Multi-Language Localization Setup

Texturge provides runtime language switching through ULocalizationSubsystem (a UGameInstanceSubsystem), and the FSemanticAnchor LCS matching mechanism ensures per-character animation correspondence survives translation.

NOTE

Texturge does not use UE5’s native UStringTable for text localization. The proprietary ULocalizationSubsystem and semantic anchoring mechanism are independent of the engine’s localization framework — they only reuse the engine’s culture switch event (FInternationalization::OnCultureChanged).

Supported Languages

The Texturge.uplugin localization targets support 15 languages (NativeCulture = en):

Language	Code		Language	Code
English	en		Polish	pl
Simplified Chinese	zh-Hans		German	de
Japanese	ja		Russian	ru
Korean	ko		French	fr
Spanish	es		Brazilian Portuguese	pt-BR
Portuguese	pt		Turkish	tr
Arabic	ar		Latin American Spanish	es-419
			Italian	it

Runtime Language Switching

// Get the subsystem
ULocalizationSubsystem* Localization = GetGameInstance()->GetSubsystem<ULocalizationSubsystem>();

// Switch language (Culture Code)
Localization->SetLanguage(TEXT("ja"));       // Japanese
Localization->SetLanguage(TEXT("zh-Hans"));  // Simplified Chinese
Localization->SetLanguage(TEXT("ar"));       // Arabic

SetLanguage sets the current culture and broadcasts OnLanguageChanged(FString NewCulture) (multicast delegate). UAnimatedTextBlock / UAnimatedRichTextBlock listen to this event and automatically trigger re-bake and replay after a language switch.

Semantic Anchoring

FSemanticAnchor (AnchorId / ContextualSnippet / SourceOffset) uses the LCS (longest common substring) algorithm to locate anchors across translated texts.

The Problem Scenario

Translated text differs in length and character sequence:

• English "Hello" (5 chars) → Simplified Chinese "你好" (2 chars)
• Each character owns independent scheduling and curve data in the animation

Anchoring Principle

static TMap<FName, int32> ULocalizationSubsystem::RelocateAnchors(
    const TArray<FSemanticAnchor>& InAnchors,
    const FString& InSourceText,
    const FString& InTargetText);

1. Exact matching (case-insensitive) takes priority
2. Unmatched anchors use FindLongestCommonSubstring for approximate matching
3. CalculateConfidence computes match confidence (0.0–1.0); low-confidence matches are deduplicated
4. Offsets are clamped within the target text range

Guaranteed Behavior

• Existing character animation curves carry over to the new language
• Deleted / added characters never misalign or crash animation data
• Animation start / end timings stay consistent across languages

Localization Asset Configuration

The plugin’s .uplugin declares 7 localization targets, all covering 15 languages:

Target	Loading policy	Content
TexturgeRuntime	Always	runtime text (node display names, prompts)
TexturgeEditor	Editor	editor UI text
TexturgeEditorTutorials	Never	editor tutorials
TexturgePropertyNames	PropertyNames	property name translations
TexturgeToolTips	ToolTips	tooltip translations
TexturgeKeywords	Editor	keywords
TexturgeCategory	Editor	category names

The preconfigured localization targets require no manual creation — they register automatically when the plugin loads.

Unicode Support

Internally Texturge uses the Texturge::Unicode namespace utilities for code-point-level classification (IsCJK / IsCJKPunctuation / IsHiragana / IsKatakana / IsHangulSyllables etc.), together with FCharBreakIterator precomputed break points (CJK attachment, surrogate-pair Emoji, Hangul syllables), ensuring correct per-glyph animation for CJK and RTL text.