Microsoft LUIS translation with Crowdin

With this app you can translate Microsoft LUIS in Crowdin and keep the app multilingual.

File Format Details

Sample File

{
  "luis_schema_version": "4.0.0",
  "versionId": "0.4",
  "name": "FAQ",
  "desc": "FAQ Chatbot",
  "culture": "en-us",
  "tokenizerVersion": "1.0.0",
  "intents": [
    {
      "name": "company_deliveries"
    },
    {
      "name": "company_mitigation"
    }
  ],
  "entities": [],
  "composites": [],
  "closedLists": [
    {
      "name": "test",
      "subLists": [
        {
          "canonicalForm": "test",
          "list": [
            "test 1",
            "test 2",
            "test 3"
          ]
        }
      ],
      "roles": []
    }
  ],
  "patternAnyEntities": [],
  "regex_entities": [],
  "prebuiltEntities": [],
  "model_features": [],
  "regex_features": [],
  "patterns": [],
  "utterances": [
    {
      "text": "test",
      "intent": "None",
      "entities": []
    }
  ],
  "settings": []
}

Translating Microsoft LUIS

Notes

• Make sure to configure the proper export patterns for your Microsoft LUIS files you upload to Crowdin. When downloading you should receive files named like "alarm.check_usersays_en.json", "alarm.check_usersays_fr.json". It's recommended to use Crowdin CLI or CVS integration to get translated package ready for import to the Microsoft LUIS.

• Every entry in Intent files woul require langauge code. This langauge code will not be set automatically, either you or a translators should put proper values using Crowdin Editor. If you enable automatic duplicates translation in your project settings, this action can be done once per every target language.