mindee 4.2.0 → 4.4.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: e6c09d5108c2470661075484f14849c7205cbef88d481427bcb612d36ede9835
-  data.tar.gz: e7b6fa99499a9c3a3d82a9c87ab40dc07cbecd48a4f2d82200cfea2729f9d5cb
+  metadata.gz: f9a2e1c98c3971454ebfaefe6902121bfab6e577f693e36a525739eb2dfbeeb9
+  data.tar.gz: 7f35351f0f647a165c82f512395e8883647e96165968869bf3904f4c2bf6b4ab
 SHA512:
-  metadata.gz: b97c068ddeefb0908e7d9753c0439c86393d1a04b13bf8f256c3dbc93399f213b0405375a036715906a8087eba3eb5126d8e69880c819ef4b0d5b7bf61459e88
-  data.tar.gz: a15ff1b478675ec58dfd61d0f0483dcc7ab23df1fc02624e8133d5f636101e82f1eeb4f823ebe19a3a06ce70c4288270ebaca7029514e91f204298ad5843c924
+  metadata.gz: '072284cd8bb18b6190b870aa014fb278744a6db35d1210ca57dbddb1d78c71a2675f2675a7ce581b83cbb1fdcd88cdeb6dedd6892c2402b43cf3f711425f9e46'
+  data.tar.gz: 4bad8302b9bef1496fc47e5f04cb301dd3c81a99dff5800ffbc1aaf3ef4bf1f5a6493f4908178b337e7f25596b7f9ce2946d47d7526b38065ccfbd7ac045624a

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
 
+## v4.4.0 - 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
+* :sparkles: add support for Invoices V4.10
+* :sparkles: add support for US Healthcare Cards V1.2
+
+
 ## v4.2.0 - 2025-03-28
 ### Changes
 * :coffin: remove support for US W9

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/docs/global_products/financial_document_v1.md

@@ -12,7 +12,7 @@ The Ruby Client Library supports the [Financial Document API](https://platform.m
 > | Specification                  | Details                                            |
 > | ------------------------------ | -------------------------------------------------- |
 > | Endpoint Name                  | `financial_document`                               |
-> | Recommended Version            | `v1.11`                                            |
+> | Recommended Version            | `v1.12`                                            |
 > | Supports Polling/Webhooks      | ✔️ Yes                                             |
 > | Support Synchronous HTTP Calls | ✔️ Yes                                             |
 > | Geography                      | 🌐 Global                                          |
@@ -93,7 +93,7 @@ puts result.document
 ########
 Document
 ########
-:Mindee ID: f469a24d-3875-4a83-ad43-e0d5aa9da604
+:Mindee ID: a80ac0ee-26f6-4e2e-988a-960b240d5ba7
 :Filename: default_sample.jpg
 
 Inference
@@ -118,7 +118,7 @@ Prediction
   +---------------+--------+----------+---------------+
   | Base          | Code   | Rate (%) | Amount        |
   +===============+========+==========+===============+
-  |               |        | 5.00     | 9.75          |
+  | 195.00        |        | 5.00     | 9.75          |
   +---------------+--------+----------+---------------+
 :Supplier Payment Details:
 :Supplier Name: JOHN SMITH
@@ -170,7 +170,7 @@ Page 0
   +---------------+--------+----------+---------------+
   | Base          | Code   | Rate (%) | Amount        |
   +===============+========+==========+===============+
-  |               |        | 5.00     | 9.75          |
+  | 195.00        |        | 5.00     | 9.75          |
   +---------------+--------+----------+---------------+
 :Supplier Payment Details:
 :Supplier Name: JOHN SMITH
@@ -382,6 +382,24 @@ puts result.document.inference.prediction.document_number.value
 puts result.document.inference.prediction.document_type.value
 ```
 
+## Document Type Extended
+**document_type_extended** ([ClassificationField](#classification-field)): Document type extended.
+
+#### Possible values include:
+ - 'CREDIT NOTE'
+ - 'INVOICE'
+ - 'OTHER'
+ - 'OTHER_FINANCIAL'
+ - 'PAYSLIP'
+ - 'PURCHASE ORDER'
+ - 'QUOTE'
+ - 'RECEIPT'
+ - 'STATEMENT'
+
+```rb
+puts result.document.inference.prediction.document_type_extended.value
+```
+
 ## Due Date
 **due_date** ([DateField](#date-field)): The date on which the payment is due.
 

data/docs/global_products/invoices_v4.md

@@ -12,7 +12,7 @@ The Ruby Client Library supports the [Invoice API](https://platform.mindee.com/m
 > | Specification                  | Details                                            |
 > | ------------------------------ | -------------------------------------------------- |
 > | Endpoint Name                  | `invoices`                                         |
-> | Recommended Version            | `v4.9`                                             |
+> | Recommended Version            | `v4.10`                                            |
 > | Supports Polling/Webhooks      | ✔️ Yes                                             |
 > | Support Synchronous HTTP Calls | ✔️ Yes                                             |
 > | Geography                      | 🌐 Global                                          |
@@ -93,7 +93,7 @@ puts result.document
 ########
 Document
 ########
-:Mindee ID: 86b1833f-138b-4a01-8387-860204b0e631
+:Mindee ID: b55db8f9-ae3b-4f05-b2f1-ec0ced5e5b70
 :Filename: default_sample.jpg
 
 Inference
@@ -117,7 +117,7 @@ Prediction
   +---------------+--------+----------+---------------+
   | Base          | Code   | Rate (%) | Amount        |
   +===============+========+==========+===============+
-  |               |        | 8.00     | 193.20        |
+  | 2145.00       |        | 8.00     | 193.20        |
   +---------------+--------+----------+---------------+
 :Supplier Payment Details:
 :Supplier Name: TURNPIKE DESIGNS
@@ -163,7 +163,7 @@ Page 0
   +---------------+--------+----------+---------------+
   | Base          | Code   | Rate (%) | Amount        |
   +===============+========+==========+===============+
-  |               |        | 8.00     | 193.20        |
+  | 2145.00       |        | 8.00     | 193.20        |
   +---------------+--------+----------+---------------+
 :Supplier Payment Details:
 :Supplier Name: TURNPIKE DESIGNS
@@ -344,6 +344,24 @@ puts result.document.inference.prediction.date.value
 puts result.document.inference.prediction.document_type.value
 ```
 
+## Document Type Extended
+**document_type_extended** ([ClassificationField](#classification-field)): Document type extended.
+
+#### Possible values include:
+ - 'CREDIT NOTE'
+ - 'INVOICE'
+ - 'OTHER'
+ - 'OTHER_FINANCIAL'
+ - 'PAYSLIP'
+ - 'PURCHASE ORDER'
+ - 'QUOTE'
+ - 'RECEIPT'
+ - 'STATEMENT'
+
+```rb
+puts result.document.inference.prediction.document_type_extended.value
+```
+
 ## Due Date
 **due_date** ([DateField](#date-field)): The date on which the payment is due.
 

data/docs/localized_products/us_healthcare_cards_v1.md

@@ -12,7 +12,7 @@ The Ruby Client Library supports the [Healthcare Card API](https://platform.mind
 > | Specification                  | Details                                            |
 > | ------------------------------ | -------------------------------------------------- |
 > | Endpoint Name                  | `us_healthcare_cards`                              |
-> | Recommended Version            | `v1.1`                                             |
+> | Recommended Version            | `v1.2`                                             |
 > | Supports Polling/Webhooks      | ✔️ Yes                                             |
 > | Support Synchronous HTTP Calls | ❌ No                                              |
 > | Geography                      | 🇺🇸 United States                                   |
@@ -138,6 +138,15 @@ A `HealthcareCardV1Copay` implements the following attributes:
 * `service_fees` (Float): The price of service.
 * `service_name` (String): The name of service of the copay.
 
+#### Possible values include:
+ - primary_care
+ - emergency_room
+ - urgent_care
+ - specialist
+ - office_visit
+ - prescription
+        
+
 # Attributes
 The following fields are extracted for Healthcare Card V1:
 
@@ -222,6 +231,13 @@ puts result.document.inference.prediction.rx_bin.value
 puts result.document.inference.prediction.rx_grp.value
 ```
 
+## RX ID
+**rx_id** ([StringField](#string-field)): The ID number for prescription drug coverage.
+
+```rb
+puts result.document.inference.prediction.rx_id.value
+```
+
 ## RX PCN
 **rx_pcn** ([StringField](#string-field)): The PCN number for prescription drug coverage.
 

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.

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