TextLanguageRouter

Route text strings to different pipeline branches based on their detected language.

Key Features

• Detects the language of a text string using the langdetect library.
• Routes text to named output connections based on the detected language.
• Supports configurable list of target languages.
• Text in languages not in the configured list is routed to the unmatched output.
• Useful for building multilingual query pipelines that need different processing per language.

Configuration

1. Drag the TextLanguageRouter component onto the canvas from the Component Library.
2. Click on the component to open the configuration panel.
3. On the General tab:
   • Set languages to the list of ISO 639-1 language codes you want to support (for example, ["en", "de"]). Each language creates a separate named output connection.

Connections

TextLanguageRouter receives a text string and routes it to one of its named output connections based on the detected language. Each configured language gets its own named output; text that doesn't match goes to the unmatched output. Connect each output to the appropriate retriever or processing component for that language.

Source Code

To check this component's source code, open text_language_router.py in the Haystack Core Integrations repository.

Usage Examples

Basic Configuration

  TextLanguageRouter:
    type: haystack_integrations.components.routers.langdetect.text_language_router.TextLanguageRouter
    init_parameters:
      languages:
      - en
      - de

Using the Component in a Pipeline

# haystack-pipeline
components:
  TextLanguageRouter:
    type: haystack_integrations.components.routers.langdetect.text_language_router.TextLanguageRouter
    init_parameters:
      languages:
      - en

  en_retriever:
    type: haystack_integrations.components.retrievers.opensearch.bm25_retriever.OpenSearchBM25Retriever
    init_parameters:
      top_k: 10
      document_store:
        type: haystack_integrations.document_stores.opensearch.document_store.OpenSearchDocumentStore
        init_parameters:
          hosts:
          index: english-index

connections:
- sender: TextLanguageRouter.en
  receiver: en_retriever.query

max_runs_per_component: 100

metadata: {}

inputs:
  query:
  - TextLanguageRouter.text

Parameters

Inputs

Parameter	Type	Description
text	str	The text string to route based on its detected language.

Outputs

The outputs are dynamic and depend on the languages configured at initialization. There is one output per language (named with the ISO code), plus an unmatched output for text in other languages.

Parameter	Type	Description
<language_code>	str	The input text, routed to the connection matching the detected language.
unmatched	str	The input text, if its language is not in the configured list.

Init Parameters

These are the parameters you can configure in Pipeline Builder:

Parameter	Type	Default	Description
languages	Optional[List[str]]	["en"]	A list of ISO 639-1 language codes to route text to. Each code creates a named output connection. Text that doesn't match any configured language is routed to the unmatched output.

Run Method Parameters

These are the parameters you can configure for the component's run() method. This means you can pass these parameters at query time through the API, in Playground, or when running a job. For details, see Modify Pipeline Parameters at Query Time.

Parameter	Type	Default	Description
text	str		The text string to detect language for and route.

Related Information

• DocumentLanguageClassifier
• Pipeline Components