mindee
             "\n  #{entry.to_table_line}\n#{self.class.line_items_separator('-')}"
           end.join
           out_str = String.new
-          out_str << ("\n#{self.class.line_items_separator('-')}\n ")
+          out_str << "\n#{self.class.line_items_separator('-')}\n "
           out_str << ' | Description                         '
           out_str << ' | Product code'
           out_str << ' | Quantity'
@@ -53,7 +53,7 @@ module Mindee
           out_str << ' | Total Amount'
           out_str << ' | Unit of measure'
           out_str << ' | Unit Price'
-          out_str << (" |\n#{self.class.line_items_separator('=')}")
+          out_str << " |\n#{self.class.line_items_separator('=')}"
           out_str + lines
         end
       end

data/lib/mindee/product/fr/bank_statement/bank_statement_v2_transactions.rb

@@ -40,11 +40,11 @@ module Mindee
               "\n  #{entry.to_table_line}\n#{self.class.line_items_separator('-')}"
             end.join
             out_str = String.new
-            out_str << ("\n#{self.class.line_items_separator('-')}\n ")
+            out_str << "\n#{self.class.line_items_separator('-')}\n "
             out_str << ' | Amount    '
             out_str << ' | Date      '
             out_str << ' | Description                         '
-            out_str << (" |\n#{self.class.line_items_separator('=')}")
+            out_str << " |\n#{self.class.line_items_separator('=')}"
             out_str + lines
           end
         end

data/lib/mindee/product/fr/energy_bill/energy_bill_v1_energy_usages.rb

@@ -50,7 +50,7 @@ module Mindee
               "\n  #{entry.to_table_line}\n#{self.class.line_items_separator('-')}"
             end.join
             out_str = String.new
-            out_str << ("\n#{self.class.line_items_separator('-')}\n ")
+            out_str << "\n#{self.class.line_items_separator('-')}\n "
             out_str << ' | Consumption'
             out_str << ' | Description                         '
             out_str << ' | End Date  '
@@ -59,7 +59,7 @@ module Mindee
             out_str << ' | Total    '
             out_str << ' | Unit of Measure'
             out_str << ' | Unit Price'
-            out_str << (" |\n#{self.class.line_items_separator('=')}")
+            out_str << " |\n#{self.class.line_items_separator('=')}"
             out_str + lines
           end
         end

data/lib/mindee/product/fr/energy_bill/energy_bill_v1_subscriptions.rb

@@ -48,14 +48,14 @@ module Mindee
               "\n  #{entry.to_table_line}\n#{self.class.line_items_separator('-')}"
             end.join
             out_str = String.new
-            out_str << ("\n#{self.class.line_items_separator('-')}\n ")
+            out_str << "\n#{self.class.line_items_separator('-')}\n "
             out_str << ' | Description                         '
             out_str << ' | End Date  '
             out_str << ' | Start Date'
             out_str << ' | Tax Rate'
             out_str << ' | Total    '
             out_str << ' | Unit Price'
-            out_str << (" |\n#{self.class.line_items_separator('=')}")
+            out_str << " |\n#{self.class.line_items_separator('=')}"
             out_str + lines
           end
         end

data/lib/mindee/product/fr/energy_bill/energy_bill_v1_taxes_and_contributions.rb

@@ -48,14 +48,14 @@ module Mindee
               "\n  #{entry.to_table_line}\n#{self.class.line_items_separator('-')}"
             end.join
             out_str = String.new
-            out_str << ("\n#{self.class.line_items_separator('-')}\n ")
+            out_str << "\n#{self.class.line_items_separator('-')}\n "
             out_str << ' | Description                         '
             out_str << ' | End Date  '
             out_str << ' | Start Date'
             out_str << ' | Tax Rate'
             out_str << ' | Total    '
             out_str << ' | Unit Price'
-            out_str << (" |\n#{self.class.line_items_separator('=')}")
+            out_str << " |\n#{self.class.line_items_separator('=')}"
             out_str + lines
           end
         end

data/lib/mindee/product/fr/payslip/payslip_v2_salary_details.rb

@@ -48,12 +48,12 @@ module Mindee
               "\n  #{entry.to_table_line}\n#{self.class.line_items_separator('-')}"
             end.join
             out_str = String.new
-            out_str << ("\n#{self.class.line_items_separator('-')}\n ")
+            out_str << "\n#{self.class.line_items_separator('-')}\n "
             out_str << ' | Amount      '
             out_str << ' | Base     '
             out_str << ' | Description                         '
             out_str << ' | Rate     '
-            out_str << (" |\n#{self.class.line_items_separator('=')}")
+            out_str << " |\n#{self.class.line_items_separator('=')}"
             out_str + lines
           end
         end

data/lib/mindee/product/fr/payslip/payslip_v3_paid_time_offs.rb

@@ -49,13 +49,13 @@ module Mindee
               "\n  #{entry.to_table_line}\n#{self.class.line_items_separator('-')}"
             end.join
             out_str = String.new
-            out_str << ("\n#{self.class.line_items_separator('-')}\n ")
+            out_str << "\n#{self.class.line_items_separator('-')}\n "
             out_str << ' | Accrued  '
             out_str << ' | Period'
             out_str << ' | Type       '
             out_str << ' | Remaining'
             out_str << ' | Used     '
-            out_str << (" |\n#{self.class.line_items_separator('=')}")
+            out_str << " |\n#{self.class.line_items_separator('=')}"
             out_str + lines
           end
         end

data/lib/mindee/product/fr/payslip/payslip_v3_salary_details.rb

@@ -49,13 +49,13 @@ module Mindee
               "\n  #{entry.to_table_line}\n#{self.class.line_items_separator('-')}"
             end.join
             out_str = String.new
-            out_str << ("\n#{self.class.line_items_separator('-')}\n ")
+            out_str << "\n#{self.class.line_items_separator('-')}\n "
             out_str << ' | Amount      '
             out_str << ' | Base     '
             out_str << ' | Description                         '
             out_str << ' | Number'
             out_str << ' | Rate     '
-            out_str << (" |\n#{self.class.line_items_separator('=')}")
+            out_str << " |\n#{self.class.line_items_separator('=')}"
             out_str + lines
           end
         end

data/lib/mindee/product/invoice/invoice_v4_line_items.rb

@@ -44,7 +44,7 @@ module Mindee
             "\n  #{entry.to_table_line}\n#{self.class.line_items_separator('-')}"
           end.join
           out_str = String.new
-          out_str << ("\n#{self.class.line_items_separator('-')}\n ")
+          out_str << "\n#{self.class.line_items_separator('-')}\n "
           out_str << ' | Description                         '
           out_str << ' | Product code'
           out_str << ' | Quantity'
@@ -53,7 +53,7 @@ module Mindee
           out_str << ' | Total Amount'
           out_str << ' | Unit of measure'
           out_str << ' | Unit Price'
-          out_str << (" |\n#{self.class.line_items_separator('=')}")
+          out_str << " |\n#{self.class.line_items_separator('=')}"
           out_str + lines
         end
       end

data/lib/mindee/product/invoice_splitter/invoice_splitter_v1_invoice_page_groups.rb

@@ -37,9 +37,9 @@ module Mindee
             "\n  #{entry.to_table_line}\n#{self.class.line_items_separator('-')}"
           end.join
           out_str = String.new
-          out_str << ("\n#{self.class.line_items_separator('-')}\n ")
+          out_str << "\n#{self.class.line_items_separator('-')}\n "
           out_str << ' | Page Indexes                                                            '
-          out_str << (" |\n#{self.class.line_items_separator('=')}")
+          out_str << " |\n#{self.class.line_items_separator('=')}"
           out_str + lines
         end
       end

data/lib/mindee/product/nutrition_facts_label/nutrition_facts_label_v1_nutrients.rb

@@ -53,13 +53,13 @@ module Mindee
             "\n  #{entry.to_table_line}\n#{self.class.line_items_separator('-')}"
           end.join
           out_str = String.new
-          out_str << ("\n#{self.class.line_items_separator('-')}\n ")
+          out_str << "\n#{self.class.line_items_separator('-')}\n "
           out_str << ' | Daily Value'
           out_str << ' | Name                '
           out_str << ' | Per 100g'
           out_str << ' | Per Serving'
           out_str << ' | Unit'
-          out_str << (" |\n#{self.class.line_items_separator('=')}")
+          out_str << " |\n#{self.class.line_items_separator('=')}"
           out_str + lines
         end
       end

data/lib/mindee/product/receipt/receipt_v5_line_items.rb

@@ -40,12 +40,12 @@ module Mindee
             "\n  #{entry.to_table_line}\n#{self.class.line_items_separator('-')}"
           end.join
           out_str = String.new
-          out_str << ("\n#{self.class.line_items_separator('-')}\n ")
+          out_str << "\n#{self.class.line_items_separator('-')}\n "
           out_str << ' | Description                         '
           out_str << ' | Quantity'
           out_str << ' | Total Amount'
           out_str << ' | Unit Price'
-          out_str << (" |\n#{self.class.line_items_separator('=')}")
+          out_str << " |\n#{self.class.line_items_separator('=')}"
           out_str + lines
         end
       end

data/lib/mindee/product/resume/resume_v1_certificates.rb

@@ -44,12 +44,12 @@ module Mindee
             "\n  #{entry.to_table_line}\n#{self.class.line_items_separator('-')}"
           end.join
           out_str = String.new
-          out_str << ("\n#{self.class.line_items_separator('-')}\n ")
+          out_str << "\n#{self.class.line_items_separator('-')}\n "
           out_str << ' | Grade     '
           out_str << ' | Name                          '
           out_str << ' | Provider                 '
           out_str << ' | Year'
-          out_str << (" |\n#{self.class.line_items_separator('=')}")
+          out_str << " |\n#{self.class.line_items_separator('=')}"
           out_str + lines
         end
       end

data/lib/mindee/product/resume/resume_v1_educations.rb

@@ -47,7 +47,7 @@ module Mindee
             "\n  #{entry.to_table_line}\n#{self.class.line_items_separator('-')}"
           end.join
           out_str = String.new
-          out_str << ("\n#{self.class.line_items_separator('-')}\n ")
+          out_str << "\n#{self.class.line_items_separator('-')}\n "
           out_str << ' | Domain         '
           out_str << ' | Degree                   '
           out_str << ' | End Month'
@@ -55,7 +55,7 @@ module Mindee
           out_str << ' | School                   '
           out_str << ' | Start Month'
           out_str << ' | Start Year'
-          out_str << (" |\n#{self.class.line_items_separator('=')}")
+          out_str << " |\n#{self.class.line_items_separator('=')}"
           out_str + lines
         end
       end

data/lib/mindee/product/resume/resume_v1_languages.rb

@@ -42,10 +42,10 @@ module Mindee
             "\n  #{entry.to_table_line}\n#{self.class.line_items_separator('-')}"
           end.join
           out_str = String.new
-          out_str << ("\n#{self.class.line_items_separator('-')}\n ")
+          out_str << "\n#{self.class.line_items_separator('-')}\n "
           out_str << ' | Language'
           out_str << ' | Level               '
-          out_str << (" |\n#{self.class.line_items_separator('=')}")
+          out_str << " |\n#{self.class.line_items_separator('=')}"
           out_str + lines
         end
       end

data/lib/mindee/product/resume/resume_v1_professional_experiences.rb

@@ -49,7 +49,7 @@ module Mindee
             "\n  #{entry.to_table_line}\n#{self.class.line_items_separator('-')}"
           end.join
           out_str = String.new
-          out_str << ("\n#{self.class.line_items_separator('-')}\n ")
+          out_str << "\n#{self.class.line_items_separator('-')}\n "
           out_str << ' | Contract Type  '
           out_str << ' | Department'
           out_str << ' | Description                         '
@@ -59,7 +59,7 @@ module Mindee
           out_str << ' | Role                '
           out_str << ' | Start Month'
           out_str << ' | Start Year'
-          out_str << (" |\n#{self.class.line_items_separator('=')}")
+          out_str << " |\n#{self.class.line_items_separator('=')}"
           out_str + lines
         end
       end

data/lib/mindee/product/resume/resume_v1_social_networks_urls.rb

@@ -42,10 +42,10 @@ module Mindee
             "\n  #{entry.to_table_line}\n#{self.class.line_items_separator('-')}"
           end.join
           out_str = String.new
-          out_str << ("\n#{self.class.line_items_separator('-')}\n ")
+          out_str << "\n#{self.class.line_items_separator('-')}\n "
           out_str << ' | Name                '
           out_str << ' | URL                                               '
-          out_str << (" |\n#{self.class.line_items_separator('=')}")
+          out_str << " |\n#{self.class.line_items_separator('=')}"
           out_str + lines
         end
       end

data/lib/mindee/product/us/healthcare_card/healthcare_card_v1_copays.rb

@@ -39,10 +39,10 @@ module Mindee
               "\n  #{entry.to_table_line}\n#{self.class.line_items_separator('-')}"
             end.join
             out_str = String.new
-            out_str << ("\n#{self.class.line_items_separator('-')}\n ")
+            out_str << "\n#{self.class.line_items_separator('-')}\n "
             out_str << ' | Service Fees'
             out_str << ' | Service Name        '
-            out_str << (" |\n#{self.class.line_items_separator('=')}")
+            out_str << " |\n#{self.class.line_items_separator('=')}"
             out_str + lines
           end
         end

data/lib/mindee/product/us/us_mail/us_mail_v2_recipient_addresses.rb

@@ -45,7 +45,7 @@ module Mindee
               "\n  #{entry.to_table_line}\n#{self.class.line_items_separator('-')}"
             end.join
             out_str = String.new
-            out_str << ("\n#{self.class.line_items_separator('-')}\n ")
+            out_str << "\n#{self.class.line_items_separator('-')}\n "
             out_str << ' | City           '
             out_str << ' | Complete Address                   '
             out_str << ' | Is Address Change'
@@ -53,7 +53,7 @@ module Mindee
             out_str << ' | Private Mailbox Number'
             out_str << ' | State'
             out_str << ' | Street                   '
-            out_str << (" |\n#{self.class.line_items_separator('=')}")
+            out_str << " |\n#{self.class.line_items_separator('=')}"
             out_str + lines
           end
         end

data/lib/mindee/product/us/us_mail/us_mail_v3_recipient_addresses.rb

@@ -46,7 +46,7 @@ module Mindee
               "\n  #{entry.to_table_line}\n#{self.class.line_items_separator('-')}"
             end.join
             out_str = String.new
-            out_str << ("\n#{self.class.line_items_separator('-')}\n ")
+            out_str << "\n#{self.class.line_items_separator('-')}\n "
             out_str << ' | City           '
             out_str << ' | Complete Address                   '
             out_str << ' | Is Address Change'
@@ -55,7 +55,7 @@ module Mindee
             out_str << ' | State'
             out_str << ' | Street                   '
             out_str << ' | Unit           '
-            out_str << (" |\n#{self.class.line_items_separator('=')}")
+            out_str << " |\n#{self.class.line_items_separator('=')}"
             out_str + lines
           end
         end

data/lib/mindee/version.rb

@@ -3,7 +3,7 @@
 # Mindee
 module Mindee
   # Current version.
-  VERSION = '4.3.0'
+  VERSION = '4.4.1'
 
   # Finds and return the current platform.
   # @return [Symbol, Hash[Symbol | String, Regexp], Nil?]

data/mindee.gemspec

@@ -37,7 +37,7 @@ Gem::Specification.new do |spec|
   spec.add_development_dependency 'rake', '~> 13.2'
   spec.add_development_dependency 'rbs', '~> 3.6'
   spec.add_development_dependency 'rspec', '~> 3.13'
-  spec.add_development_dependency 'rubocop', '~> 1.70'
+  spec.add_development_dependency 'rubocop', '~> 1.75'
   spec.add_development_dependency 'steep', '~> 1.7'
   spec.add_development_dependency 'yard', '~> 0.9'
 end

data/sig/custom/net_http.rbs

@@ -29,6 +29,9 @@ module Net
   class HTTPResponse
     def self.body: -> untyped
     def body: -> untyped
+    def []: (untyped) -> untyped
+    def key?: (untyped) -> bool
+    def code: -> String
   end
 
   class HTTPRedirection

data/sig/mindee/client.rbs

@@ -17,6 +17,8 @@ module Mindee
     attr_accessor close_file: bool
     attr_accessor page_options: PageOptions
     attr_accessor cropper: bool
+    attr_accessor rag: bool
+    attr_accessor workflow_id: String | nil
     attr_accessor initial_delay_sec: Integer | Float
     attr_accessor delay_sec: Integer | Float
     attr_accessor max_retries: Integer
@@ -36,7 +38,7 @@ module Mindee
   end
 
   class Client
-    def initialize: (api_key: String) -> void
+    def initialize: (?api_key: String) -> void
     def logger: () -> untyped
     def parse: (Input::Source::LocalInputSource | Input::Source::URLInputSource, untyped, ?endpoint: (HTTP::Endpoint?), options: ParseOptions | Hash[Symbol | String, untyped]) -> Parsing::Common::ApiResponse
     def parse_sync: (Input::Source::LocalInputSource | Input::Source::URLInputSource, untyped, HTTP::Endpoint, ParseOptions) -> Parsing::Common::ApiResponse
@@ -57,7 +59,7 @@ module Mindee
     def process_pdf_if_required: (Input::Source::LocalInputSource, ParseOptions | WorkflowOptions) -> void
     def initialize_endpoint: (singleton(Parsing::Common::Inference), ?endpoint_name: String, ?account_name: String, ?version: String) -> HTTP::Endpoint
     def validate_async_params: (Integer | Float, Integer | Float, Integer) -> void
-    def fix_endpoint_name: (singleton(Parsing::Common::Inference), String?) -> String?
+    def fix_endpoint_name: (singleton(Parsing::Common::Inference), String?) -> String
     def fix_account_name: (String) -> String
     def fix_version: (singleton(Parsing::Common::Inference), String) -> String
   end

data/sig/mindee/http/endpoint.rbs

@@ -13,14 +13,15 @@ module Mindee
       def api_key: -> String?
       def request_timeout: -> Integer
       def url_root: -> String
-      def initialize: (untyped, untyped, untyped, ?api_key: String) -> String
-      def predict: (untyped, untyped, untyped, untyped, untyped) -> [untyped, untyped]
-      def predict_async: (untyped, untyped, untyped, untyped, untyped) -> [untyped, untyped]
-      def parse_async: (untyped) -> [untyped, untyped]
-      def predict_req_post: (untyped, ?all_words: false, ?full_text: false, ?close_file: true, ?cropper: false) -> untyped
-      def document_queue_req_get: (untyped, untyped, untyped, untyped, untyped) -> untyped
-      def document_queue_req: (untyped) -> untyped
-      def check_api_key: -> nil
+      def base_url: -> String
+      def initialize: (String, String, String | nil, ?api_key: String) -> String
+      def predict: (Input::Source::LocalInputSource | Input::Source::URLInputSource, ParseOptions) -> [Net::HTTPResponse, Hash[Symbol, untyped]]
+      def predict_async: (Input::Source::LocalInputSource | Input::Source::URLInputSource, ParseOptions) -> [Net::HTTPResponse, Hash[Symbol, untyped]]
+      def parse_async: (String) -> [Net::HTTPResponse, Hash[Symbol, untyped]]
+      def predict_req_post: (Input::Source::LocalInputSource | Input::Source::URLInputSource, ParseOptions) -> Net::HTTPResponse
+      def document_queue_req_post: (Input::Source::LocalInputSource | Input::Source::URLInputSource, ParseOptions) -> Net::HTTPResponse
+      def document_queue_req_get: (untyped) -> Net::HTTPResponse
+      def check_api_key: -> void
     end
   end
 end

data/sig/mindee/http/response_validation.rbs

@@ -2,9 +2,9 @@
 module Mindee
   module HTTP
     module ResponseValidation
-      def self.valid_sync_response?: (untyped) -> bool
-      def self.valid_async_response?: (untyped) -> bool
-      def self.clean_request!: (untyped) -> nil
+      def self.valid_sync_response?: (Net::HTTPResponse) -> bool
+      def self.valid_async_response?: (Net::HTTPResponse) -> bool
+      def self.clean_request!: (Net::HTTPResponse) -> void
     end
   end
 end

data/sig/mindee/parsing/common/api_response.rbs

@@ -20,7 +20,7 @@ module Mindee
         def job: -> Parsing::Common::Job?
         def api_request: -> Parsing::Common::ApiRequest?
         def raw_http: -> String
-        def initialize: (singleton(Parsing::Common::Inference), Hash[Symbol | String, untyped], String) -> void
+        def initialize: (singleton(Parsing::Common::Inference), Hash[Symbol | String, untyped] | Net::HTTPResponse, String) -> void
       end
     end
   end

data/sig/mindee/parsing/common/extras/full_text_ocr_extra.rbs

@@ -4,8 +4,8 @@ module Mindee
     module Common
       module Extras
         class FullTextOCRExtra
-          def contents: -> untyped
-          def language: -> untyped
+          def contents: -> String?
+          def language: -> String?
           def initialize: (untyped) -> nil
           def to_s: -> String
         end

data/sig/mindee/parsing/common/extras/rag_extra.rbs

@@ -0,0 +1,15 @@
+# lib/mindee/parsing/common/extras/rag_extra.rb
+
+module Mindee
+  module Parsing
+    module Common
+      module Extras
+        class RAGExtra
+          def matching_document_id: -> String?
+          def initialize: (untyped) -> nil
+          def to_s: -> String
+        end
+      end
+    end
+  end
+end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: mindee
 version: !ruby/object:Gem::Version
-  version: 4.3.0
+  version: 4.4.1
 platform: ruby
 authors:
 - Mindee, SA
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2025-04-08 00:00:00.000000000 Z
+date: 2025-05-13 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: base64
@@ -148,14 +148,14 @@ dependencies:
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '1.70'
+        version: '1.75'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '1.70'
+        version: '1.75'
 - !ruby/object:Gem::Dependency
   name: steep
   requirement: !ruby/object:Gem::Requirement
@@ -246,6 +246,7 @@ files:
 - docs/code_samples/us_mail_v2_async.txt
 - docs/code_samples/us_mail_v3_async.txt
 - docs/code_samples/workflow_execution.txt
+- docs/code_samples/workflow_polling.txt
 - docs/getting_started.md
 - docs/global_products.md
 - docs/global_products/barcode_reader_v1.md
@@ -327,6 +328,7 @@ files:
 - lib/mindee/parsing/common/extras/cropper_extra.rb
 - lib/mindee/parsing/common/extras/extras.rb
 - lib/mindee/parsing/common/extras/full_text_ocr_extra.rb
+- lib/mindee/parsing/common/extras/rag_extra.rb
 - lib/mindee/parsing/common/inference.rb
 - lib/mindee/parsing/common/job.rb
 - lib/mindee/parsing/common/ocr.rb
@@ -575,6 +577,7 @@ files:
 - sig/mindee/parsing/common/extras/cropper_extra.rbs
 - sig/mindee/parsing/common/extras/extras.rbs
 - sig/mindee/parsing/common/extras/full_text_ocr_extra.rbs
+- sig/mindee/parsing/common/extras/rag_extra.rbs
 - sig/mindee/parsing/common/inference.rbs
 - sig/mindee/parsing/common/job.rbs
 - sig/mindee/parsing/common/ocr/mvision_v1.rbs