Package com.sap.ai.sdk.orchestration

Class AzureContentFilter

java.lang.Object

com.sap.ai.sdk.orchestration.AzureContentFilter

All Implemented Interfaces:

ContentFilter

public class AzureContentFilter
extends Object
implements ContentFilter

A content filter wrapping Azure Content Safety.

This class allows setting filtration thresholds for different content categories such as hate, self-harm, sexual, and violence.

Example usage:

 AzureContentFilter filter = new AzureContentFilter()
     .hate(AzureFilterThreshold.ALLOW_SAFE)
     .selfHarm(AzureFilterThreshold.ALLOW_SAFE_LOW);
 

Constructor Summary

Constructors

Constructor	Description
AzureContentFilter()

Method Summary

Modifier and Type	Method	Description
AzureContentSafetyInputFilterConfig	createInputFilterConfig()	Converts AzureContentFilter to its serializable counterpart AzureContentSafetyInputFilterConfig.
AzureContentSafetyOutputFilterConfig	createOutputFilterConfig()	Converts AzureContentFilter to its serializable counterpart AzureContentSafetyOutput.
AzureContentFilter	hate(AzureFilterThreshold hate)	The filter category for hate content.
AzureContentFilter	promptShield(Boolean promptShield)	A flag to set prompt shield on input filer.
AzureContentFilter	protectedMaterialCode(boolean protectedMaterialCode)	The filter category for protected material content.
AzureContentFilter	selfHarm(AzureFilterThreshold selfHarm)	The filter category for self-harm content.
AzureContentFilter	sexual(AzureFilterThreshold sexual)	The filter category for sexual content.
AzureContentFilter	violence(AzureFilterThreshold violence)	The filter category for violence content.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Details

AzureContentFilter

public AzureContentFilter()

Method Details

createInputFilterConfig

@Nonnull
public AzureContentSafetyInputFilterConfig createInputFilterConfig()

Converts AzureContentFilter to its serializable counterpart AzureContentSafetyInputFilterConfig.

Specified by:

createInputFilterConfig in interface ContentFilter

Returns:

the corresponding AzureContentSafetyInputFilterConfig object.

Throws:

IllegalArgumentException - if no policies are set.

createOutputFilterConfig

@Nonnull
public AzureContentSafetyOutputFilterConfig createOutputFilterConfig()

Converts AzureContentFilter to its serializable counterpart AzureContentSafetyOutput.

Specified by:

createOutputFilterConfig in interface ContentFilter

Returns:

the corresponding AzureContentSafetyOutputFilterConfig object.

Throws:

IllegalArgumentException - if no policies are set.

hate

public AzureContentFilter hate(@Nullable
 AzureFilterThreshold hate)

The filter category for hate content.

Returns:

this.

selfHarm

public AzureContentFilter selfHarm(@Nullable
 AzureFilterThreshold selfHarm)

The filter category for self-harm content.

Returns:

this.

sexual

public AzureContentFilter sexual(@Nullable
 AzureFilterThreshold sexual)

The filter category for sexual content.

Returns:

this.

violence

public AzureContentFilter violence(@Nullable
 AzureFilterThreshold violence)

The filter category for violence content.

Returns:

this.

protectedMaterialCode

public AzureContentFilter protectedMaterialCode(boolean protectedMaterialCode)

The filter category for protected material content.

Returns:

this.

promptShield

public AzureContentFilter promptShield(@Nullable
 Boolean promptShield)

A flag to set prompt shield on input filer.

Returns:

this.

Since:

1.9.0