Package com.smartgwt.client.ai

Class AI

java.lang.Object
com.smartgwt.client.ai.AI
public class AI extends Object
Provides class methods for enabling and disabling the use of AI technology, registering and unregistering AI engines, and performing high-level operations using installed AI engines.
See Also:

  • Constructor Details

    • AI

      public AI()

  • Method Details

    • setDefaultEngineId

      public static void setDefaultEngineId(String defaultEngineId)
      The ID of the default AIEngine to use.
      Parameters:
      defaultEngineId - new defaultEngineId. Default value is "gpt-4.1".

    • getDefaultEngineId

      public static String getDefaultEngineId()
      The ID of the default AIEngine to use.
      Returns:
      current value of defaultEngineId

    • setDefaultMaxRetries

      public static void setDefaultMaxRetries(int defaultMaxRetries)
      The defualt maximum number of retries for any one particular request to AI.
      Parameters:
      defaultMaxRetries - new defaultMaxRetries. Default value is 2.

    • getDefaultMaxRetries

      public static int getDefaultMaxRetries()
      The defualt maximum number of retries for any one particular request to AI.
      Returns:
      current value of defaultMaxRetries

    • setDisabled

      public static void setDisabled(Boolean disabled)
      Whether AI is disabled.

      By default, AI is disabled. This static property must be set to false and the default AIEngine must be registered in order to enable the use of AI in the application.

      Parameters:
      disabled - new disabled. Default value is true.

    • getDisabled

      public static Boolean getDisabled()
      Whether AI is disabled.

      By default, AI is disabled. This static property must be set to false and the default AIEngine must be registered in order to enable the use of AI in the application.

      Returns:
      current value of disabled

    • addMockResponses

      public static void addMockResponses(Map[] responses)
      Alias for addSpoofedResponses().
      Parameters:
      responses - mock response entries

    • askDataQuestion

      public static void askDataQuestion(String question)
      Asks AI to answer a question about the data of the application.
      Parameters:
      question - The text of the end-user's question or their request for an answer to a data question.

    • askDataQuestion

      public static void askDataQuestion(UserAIRequest question)
      Asks AI to answer a question about the data of the application.
      Parameters:
      question - The text of the end-user's question or their request for an answer to a data question.

    • askDataQuestion

      public static void askDataQuestion(String question, DataSource... dataSources)
      See askDataQuestion()

    • askDataQuestion

      public static void askDataQuestion(UserAIRequest question, DataSource... dataSources)
      See askDataQuestion()

    • askDataQuestion

      public static void askDataQuestion(String question, DataSource[] dataSources, DataQuestionSettings settings)
      See askDataQuestion()

    • askDataQuestion

      public static void askDataQuestion(UserAIRequest question, DataSource[] dataSources, DataQuestionSettings settings)
      See askDataQuestion()

    • askDataQuestion

      public static void askDataQuestion(String question, DataSource[] dataSources, DataQuestionSettings settings, AskDataQuestionResultCallback callback)
      Asks AI to answer a question about the data of the application.
      Parameters:
      question - The text of the end-user's question or their request for an answer to a data question.
      dataSources - The available data sources. All data sources in the array must have a global ID. If null, then the array of all DataSources available to the AI module is used.
      settings - Settings to use when answering the data question.
      callback - The callback to call with the result.

    • askDataQuestion

      public static void askDataQuestion(UserAIRequest question, DataSource[] dataSources, DataQuestionSettings settings, AskDataQuestionResultCallback callback)
      Asks AI to answer a question about the data of the application.
      Parameters:
      question - The text of the end-user's question or their request for an answer to a data question.
      dataSources - The available data sources. All data sources in the array must have a global ID. If null, then the array of all DataSources available to the AI module is used.
      settings - Settings to use when answering the data question.
      callback - The callback to call with the result.

    • cancelDataQuestion

      public static void cancelDataQuestion(String questionId)
      Cancels a data question if not already canceled.
      Parameters:
      questionId - The ID of the data question to cancel.

    • clearMockResponses

      public static void clearMockResponses()
      Alias for clearSpoofedResponses().

    • getDataSourceSummaries

      public static Map[] getDataSourceSummaries()
      Returns an array of summaries (see getDataSourceSummary()) for the specified DataSources, or all registered application DataSources if no list is provided. The settings parameter is passed through to each getDataSourceSummary() call.
      Returns:
      array of summary objects

    • getDataSourceSummaries

      public static Map[] getDataSourceSummaries(String[] dataSourceNames)
      See getDataSourceSummaries()

    • getDataSourceSummaries

      public static Map[] getDataSourceSummaries(String[] dataSourceNames, Map settings)
      Returns an array of summaries (see getDataSourceSummary()) for the specified DataSources, or all registered application DataSources if no list is provided. The settings parameter is passed through to each getDataSourceSummary() call.
      Parameters:
      dataSourceNames - DataSource IDs to summarize; defaults to getDataSourceNames()
      settings - passed through to getDataSourceSummary()
      Returns:
      array of summary objects

    • getDataSourceSummary

      public static Map getDataSourceSummary(DataSource dataSource)
      Returns a structured summary of a DataSource at a configurable detail level. In "compact" mode (default), only identity metadata, primary key, and foreign key relationships are included ? suitable for large-app DS selection where full field listings would exceed the AI context window. Higher detail levels progressively add field information via an otherFields property.

      Detail levels (set via settings.detail):

      • "compact" ? ID, title, description, fieldCount, pk, fk. No field list.
      • "salientNames" ? adds otherFields as an array of field name strings for salient/critical fields (PK and FK fields already shown in pk/fk are excluded).
      • "salientTyped" ? like salientNames but each entry is an object with name, type, and optionally required and valueMap.
      • "allTyped" ? like salientTyped but includes all fields, not just salient ones.
      • "full" ? all fields with the full salientFieldAttributes attribute mask.

      Settings:

      • detail (String) ? one of the levels above; default "compact".
      • maxFields (Integer) ? cap on the number of fields in otherFields.
      • excludeHousekeeping (Boolean) ? omit auto-populated fields (creator, modifier, creatorTimestamp, modifierTimestamp); default true for all modes except "full".
      • includeData (String) ? if set, includes representative data samples from available client-side caches. Values: "fieldValues" (per-field unique sample values added as sampleValues on each field entry) or "records" (whole sample records added as sampleRecords). Data is sourced from clientOnly cacheData, cacheAllData cacheResultSet, or any default-fetch (no operationId) ResultSet observing the DataSource ? no server requests are triggered.
      • dataSampleCount (Integer) ? maximum unique sample values per field (fieldValues mode) or sample records (records mode); default 3.
      • dataSampleMaxSearch (Integer) ? maximum cache records to scan when collecting unique sample values; default 100.
      Parameters:
      dataSource - DataSource or ID
      Returns:
      structured summary
      See Also:

    • getDataSourceSummary

      public static Map getDataSourceSummary(String dataSource)
      Returns a structured summary of a DataSource at a configurable detail level. In "compact" mode (default), only identity metadata, primary key, and foreign key relationships are included ? suitable for large-app DS selection where full field listings would exceed the AI context window. Higher detail levels progressively add field information via an otherFields property.

      Detail levels (set via settings.detail):

      • "compact" ? ID, title, description, fieldCount, pk, fk. No field list.
      • "salientNames" ? adds otherFields as an array of field name strings for salient/critical fields (PK and FK fields already shown in pk/fk are excluded).
      • "salientTyped" ? like salientNames but each entry is an object with name, type, and optionally required and valueMap.
      • "allTyped" ? like salientTyped but includes all fields, not just salient ones.
      • "full" ? all fields with the full salientFieldAttributes attribute mask.

      Settings:

      • detail (String) ? one of the levels above; default "compact".
      • maxFields (Integer) ? cap on the number of fields in otherFields.
      • excludeHousekeeping (Boolean) ? omit auto-populated fields (creator, modifier, creatorTimestamp, modifierTimestamp); default true for all modes except "full".
      • includeData (String) ? if set, includes representative data samples from available client-side caches. Values: "fieldValues" (per-field unique sample values added as sampleValues on each field entry) or "records" (whole sample records added as sampleRecords). Data is sourced from clientOnly cacheData, cacheAllData cacheResultSet, or any default-fetch (no operationId) ResultSet observing the DataSource ? no server requests are triggered.
      • dataSampleCount (Integer) ? maximum unique sample values per field (fieldValues mode) or sample records (records mode); default 3.
      • dataSampleMaxSearch (Integer) ? maximum cache records to scan when collecting unique sample values; default 100.
      Parameters:
      dataSource - DataSource or ID
      Returns:
      structured summary
      See Also:

    • getDataSourceSummary

      public static Map getDataSourceSummary(DataSource dataSource, Map settings)
      Returns a structured summary of a DataSource at a configurable detail level. In "compact" mode (default), only identity metadata, primary key, and foreign key relationships are included ? suitable for large-app DS selection where full field listings would exceed the AI context window. Higher detail levels progressively add field information via an otherFields property.

      Detail levels (set via settings.detail):

      • "compact" ? ID, title, description, fieldCount, pk, fk. No field list.
      • "salientNames" ? adds otherFields as an array of field name strings for salient/critical fields (PK and FK fields already shown in pk/fk are excluded).
      • "salientTyped" ? like salientNames but each entry is an object with name, type, and optionally required and valueMap.
      • "allTyped" ? like salientTyped but includes all fields, not just salient ones.
      • "full" ? all fields with the full salientFieldAttributes attribute mask.

      Settings:

      • detail (String) ? one of the levels above; default "compact".
      • maxFields (Integer) ? cap on the number of fields in otherFields.
      • excludeHousekeeping (Boolean) ? omit auto-populated fields (creator, modifier, creatorTimestamp, modifierTimestamp); default true for all modes except "full".
      • includeData (String) ? if set, includes representative data samples from available client-side caches. Values: "fieldValues" (per-field unique sample values added as sampleValues on each field entry) or "records" (whole sample records added as sampleRecords). Data is sourced from clientOnly cacheData, cacheAllData cacheResultSet, or any default-fetch (no operationId) ResultSet observing the DataSource ? no server requests are triggered.
      • dataSampleCount (Integer) ? maximum unique sample values per field (fieldValues mode) or sample records (records mode); default 3.
      • dataSampleMaxSearch (Integer) ? maximum cache records to scan when collecting unique sample values; default 100.
      Parameters:
      dataSource - DataSource or ID
      settings - detail level and field options
      Returns:
      structured summary
      See Also:

    • getDataSourceSummary

      public static Map getDataSourceSummary(String dataSource, Map settings)
      Returns a structured summary of a DataSource at a configurable detail level. In "compact" mode (default), only identity metadata, primary key, and foreign key relationships are included ? suitable for large-app DS selection where full field listings would exceed the AI context window. Higher detail levels progressively add field information via an otherFields property.

      Detail levels (set via settings.detail):

      • "compact" ? ID, title, description, fieldCount, pk, fk. No field list.
      • "salientNames" ? adds otherFields as an array of field name strings for salient/critical fields (PK and FK fields already shown in pk/fk are excluded).
      • "salientTyped" ? like salientNames but each entry is an object with name, type, and optionally required and valueMap.
      • "allTyped" ? like salientTyped but includes all fields, not just salient ones.
      • "full" ? all fields with the full salientFieldAttributes attribute mask.

      Settings:

      • detail (String) ? one of the levels above; default "compact".
      • maxFields (Integer) ? cap on the number of fields in otherFields.
      • excludeHousekeeping (Boolean) ? omit auto-populated fields (creator, modifier, creatorTimestamp, modifierTimestamp); default true for all modes except "full".
      • includeData (String) ? if set, includes representative data samples from available client-side caches. Values: "fieldValues" (per-field unique sample values added as sampleValues on each field entry) or "records" (whole sample records added as sampleRecords). Data is sourced from clientOnly cacheData, cacheAllData cacheResultSet, or any default-fetch (no operationId) ResultSet observing the DataSource ? no server requests are triggered.
      • dataSampleCount (Integer) ? maximum unique sample values per field (fieldValues mode) or sample records (records mode); default 3.
      • dataSampleMaxSearch (Integer) ? maximum cache records to scan when collecting unique sample values; default 100.
      Parameters:
      dataSource - DataSource or ID
      settings - detail level and field options
      Returns:
      structured summary
      See Also:

    • getEngine

      public static AIEngine getEngine()
      Returns the AIEngine having the given engine ID.
      Returns:
      the AIEngine, or null if the AIEngine could not be found.

    • getEngine

      public static AIEngine getEngine(String engineId)
      Returns the AIEngine having the given engine ID.
      Parameters:
      engineId - the engineId of the AIEngine to get. If not specified, defaultEngineId is used.
      Returns:
      the AIEngine, or null if the AIEngine could not be found.

    • getMockingPolicy

      public static String getMockingPolicy()
      Returns the current mockingPolicy. The return value is always one of "none", "auto", or "interactive"; the legacy responseSpoofingMode values "hybrid" and "full" are normalized before returning.
      Returns:
      The current mocking policy: "none", "auto", or "interactive"

    • isEnabled

      public static boolean isEnabled()
      Determines whether AI support is enabled. disabled must be set to false and the default AIEngine must be registered in order to enable the use of AI.
      Returns:
      true if the use of AI is enabled; false otherwise.
      See Also:
      • com.smartgwt.client.ai.AI#disabled
      • com.smartgwt.client.ai.AI#defaultEngineId

    • pauseDataQuestion

      public static void pauseDataQuestion(String questionId)
      Pauses a data question if not already paused or canceled.
      Parameters:
      questionId - The ID of the data question to pause.

    • registerEngine

      public static boolean registerEngine(AIEngine engine)
      Registers the given AIEngine.
      Parameters:
      engine - The AIEngine to register.
      Returns:
      true if the AIEngine was successfully registered; false otherwise.
      See Also:

    • resumeDataQuestion

      public static void resumeDataQuestion(String questionId)
      Resumes a data question if paused.
      Parameters:
      questionId - The ID of the data question to resume.

    • sendPrompt

      public static void sendPrompt(String dynamicString, Map context, AIResponseCallback callback)
      Evaluates the given dynamic string to form a prompt string that is then sent as the request to the default AI engine.

      Within dynamicString, any evaluated JavaScript expressions have access to all of the values in the context ValueMap.

      Parameters:
      dynamicString - A dynamic string. See DynamicString
      context - A map from each in-scope Identifier to its value.
      callback - The callback to fire with the response from AI.

    • setMockingPolicy

      public static void setMockingPolicy(String policy)
      Sets mockingPolicy. Pass one of "none", "auto", or "interactive".
      Parameters:
      policy - The mocking policy: "none", "auto", or "interactive"

    • unregisterEngine

      public static boolean unregisterEngine(String engineId)
      Unregisters an AIEngine specified by its ID.
      Parameters:
      engineId - the ID of the AIEngine to unregister.
      Returns:
      true if the AIEngine was successfully unregistered; false otherwise.
      See Also:

    • moduleLoaded

      public static boolean moduleLoaded()
      Reports whether the optional AI module is loaded for the page.
      Returns:
      true if and only if the AI module has been loaded

    • checkModuleLoaded

      public static void checkModuleLoaded(String faultingAction)