CLIFF Localization Suite

CLIFF Localization Suite Alpha is the first public release of the plugin (0.1.0-alpha.1). It turns CLIFF 1.0 into a readable, reviewable, AI-ready intermediate format for UE localization data: import and export both run as steps of the official GatherText pipeline, and every write still goes through the engine’s manifest, archive and .locres generators.

Alpha highlights

• CLIFF 1.0 parsing and validation — FCliffDocument::ParseAndValidate / Validate cover every header, group and entry field plus the closed vocabularies, verified against 10 valid and 32 invalid fixtures
• Official pipeline import — .cliff → FLocTextHelper writes manifest / archive → GenerateTextLocalizationResource compiles .locmeta / .locres → FTextLocalizationManager live refresh
• Official pipeline export — manifest plus every culture archive (or a compiled .locres) → one <target-language>/<clan>.cliff per culture plus a .cliffmap.json key sidecar
• Semantic mapping tables — clan ↔ UE Namespace and Namespace → type / emotion are configurable in Project Settings; missing information degrades to sentence with a warning
• DeepSeek AI translation — a three-stage export → translate → import pipeline with concurrent cultures, placeholder validation, self-translation and empty-target rejection, and an automatic CLIFF repair loop
• Official entry points — Dashboard toolbar buttons, the official Slate commandlet progress window, and CliffLocalizationSuite.Import/Export/AITranslate console commands

Documentation

Tutorials

Step-by-step guides from installation to a complete CLIFF round trip.

• Installation → — install the plugin, enable the modules and verify
• Dashboard workflow → — your first CLIFF export and import
• Your first AI translation → — configure DeepSeek and fill in missing targets
• CLIFF in five minutes → — read a .cliff file with confidence

Guides

How the pipeline, the semantic mapping and the AI translation are designed.

• Architecture overview → — modules, official integration points and trade-offs
• Data flow → — the import, export and AI translation chains end to end
• Import pipeline → — manifest / archive writes, metadata, compilation and live refresh
• Export pipeline → — grouping, stable IDs, semantic fallback and the sidecar
• Semantic mapping → — clan ↔ Namespace and type / emotion derivation
• AI translation setup → — model, prompts, translation standard and batching
• Write-back safety and repair → — self-translations, empty targets, placeholders and the repair loop
• Troubleshooting → — common errors, causes and fixes

Reference

Complete references for settings, commands and data formats.

• Settings reference → — every UCliffLocalizationSuiteSettings field
• Commands and commandlets → — console commands, GatherText steps and parameters
• Vocabularies and mapping → — type / emotion / status vocabularies and the UE mapping
• Key sidecar → — .cliffmap.json structure and restore rules
• FAQ → — frequent questions and behavioural notes

API

The complete C++ API and data structure documentation.

• Runtime module overview → — entry points for parsing, validation, serialization and mapping
• Runtime class reference → — FCliffDocument, FCliffEntry, FCliffSemanticMapping and friends
• Importer and exporter → — FCliffImporter / FCliffExporter and their settings structs
• AI translation and commandlets → — FCliffAITranslator and the UGatherTextCommandletBase subclasses