Package org.egothor.methodatlas.ai

Class OllamaClient

java.lang.Object

org.egothor.methodatlas.ai.OllamaClient

All Implemented Interfaces:

AiProviderClient

public final class OllamaClient
extends Object
implements AiProviderClient

AiProviderClient implementation for a locally running Ollama inference service.

This client submits taxonomy-guided classification prompts to the Ollama HTTP API and converts the returned model response into the internal AiClassSuggestion representation used by the MethodAtlas AI subsystem.

Operational Responsibilities

• verifying local Ollama availability
• constructing chat-style inference requests
• injecting the system prompt and taxonomy-guided user prompt
• executing HTTP requests against the Ollama API
• extracting and normalizing JSON classification results

The client uses the Ollama /api/chat endpoint for inference and the /api/tags endpoint as a lightweight availability probe.

This implementation is intended primarily for local, offline, or privacy-preserving inference scenarios where source code should not be sent to an external provider.

See Also:

• AiProviderClient
• AiProviderFactory
• AiSuggestionEngine

Constructor Summary

Constructors

Constructor	Description
OllamaClient(AiOptions options)	Creates a new Ollama client using the supplied runtime configuration.

Method Summary

Modifier and Type	Method	Description
boolean	isAvailable()	Determines whether the configured Ollama service is reachable.
AiClassSuggestion	suggestForClass(String fqcn, String classSource, String taxonomyText, List<PromptBuilder.TargetMethod> targetMethods)	Submits a classification request to the Ollama chat API for the specified test class.

Methods inherited from class java.lang.Object

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

Constructor Details

OllamaClient

public OllamaClient(AiOptions options)

Creates a new Ollama client using the supplied runtime configuration.

The configuration determines the base URL of the Ollama service, the model identifier, and request timeout values used by this client.

Parameters:

options - AI runtime configuration

Method Details

isAvailable

public boolean isAvailable()

Determines whether the configured Ollama service is reachable.

The method performs a lightweight availability probe against the /api/tags endpoint. If the endpoint responds successfully, the provider is considered available.

Any exception raised during the probe is treated as an indication that the provider is unavailable.

Specified by:

isAvailable in interface AiProviderClient

Returns:

true if the Ollama service is reachable; false otherwise

suggestForClass

public AiClassSuggestion suggestForClass(String fqcn,
 String classSource,
 String taxonomyText,
 List<PromptBuilder.TargetMethod> targetMethods)
                                  throws AiSuggestionException

Submits a classification request to the Ollama chat API for the specified test class.

The request consists of:

• a system prompt enforcing strict JSON output
• a user prompt containing the test class source and taxonomy text
• provider options such as deterministic temperature settings

The returned response is expected to contain a JSON object in the message content field. That JSON text is extracted, deserialized into an AiClassSuggestion, and then normalized before being returned.

Specified by:

suggestForClass in interface AiProviderClient

Parameters:

fqcn - fully qualified class name being analyzed

classSource - complete source code of the class being analyzed

taxonomyText - taxonomy definition guiding classification

targetMethods - deterministically extracted JUnit test methods that must be classified

Returns:

normalized AI classification result

Throws:

AiSuggestionException - if the request fails, if the provider returns invalid content, or if response deserialization fails

See Also:

• AiClassSuggestion
• AiMethodSuggestion