mindee 3.10.0 → 3.11.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: bdc0f8cc57ce32684b3da523ce0a6666a251ff90d7a6d894a7b4601231bed3e4
-  data.tar.gz: e249e12de6c86cc4a7683392504d5cd914afb941a99b5b00bff3f90fd0f4ae36
+  metadata.gz: cdf4542787e1ef24c74bf2104d3eacc5205afff3066ea24b0bea52b7d43d6993
+  data.tar.gz: 6d816b54bb34feb7d6c59db0f0a1f90cc34c5b6dd8155be315dfb698c88838d4
 SHA512:
-  metadata.gz: 713cf8cb6259b54637865f6799db1f5da7d7f27c436960316f84e2d404fe743d343d2d307feb2023c0fcab580421e5e0004bc70d34735d5800f0530d135c70ea
-  data.tar.gz: 535dc099fe76a27b3af26bb5cb726756e1b158056027782b607820e2648b92dbc32e65cf453d356c33355c5b94a3df9ffbbd86fabfc18fd5f439b48efa68816c
+  metadata.gz: 3ee7621e86789bf3f3ff294d505d71272018f342936d05f74f17a57e7b15916f8cde1218ffe1fa7ff7e7d2331c2982562725994bbf3565290645b870756d113e
+  data.tar.gz: 9b48e7088ffcc75136714e10e0fca34975b6b94f0f8a5241fd775fbcde0e8a05a836ca9f8703a7dc1d00c2820717ccc0cc989c8c1ea390716a3119114347e111

data/CHANGELOG.md

@@ -1,5 +1,10 @@
 # Mindee Ruby API Library Changelog
 
+## v3.11.0 - 2024-06-10
+### Changes
+* :sparkles: add custom tax extraction feature (#76)
+
+
 ## v3.10.0 - 2024-05-31
 ### Changes
 * :sparkles: add support for us mail v2 (#98)

data/lib/mindee/extraction/ocr_extractor.rb

@@ -0,0 +1,110 @@
+# frozen_string_literal: true
+
+module Mindee
+  module Extraction
+    # Generic extractor class
+    class OcrExtractor
+      # Checks for a list of possible matches in a string & returns the index of the first found candidate.
+      # Case & diacritics insensitive.
+      # @param text [String] string to search for matches.
+      # @param str_candidates [Array<String>] array of values to look for
+      # @return [Integer, nil]
+      def self.match_index(text, str_candidates)
+        idx = nil
+        str_candidates.each do |str_candidate|
+          found_idx = remove_accents(text.downcase).index(remove_accents(str_candidate.downcase))
+          idx = found_idx if idx.nil?
+          idx = found_idx if !found_idx.nil? && found_idx >= idx
+        end
+        idx
+      end
+
+      # Normalizes text by removing diacritics.
+      # @param input_str [String] string to handle.
+      # @return [String]
+      def self.remove_accents(input_str)
+        diacritics = [*0x1DC0..0x1DFF, *0x0300..0x036F, *0xFE20..0xFE2F].pack('U*')
+        input_str
+          .unicode_normalize(:nfd)
+          .tr(diacritics, '')
+          .unicode_normalize(:nfc)
+          .scrub
+      end
+
+      # Checks if a given percentage value is within the allowed range
+      # @param value [Integer] The value to check
+      # @param min_rate_percentage [Integer] Minimum allowed rate on the tax.
+      # @param max_rate_percentage [Integer] Maximum allowed rate on the tax.
+      # @return [Boolean]
+      def self.valid_percentage?(value, min_rate_percentage, max_rate_percentage)
+        return false if value.nil?
+
+        value > min_rate_percentage && value < max_rate_percentage
+      end
+
+      # Parses a percentage from a string, and returns it as a float.
+      # Returns nil if candidate isn't a valid percentage.
+      # @param percentage_str [String] String candidate.
+      # @return [Float, nil]
+      def self.parse_percentage(percentage_str)
+        percentage_str.gsub!('%', '')
+        percentage_str.strip
+        percentage_str.gsub!(',', '.')
+        Float(percentage_str.scrub)
+      rescue ArgumentError
+        nil
+      end
+
+      # Parses an amount from a string, and returns it as a float.
+      # Returns nil if candidate isn't a valid amount.
+      # @param amount_str [String] String candidate.
+      # @return [Float, nil]
+      def self.parse_amount(amount_str)
+        cleaned_str = amount_str.gsub(' ', '')
+        cleaned_str = standardize_delimiters(cleaned_str)
+        Float(cleaned_str)
+      rescue ArgumentError
+        nil
+      end
+
+      private
+
+      def self.standardize_delimiters(str)
+        if comma_decimal?(str)
+          str.gsub('.', '').gsub(',', '.')
+        elsif dot_decimal?(str)
+          str.gsub(',', '')
+        else
+          str
+        end
+      end
+
+      def self.comma_decimal?(str)
+        (str.length > 3 && str[-3] == ',') || str[-2] == ','
+      end
+
+      def self.dot_decimal?(str)
+        (str.length > 3 && str[-3] == '.') || str[-2] == '.'
+      end
+
+      # Removes most common currency symbols from string
+      # @param input_string [String] string to remove the symbols from
+      # @return [String]
+      def self.remove_currency_symbols(input_string)
+        # Define an array of common currency symbols
+        currency_symbols = ['$', '€', '£', '¥', '₹', '₽', '฿', '₺', '₴', '₿', '₡', '₮', '₱', '₲', '₪', '₫', '₩', '₵',
+                            '₦', '₢', '₤', '₣', '₧', '₯', '₠', '₶', '₸', '₷', '₼', '₾', '₺', '﹩', '₨', '₹', '＄', '﹫']
+
+        # Iterate over each currency symbol and remove it from the input string
+        currency_symbols.each do |symbol|
+          input_string.gsub!(symbol, '')
+        end
+
+        input_string
+      end
+
+      private_class_method :remove_accents, :match_index, :parse_amount, :parse_percentage, :remove_currency_symbols,
+                           :valid_percentage?, :comma_decimal?, :dot_decimal?, :standardize_delimiters
+    end
+  end
+end

data/lib/mindee/extraction/tax_extractor.rb

@@ -0,0 +1,322 @@
+# frozen_string_literal: true
+
+require_relative 'ocr_extractor'
+
+module Mindee
+  module Extraction
+    # Tax extractor class
+    class TaxExtractor < OcrExtractor
+      # Extracts the most relevant candidate.
+      # @param candidates [Array<Hash>] a candidate for the tax.
+      # @param tax_names [Array<String>] list of all possible names the tax can have.
+      # @return [Hash, nil]
+      def self.pick_best(candidates, tax_names)
+        return candidates[0] if candidates.size == 1
+        return nil if candidates.empty?
+
+        picked = 0
+        picked_score = 0
+
+        candidates.each_with_index do |candidate, i|
+          next unless valid_candidate?(candidate, tax_names)
+
+          sum_fields_score = calculate_score(candidate, i)
+
+          if picked_score < sum_fields_score
+            picked_score = sum_fields_score
+            picked = i
+          end
+        end
+
+        candidates[picked]
+      end
+
+      # Checks whether a tax code has been properly read. Shouldn't trigger except in case of very specific regex breaks
+      # due to unsupported diacritics.
+      # @param candidate [Hash] A candidate for the tax.
+      # @param tax_names [Array<String>] list of all possible names the tax can have.
+      # @return [Boolean]
+      def self.valid_candidate?(candidate, tax_names)
+        return false if tax_names.empty? || candidate.nil? || candidate['code'].nil?
+
+        tax_names.each do |tax_name|
+          return true if remove_accents(tax_name.downcase) == remove_accents(candidate['code'].downcase)
+        end
+        false
+      end
+
+      # [Experimental] computes the score of a valid candidate for a tax.
+      # @param candidate [Hash] A candidate for the tax.
+      # @param index [Integer]
+      def self.calculate_score(candidate, index)
+        score = index + 1
+        unless candidate['rate'].nil?
+          score += 1
+          score -= 2 if candidate['rate'] > 100
+          score -= 1 if candidate['rate'] > 30
+        end
+        score += 4 unless candidate['value'].nil?
+        score += 1 unless candidate['base'].nil?
+        score
+      end
+
+      # Curates tax values based on simple rules to avoid improbable data
+      # @param found_hash [Hash] Hash of currently retrieved values
+      # @param min_rate_percentage [Integer] Minimum allowed rate on the tax.
+      # @param max_rate_percentage [Integer] Maximum allowed rate on the tax.
+      # @return [Hash]
+      def self.curate_values(found_hash, min_rate_percentage, max_rate_percentage)
+        reconstructed_hash = { 'code' => nil, 'page_id' => nil, 'rate' => nil, 'base' => nil, 'value' => nil }
+        return reconstructed_hash if found_hash.nil?
+
+        reconstructed_hash['code'] =
+          found_hash['code'].nil? ? found_hash['code'] : found_hash['code'].sub(%r{\s*\.*\s*$}, '')
+
+        if found_hash['rate'] && found_hash['rate'] < 1 && (found_hash['rate']).positive?
+          found_hash['rate'] =
+            found_hash['rate'] * 100
+        end
+        found_hash = swap_rates_if_needed(found_hash, min_rate_percentage, max_rate_percentage)
+        found_hash = decimate_rates_if_needed(found_hash)
+        found_hash = fix_rate(found_hash)
+        reconstructed_hash['rate'] = found_hash['rate']
+        set_base_and_value(reconstructed_hash, found_hash)
+      end
+
+      # Swaps the rate with base or value if rate is out of bounds
+      # @param found_hash [Hash] Hash of currently retrieved values
+      # @param min_rate_percentage [Integer] Minimum allowed rate on the tax.
+      # @param max_rate_percentage [Integer] Maximum allowed rate on the tax.
+      # @return [Hash]
+      def self.swap_rates_if_needed(found_hash, min_rate_percentage, max_rate_percentage)
+        if found_hash['rate'] && (found_hash['rate'] > max_rate_percentage || found_hash['rate'] < min_rate_percentage)
+          if valid_percentage?(found_hash['base'], min_rate_percentage, max_rate_percentage)
+            found_hash['rate'], found_hash['base'] = found_hash['base'], found_hash['rate']
+          elsif valid_percentage?(found_hash['value'], min_rate_percentage, max_rate_percentage)
+            found_hash['rate'], found_hash['value'] = found_hash['value'], found_hash['rate']
+          end
+        end
+        found_hash
+      end
+
+      # Rates can't be negative if set.
+      # @param found_hash [Hash] Hash of currently retrieved values
+      def self.fix_rate(found_hash)
+        found_hash['rate'] = found_hash['rate'].abs unless found_hash['rate'].nil?
+        found_hash
+      end
+
+      # Swaps the rate with base or value if rate is out of bounds
+      # @param found_hash [Hash] Hash of currently retrieved values
+      # @return [Hash]
+      def self.decimate_rates_if_needed(found_hash)
+        if found_hash['rate'] && found_hash['rate'] > 100
+          if !found_hash['base'].nil? && found_hash['rate'] > found_hash['base']
+            found_hash['rate'], found_hash['base'] = found_hash['base'], found_hash['rate']
+          elsif !found_hash['value'].nil? && found_hash['rate'] > found_hash['value']
+            found_hash['rate'], found_hash['value'] = found_hash['value'], found_hash['rate']
+          end
+        end
+        found_hash
+      end
+
+      # Sets the base and value in the reconstructed hash based on certain conditions
+      # @param reconstructed_hash [Hash] Hash being reconstructed with new values
+      # @param found_hash [Hash] Hash of currently retrieved values
+      # @return [Hash]
+      def self.set_base_and_value(reconstructed_hash, found_hash)
+        if found_hash['base'].nil?
+          reconstructed_hash['base'] = found_hash['base']
+          reconstructed_hash['value'] = found_hash['value']
+        elsif found_hash['value'].nil? && found_hash['base'] < found_hash['value']
+          reconstructed_hash['base'] = found_hash['value']
+          reconstructed_hash['value'] = found_hash['base']
+        else
+          reconstructed_hash['value'] = found_hash['value']
+        end
+        reconstructed_hash
+      end
+
+      # Extracts a single custom type of tax.
+      # For the sake of simplicity, this only extracts the first example, unless specifically instructed otherwise.
+      # @param ocr_result [Mindee::Parsing::Common::Ocr::Ocr] result of the OCR.
+      # @param tax_names [Array<String>] list of all possible names the tax can have.
+      # @param min_rate_percentage [Integer] Minimum allowed rate on the tax.
+      # @param max_rate_percentage [Integer] Maximum allowed rate on the tax.
+      # @return [Mindee::Parsing::Standard::TaxField, nil]
+      def self.extract_custom_tax(ocr_result, tax_names, min_rate_percentage = 0, max_rate_percentage = 100)
+        return nil if ocr_result.is_a?(Mindee::Parsing::Common::Ocr) || tax_names.empty?
+
+        tax_names.sort!
+        found_hash = pick_best(extract_horizontal_tax(ocr_result, tax_names), tax_names)
+        # a tax is considered found horizontally if it has a value, otherwise it is vertical
+        if found_hash.nil? || found_hash['value'].nil?
+          found_hash = extract_vertical_tax(ocr_result, tax_names,
+                                            found_hash)
+        end
+        found_hash = curate_values(found_hash, min_rate_percentage, max_rate_percentage)
+
+        return if found_hash.nil? || found_hash.empty?
+
+        create_tax_field(found_hash)
+      end
+
+      # Creates a tax field from a given hash.
+      # @param found_hash [Hash] Hash of currently retrieved values
+      # @return [Mindee::Parsing::Standard::TaxField]
+      def self.create_tax_field(found_hash)
+        Mindee::Parsing::Standard::TaxField.new(
+          found_hash,
+          found_hash.key?('page_id') ? found_hash['page_id'] : nil
+        )
+      end
+
+      # Extracts the rate and code, if found, from matches into the found_hash.
+      # @param matches [MatchData] RegEx matches of the values for taxes
+      # @param found_hash [Hash] Hash of currently retrieved values
+      # @param percent_first [Boolean] Whether the percentage was found before or after the tax name.
+      # @return [Hash]
+      def self.extract_percentage_from_tax(matches, found_hash, percent_first)
+        if percent_first
+          found_hash['code'] = matches[2].strip unless matches[2].nil?
+          found_hash['rate'] = parse_amount(matches[1].gsub('%', '')) unless matches[1].nil?
+        else
+          found_hash['code'] = matches[1].strip unless matches[1].nil?
+          found_hash['rate'] = parse_amount(matches[2].gsub('%', '')) unless matches[2].nil?
+        end
+        found_hash
+      end
+
+      # rubocop:disable Metrics/CyclomaticComplexity
+      # rubocop:disable Metrics/PerceivedComplexity
+
+      # Extracts the basis and value of a tax from regex matches, independent of the order.
+      # @param matches [MatchData] RegEx matches of the values for taxes
+      # @param found_hash [Hash] Hash of currently retrieved values
+      # @return [Hash]
+      def self.extract_basis_and_value(matches, found_hash)
+        if matches[4].nil? && !matches[3].nil?
+          found_hash['value'] = parse_amount(matches[3]) unless matches[3].nil?
+        elsif matches[3].nil? && !matches[4].nil?
+          found_hash['value'] = parse_amount(matches[4]) unless matches[4].nil?
+        elsif !matches[3].nil? && !matches[4].nil?
+          found_hash['base'] = parse_amount(matches[3]) unless matches[3].nil?
+          found_hash['value'] = parse_amount(matches[4]) unless matches[4].nil?
+        end
+        found_hash
+      end
+
+      # rubocop:enable Metrics/CyclomaticComplexity
+      # rubocop:enable Metrics/PerceivedComplexity
+
+      # Extracts tax information from a horizontal line.
+      # @param line [String] Line to be processed.
+      # @param pattern [Regexp] RegEx pattern to search the line with.
+      # @param percent_first [Boolean] Whether the percentage was found before or after the tax name.
+      # @return [Hash]
+      def self.extract_tax_from_horizontal_line(line, pattern, page_id, percent_first)
+        found_hash = {}
+
+        matches = line.match(pattern)
+
+        # Edge case for when the tax is split-up between two pages, we'll consider that
+        # the answer belongs to the first one.
+        found_hash['page_id'] = page_id unless found_hash.key?('page_id')
+        return found_hash if matches.nil?
+
+        found_hash = extract_percentage_from_tax(matches, found_hash, percent_first)
+        extract_basis_and_value(matches, found_hash)
+      end
+
+      # rubocop:disable Metrics/CyclomaticComplexity
+      # rubocop:disable Metrics/PerceivedComplexity
+
+      # Processes a horizontal line for tax extraction. Returns a hash of collected values.
+      # @param ocr_result [Mindee::Parsing::Common::Ocr::Ocr] Processed OCR results.
+      # @param tax_names [Array<String>] Possible tax names candidates.
+      # @return [Array<Hash>]
+      def self.extract_horizontal_tax(ocr_result, tax_names)
+        candidates = [{ 'code' => nil, 'value' => nil, 'base' => nil, 'rate' => nil }]
+        linear_pattern_percent_first = %r{
+          ((?:\s*-\s*)?(?:\d*[.,])*\d+[ ]?%?|%?[ ]?(?:\s*-\s*)?(?:\d*[.,])*\d+)?[ .]?
+          ([a-zA-ZÀ-ÖØ-öø-ÿ .]*[a-zA-ZÀ-ÖØ-öø-ÿ]?)[ .]?
+          ((?:\s*-\s*)?(?:\d*[.,])+\d{2,})?[ .]*
+          ((?:\s*-\s*)?(\d*[.,])*\d{2,})?
+        }x
+        linear_pattern_percent_second = %r{
+          ([a-zA-ZÀ-ÖØ-öø-ÿ .]*[a-zA-ZÀ-ÖØ-öø-ÿ]?)[ .]*
+          ((?:\s*-\s*)?(?:\d*[.,])*\d+[ ]?%?|%?[ ]?(?:\s*-\s*)?(?:\d*[.,])*\d+)?[ .]?
+          ((?:\s*-\s*)?(?:\d*[.,])+\d{2,})?[ .]*
+          ((?:\s*-\s*)?(\d*[.,])*\d{2,})?
+        }x
+        ocr_result.mvision_v1.pages.each.with_index do |page, page_id|
+          page.all_lines.each do |line|
+            clean_line = remove_currency_symbols(line.to_s.scrub.gsub(%r{[+(\[)\]¿?*_]}, '')).gsub(%r{\.{2,}}, ' ')
+                                                                                             .gsub(%r{ +}, ' ').strip
+
+            next if match_index(clean_line, tax_names).nil?
+
+            unless clean_line.match(linear_pattern_percent_second).nil?
+              candidates.append(extract_tax_from_horizontal_line(clean_line[match_index(clean_line, tax_names)..],
+                                                                 linear_pattern_percent_second, page_id, false))
+            end
+            if clean_line.include?('%') && !clean_line.match(linear_pattern_percent_first).nil?
+              candidates.append(extract_tax_from_horizontal_line(clean_line[clean_line.index(%r{\d*[.,]?\d* ?%})..],
+                                                                 linear_pattern_percent_first, page_id, true))
+            elsif !clean_line.match(linear_pattern_percent_first).nil?
+              candidates.append(extract_tax_from_horizontal_line(clean_line,
+                                                                 linear_pattern_percent_first, page_id, true))
+            end
+          end
+        end
+        candidates
+      end
+      # rubocop:enable Metrics/CyclomaticComplexity
+      # rubocop:enable Metrics/PerceivedComplexity
+
+      # Processes a vertical reconstructed line for tax extraction. Returns a hash of collected values.
+      # @param line [Mindee::Parsing::Common::Ocr::OcrLine] Processed OCR results.
+      # @param found_hash [Hash] Hash containing previously found values, if any.
+      # @return [Hash]
+      def self.extract_vertical_tax_values(line, found_hash)
+        amounts = []
+        line.each do |reconstructed_word|
+          amounts.push(parse_amount(reconstructed_word.text)) unless parse_amount(reconstructed_word.text).nil?
+        end
+        if amounts.length == 1 && !found_hash.key?('value')
+          found_hash['value'] = amounts[0]
+        else
+          found_hash['rate'] = amounts[0] if found_hash['rate'].nil?
+          found_hash['value'] = amounts[1] if found_hash['value'].nil?
+        end
+        found_hash
+      end
+
+      # Extracts tax data from a vertical reconstructed row.
+      # @param ocr_result [Mindee::Parsing::Common::Ocr] OCR raw results
+      # @param tax_names [Array<String>] Array of possible names a tax can have
+      # @param found_hash [Hash] Hash of currently retrieved values
+      def self.extract_vertical_tax(ocr_result, tax_names, found_hash)
+        found_hash = { 'code' => nil, 'page_id' => nil } if found_hash.nil?
+
+        ocr_result.mvision_v1.pages.each_with_index do |page, page_id|
+          page.all_words.each do |word|
+            next if match_index(word.text, tax_names).nil?
+
+            reconstructed_line = ocr_result.reconstruct_vertically(word.polygon, page_id)
+            found_hash['page_id'] = page_id if found_hash['page_id'].nil?
+            found_hash['code'] = word.text.strip if found_hash['code'].nil?
+            found_hash = extract_vertical_tax_values(reconstructed_line, found_hash)
+          end
+        end
+        found_hash
+      end
+
+      private_class_method :extract_percentage_from_tax, :extract_basis_and_value, :extract_tax_from_horizontal_line,
+                           :extract_horizontal_tax, :extract_vertical_tax_values, :extract_vertical_tax,
+                           :create_tax_field, :fix_rate, :pick_best, :calculate_score, :curate_values,
+                           :decimate_rates_if_needed, :extract_basis_and_value, :set_base_and_value, :valid_candidate?,
+                           :swap_rates_if_needed
+    end
+  end
+end

data/lib/mindee/extraction.rb

@@ -0,0 +1,3 @@
+# frozen_string_literal: true
+
+require_relative 'extraction/tax_extractor'

data/lib/mindee/geometry/utils.rb

@@ -77,5 +77,24 @@ module Mindee
       coords = points.map(&:x)
       MinMax.new(coords.min, coords.max)
     end
+
+    # Checks whether a set of coordinates is below another on the page, with a slight margin for the lateral value.
+    # @param candidate [Array<Mindee::Geometry::Point] Polygon to check
+    # @param anchor [Array<Mindee::Geometry::Point] Reference polygon
+    # @param margin_left [Float] Margin tolerance on the left of the anchor
+    # @param margin_right [Float] Margin tolerance on the right of the anchor
+    def self.below?(candidate, anchor, margin_left, margin_right)
+      return false if Geometry.get_min_max_y(candidate).min < Geometry.get_min_max_y(anchor).min
+      if Geometry.get_min_max_x(candidate).min <
+         Geometry.get_min_max_x(anchor).min - (Geometry.get_min_max_x(anchor).min * margin_left)
+        return false
+      end
+      if Geometry.get_min_max_x(candidate).max >
+         Geometry.get_min_max_x(anchor).max + (Geometry.get_min_max_x(anchor).max * margin_right)
+        return false
+      end
+
+      true
+    end
   end
 end

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

@@ -100,7 +100,7 @@ module Mindee
         attr_reader :job
         # @return [Mindee::Parsing::Common::ApiRequest]
         attr_reader :api_request
-        # @return [Hash]
+        # @return [String]
         attr_reader :raw_http
 
         # @param product_class [Class<Mindee::Product>]

data/lib/mindee/parsing/common/ocr/mvision_v1.rb

@@ -27,6 +27,22 @@ module Mindee
             end
             out_str.strip
           end
+
+          # Constructs a line from a column, located underneath given coordinates
+          # @param coordinates [Array<Mindee::Geometry::Point>] Polygon or bounding box where the reconstruction should
+          # start.
+          # @param page_id [Integer] ID of the page to start at
+          # @param x_margin [Float] Margin of misalignment for the x coordinate.
+          # @return [Mindee::Parsing::Common::Ocr::OcrLine]
+          def reconstruct_vertically(coordinates, page_id, x_margin)
+            line_arr = OcrLine.new([])
+            @pages[page_id].all_lines.each do |line|
+              line.each do |word|
+                line_arr.push(word) if Geometry.below?(word.polygon, coordinates, x_margin / 2, x_margin * 2)
+              end
+            end
+            line_arr
+          end
         end
       end
     end

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

@@ -163,6 +163,16 @@ module Mindee
           def to_s
             @mvision_v1.to_s
           end
+
+          # Constructs a line from a column, located underneath given coordinates
+          # @param coordinates [Array<Mindee::Geometry::Point>] Polygon or bounding box where the reconstruction should
+          # start
+          # @param page_id [Integer] ID of the page to start at
+          # @param x_margin [Float] Margin of misalignment for the x coordinate (default 10%)
+          # @return [Mindee::Parsing::Common::Ocr::OcrLine]
+          def reconstruct_vertically(coordinates, page_id, x_margin = 0.05)
+            @mvision_v1.reconstruct_vertically(coordinates, page_id, x_margin)
+          end
         end
       end
     end

data/lib/mindee/version.rb

@@ -3,7 +3,7 @@
 # Mindee
 module Mindee
   # Current version.
-  VERSION = '3.10.0'
+  VERSION = '3.11.0'
 
   # Finds and return the current platform.
   # @return [String]

data/lib/mindee.rb

@@ -1,6 +1,7 @@
 # frozen_string_literal: true
 
 require 'mindee/client'
+require 'mindee/extraction'
 
 module Mindee
   # Mindee internal http module.
@@ -18,6 +19,10 @@ module Mindee
     end
   end
 
+  # Custom extraction module
+  module Extraction
+  end
+
   # Parsing internals and fields.
   module Parsing
     # Common fields and functions.

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: mindee
 version: !ruby/object:Gem::Version
-  version: 3.10.0
+  version: 3.11.0
 platform: ruby
 authors:
 - Mindee, SA
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2024-05-31 00:00:00.000000000 Z
+date: 2024-06-10 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: marcel
@@ -173,6 +173,9 @@ files:
 - docs/us_w9_v1.md
 - lib/mindee.rb
 - lib/mindee/client.rb
+- lib/mindee/extraction.rb
+- lib/mindee/extraction/ocr_extractor.rb
+- lib/mindee/extraction/tax_extractor.rb
 - lib/mindee/geometry.rb
 - lib/mindee/geometry/min_max.rb
 - lib/mindee/geometry/point.rb