i18n-context-generator 0.5.0 → 0.5.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 6588c09e273cb74c40dba8cc26798e46b82b5d47e5c840f6e4609843bb1f5ed8
-  data.tar.gz: 9b3e309fd4030a99e065a2c7693240f09d205f9d95323a9744ecb3eaa1b239b1
+  metadata.gz: 2109020f4ff761dcf5fdd9817499b551619dc8ee17cef9e88acc881ac5ad647f
+  data.tar.gz: 26c9e86a187bbe4f2a0523754a7ee78912d96d9e532e261143ce905106cfd3da
 SHA512:
-  metadata.gz: 2a7b9df974f135375341c7c1a5f73e40575f10c92d11a72ecc976ecef9ae6b25aac32351f1468c79671c870d87c2df576664a96b32ef00128815fc4ff016de1c
-  data.tar.gz: 0cc9db25973ffe209dea8cb4197df899310320361f0a8da2e45fa597fc592867bea22cdd5ad5337907c18cbdf824e269955fe79550baedd23f0d919fdef73c00
+  metadata.gz: ed14b199d897b8ee27dbd793c372df01b72efc4a9b6a1122cd9fa724abb76030b93e1c967f82c02c5fafb4e104fba3b8abe10b9b1d672d95521b9b9f04d6b555
+  data.tar.gz: e13a1a8d4c17823597727717afb4f5950d0790109c35a124262989504123165a2111482fe9d27bf9baff3de930c9bc8fb6159edb1dee1ed9a8ad863beb6c346c

data/lib/i18n_context_generator/llm/client.rb

@@ -38,8 +38,18 @@ module I18nContextGenerator
         required: %w[description ui_element tone max_length confidence ambiguity_reason],
         properties: {
           description: { type: 'string' },
-          ui_element: { type: %w[string null], enum: UI_ELEMENTS + [nil] },
-          tone: { type: %w[string null], enum: TONES + [nil] },
+          ui_element: {
+            anyOf: [
+              { type: 'string', enum: UI_ELEMENTS },
+              { type: 'null' }
+            ]
+          },
+          tone: {
+            anyOf: [
+              { type: 'string', enum: TONES },
+              { type: 'null' }
+            ]
+          },
           max_length: { type: %w[integer null] },
           confidence: { type: 'string', enum: CONFIDENCE_LEVELS },
           ambiguity_reason: {

data/lib/i18n_context_generator/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module I18nContextGenerator
-  VERSION = '0.5.0'
+  VERSION = '0.5.1'
 end

data/lib/i18n_context_generator/xcstrings_document.rb

@@ -9,6 +9,25 @@ module I18nContextGenerator
   # lets diff handling map revision-specific line numbers without reformatting.
   class XcstringsDocument
     Member = Data.define(:key, :key_start, :key_end, :value_start, :value_end, :separator)
+    BYTE = {
+      backslash: '\\'.ord,
+      close_array: ']'.ord,
+      close_object: '}'.ord,
+      colon: ':'.ord,
+      comma: ','.ord,
+      newline: "\n".ord,
+      open_array: '['.ord,
+      open_object: '{'.ord,
+      quote: '"'.ord
+    }.freeze
+    JSON_WHITESPACE_BYTES = ["\t".ord, "\n".ord, "\r".ord, ' '.ord].freeze
+    PRIMITIVE_TERMINATOR_BYTES = [
+      BYTE[:comma],
+      BYTE[:close_object],
+      BYTE[:close_array],
+      *JSON_WHITESPACE_BYTES
+    ].freeze
+    private_constant :BYTE, :JSON_WHITESPACE_BYTES, :PRIMITIVE_TERMINATOR_BYTES
 
     attr_reader :catalog
 
@@ -17,10 +36,11 @@ module I18nContextGenerator
       @path = path
       @catalog = Oj.load(content, mode: :strict)
       @line_starts = [0]
-      content.each_char.with_index { |character, index| @line_starts << (index + 1) if character == "\n" }
+      content.each_byte.with_index { |byte, index| @line_starts << (index + 1) if byte == BYTE[:newline] }
       @root_members = object_members(skip_whitespace(0))
       strings_member = @root_members.find { |member| member.key == 'strings' }
-      raise TypeError, 'strings must be a mapping' unless strings_member && @content[strings_member.value_start] == '{'
+      strings_object = strings_member && byte_at(strings_member.value_start) == BYTE[:open_object]
+      raise TypeError, 'strings must be a mapping' unless strings_object
 
       @strings_member = strings_member
       @entry_members = object_members(strings_member.value_start)
@@ -40,18 +60,19 @@ module I18nContextGenerator
     def with_comments(comments_by_key)
       edits = comments_by_key.filter_map do |key, comment|
         entry_member = @entries_by_key[key]
-        next unless entry_member && @content[entry_member.value_start] == '{'
+        next unless entry_member && byte_at(entry_member.value_start) == BYTE[:open_object]
         next if @catalog.dig('strings', key, 'comment') == comment
 
         comment_edit(entry_member, comment)
       end
       return @content if edits.empty?
 
-      rendered = @content.dup
+      original_encoding = @content.encoding
+      rendered = @content.b
       edits.sort_by(&:first).reverse_each do |start_offset, end_offset, replacement|
-        rendered[start_offset...end_offset] = replacement
+        rendered[start_offset...end_offset] = replacement.b
       end
-      rendered
+      rendered.force_encoding(original_encoding)
     end
 
     private
@@ -109,8 +130,9 @@ module I18nContextGenerator
     end
 
     def line_indent(offset)
-      line_start = @content.rindex("\n", offset - 1)
-      @content[(line_start ? line_start + 1 : 0)...offset]
+      following_line = @line_starts.bsearch_index { |line_start| line_start > offset }
+      line_start = following_line ? @line_starts.fetch(following_line - 1) : @line_starts.last
+      byte_slice(line_start, offset)
     end
 
     def line_number(offset)
@@ -118,18 +140,18 @@ module I18nContextGenerator
     end
 
     def object_members(object_start)
-      raise TypeError, "expected object at byte #{object_start}" unless @content[object_start] == '{'
+      raise TypeError, "expected object at byte #{object_start}" unless byte_at(object_start) == BYTE[:open_object]
 
       members = []
       cursor = skip_whitespace(object_start + 1)
-      return members if @content[cursor] == '}'
+      return members if byte_at(cursor) == BYTE[:close_object]
 
       loop do
         key_start = cursor
         key_end = string_end(key_start)
-        key = JSON.parse(@content[key_start...key_end])
+        key = JSON.parse(byte_slice(key_start, key_end))
         cursor = skip_whitespace(key_end)
-        raise TypeError, "expected ':' at byte #{cursor}" unless @content[cursor] == ':'
+        raise TypeError, "expected ':' at byte #{cursor}" unless byte_at(cursor) == BYTE[:colon]
 
         cursor = skip_whitespace(cursor + 1)
         value_start = cursor
@@ -140,12 +162,12 @@ module I18nContextGenerator
           key_end: key_end,
           value_start: value_start,
           value_end: value_end,
-          separator: @content[key_end...value_start]
+          separator: byte_slice(key_end, value_start)
         )
         cursor = skip_whitespace(value_end)
-        break if @content[cursor] == '}'
+        break if byte_at(cursor) == BYTE[:close_object]
 
-        raise TypeError, "expected ',' at byte #{cursor}" unless @content[cursor] == ','
+        raise TypeError, "expected ',' at byte #{cursor}" unless byte_at(cursor) == BYTE[:comma]
 
         cursor = skip_whitespace(cursor + 1)
       end
@@ -154,13 +176,13 @@ module I18nContextGenerator
     end
 
     def value_end(start_offset)
-      case @content[start_offset]
-      when '"'
+      case byte_at(start_offset)
+      when BYTE[:quote]
         string_end(start_offset)
-      when '{'
-        collection_end(start_offset, '{', '}')
-      when '['
-        collection_end(start_offset, '[', ']')
+      when BYTE[:open_object]
+        collection_end(start_offset, BYTE[:open_object], BYTE[:close_object])
+      when BYTE[:open_array]
+        collection_end(start_offset, BYTE[:open_array], BYTE[:close_array])
       else
         primitive_end(start_offset)
       end
@@ -169,13 +191,13 @@ module I18nContextGenerator
     def string_end(start_offset)
       cursor = start_offset + 1
       escaped = false
-      while cursor < @content.length
-        character = @content[cursor]
+      while cursor < @content.bytesize
+        byte = byte_at(cursor)
         if escaped
           escaped = false
-        elsif character == '\\'
+        elsif byte == BYTE[:backslash]
           escaped = true
-        elsif character == '"'
+        elsif byte == BYTE[:quote]
           return cursor + 1
         end
         cursor += 1
@@ -186,10 +208,10 @@ module I18nContextGenerator
     def collection_end(start_offset, opening, closing)
       depth = 0
       cursor = start_offset
-      while cursor < @content.length
-        character = @content[cursor]
-        case character
-        when '"'
+      while cursor < @content.bytesize
+        byte = byte_at(cursor)
+        case byte
+        when BYTE[:quote]
           cursor = string_end(cursor)
           next
         when opening
@@ -205,14 +227,22 @@ module I18nContextGenerator
 
     def primitive_end(start_offset)
       cursor = start_offset
-      cursor += 1 while cursor < @content.length && !",}] \t\r\n".include?(@content[cursor])
+      cursor += 1 while cursor < @content.bytesize && !PRIMITIVE_TERMINATOR_BYTES.include?(byte_at(cursor))
       cursor
     end
 
     def skip_whitespace(offset)
       cursor = offset
-      cursor += 1 while cursor < @content.length && @content[cursor].match?(/\s/)
+      cursor += 1 while cursor < @content.bytesize && JSON_WHITESPACE_BYTES.include?(byte_at(cursor))
       cursor
     end
+
+    def byte_at(offset)
+      @content.getbyte(offset)
+    end
+
+    def byte_slice(start_offset, end_offset)
+      @content.byteslice(start_offset...end_offset)
+    end
   end
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: i18n-context-generator
 version: !ruby/object:Gem::Version
-  version: 0.5.0
+  version: 0.5.1
 platform: ruby
 authors:
 - Automattic
 autorequire:
 bindir: exe
 cert_chain: []
-date: 2026-07-23 00:00:00.000000000 Z
+date: 2026-07-24 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: concurrent-ruby