steuer 1.0.3 → 1.1.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 932dd3009320b708e4141a8e90092717d30efc0792550dcaf2161ab16f4bcc7a
-  data.tar.gz: '0219a890034b58541a5f797fa62ab6174186ff7be964c0946bacda4a122f3272'
+  metadata.gz: ecfb962a41af7819208aaf3938a9534f9c80b17ccb596e96dfaf0fb02a267d15
+  data.tar.gz: 7a6c99cd8e357c641d55dfc3be85e9ada4425b62b00e12e0890b5510c0f66e94
 SHA512:
-  metadata.gz: 8d978823b3e9a45da0f859f2acdf1a0bcaebc0b67c94f9d41229f6be5dc1a73c150806facbcf9286661c180f46cc537c038c2c077c46ca989221b2def85a2ee9
-  data.tar.gz: a560bef4902b511e2befb2b6ffe367ffae2758ecc7cea147a05fbde419561737dfa7ebc0d64032f3ca776ab6c367a044177f8045fb565b98bc1376519354c252
+  metadata.gz: 1ef79e5aa9d45bd9d379d39156c2c3291c005c078f66cded75fe22a896d03e5b00e95fd2826c871e272f81651cc7eddde287f8f7c170daea0df711795f1c6e6c
+  data.tar.gz: 309aabc613b97fe54337e86f0620b90356b1bd25859cdd2a5e6b352cd4b0509481b2cc2f1091719cbb3250588c04b264d44098a0b89d1fd66c15f48a469a8e23

data/lib/steuer/state_mapping.rb

@@ -119,6 +119,66 @@ module Steuer
       },
     }.freeze
     class << self
+      def standard_group_sizes(state_code)
+        config = STATES[state_code]
+        return unless config
+
+        sizes = config[:standard_pattern].source.delete_prefix('^').delete_suffix('$')
+          .split('/')
+          .map { |group| standard_group_size(group) }
+
+        return if sizes.any?(&:nil?) || sizes.empty?
+
+        sizes
+      end
+
+      # Ranges such as Hessen's 0(1[3-9]) and Saarland's (01[0-2]) stand in
+      # for single digits, so groups are measured rather than read off \d{n}.
+      def standard_group_size(group)
+        size = 0
+        scanner = group.dup
+        until scanner.empty?
+          case scanner
+          when /\A\\d\{(\d+)\}/
+            size += Regexp.last_match(1).to_i
+            scanner = scanner.sub(/\A\\d\{\d+\}/, '')
+          when /\A\[[^\]]*\]/
+            size += 1
+            scanner = scanner.sub(/\A\[[^\]]*\]/, '')
+          when /\A\\d/
+            size += 1
+            scanner = scanner.delete_prefix('\\d')
+          when /\A\d/
+            size += 1
+            scanner = scanner[1..]
+          when /\A[()]/
+            scanner = scanner[1..]
+          else
+            return
+          end
+        end
+        size
+      end
+
+      # Separators are presentation only: Baden-Württemberg prints 93815/08152
+      # for the ten digits the BZSt Standardschema groups as 93/815/08152.
+      def canonicalize_standard_format(tax_number, state_code)
+        sizes = standard_group_sizes(state_code)
+        return unless sizes
+
+        digits = tax_number.gsub(/\D/, '')
+        return unless digits.length == sizes.sum
+
+        offset = 0
+        groups = sizes.map do |size|
+          part = digits[offset, size]
+          offset += size
+          part
+        end
+
+        groups.join('/')
+      end
+
       def find_state_by_standard_format(tax_number)
         # Try more specific patterns first to avoid conflicts with general patterns
         # Priority: character classes with ranges > specific digit patterns > general digit patterns

data/lib/steuer/steuernummer.rb

@@ -8,7 +8,8 @@ module Steuer
 
     def initialize(tax_number, state: nil)
       @original_input = tax_number.to_s.strip
-      @state_code = normalize_state_code(state) || auto_detect_state
+      @explicit_state_code = normalize_state_code(state)
+      @state_code = @explicit_state_code || auto_detect_state
 
       validate_tax_number!
     end
@@ -61,7 +62,7 @@ module Steuer
 
       case format_type
       when :standard
-        @original_input
+        canonical_standard_input || @original_input
       when :federal_12
         convert_federal_12_to_standard
       when :federal_13
@@ -82,6 +83,8 @@ module Steuer
         :federal_12
       elsif normalized_input.length == 13 && normalized_input.match?(/^\d{13}$/)
         :federal_13
+      elsif unseparated_standard?
+        :standard
       end
 
       @format_type
@@ -178,7 +181,24 @@ module Steuer
 
     def validate_standard_format
       config = StateMapping::STATES[@state_code]
-      normalized_input.match?(config[:standard_pattern])
+      return true if normalized_input.match?(config[:standard_pattern])
+
+      canonical = canonical_standard_input
+      !canonical.nil? && canonical.match?(config[:standard_pattern])
+    end
+
+    def unseparated_standard?
+      return false unless @explicit_state_code
+      return false unless normalized_input.match?(/\A\d+\z/)
+
+      sizes = StateMapping.standard_group_sizes(@explicit_state_code)
+      !sizes.nil? && normalized_input.length == sizes.sum
+    end
+
+    def canonical_standard_input
+      return unless @state_code
+
+      StateMapping.canonicalize_standard_format(normalized_input, @state_code)
     end
 
     def validate_federal_12_format
@@ -200,7 +220,8 @@ module Steuer
     def convert_standard_to_federal_12
       config = StateMapping::STATES[@state_code]
 
-      match = normalized_input.match(config[:standard_pattern])
+      match = normalized_input.match(config[:standard_pattern]) ||
+        canonical_standard_input&.match(config[:standard_pattern])
       return unless match
 
       parts = match.captures

data/lib/steuer/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Steuer
-  VERSION = '1.0.3'
+  VERSION = '1.1.0'
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: steuer
 version: !ruby/object:Gem::Version
-  version: 1.0.3
+  version: 1.1.0
 platform: ruby
 authors:
 - Olumuyiwa Osiname
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2025-12-23 00:00:00.000000000 Z
+date: 2026-09-04 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: pry