Module de.seggebaeing.sqlanalyzer

Package de.seggebaeing.sqlanalyzer.logic.promptable.impl.llm

Class DeepSeekPromptHandler

java.lang.Object

de.seggebaeing.sqlanalyzer.logic.promptable.impl.llm.AbstractLLMHandler

de.seggebaeing.sqlanalyzer.logic.promptable.impl.llm.DeepSeekPromptHandler

All Implemented Interfaces:

Promptable

public class DeepSeekPromptHandler
extends AbstractLLMHandler

Handles communication with the DeepSeek API for generating chat completions.

Builds and sends requests to the DeepSeek endpoint, processes responses, and extracts the generated text. Handles error responses, including unexpected formats, API errors, and potential rate limits.

Field Summary

Fields inherited from class de.seggebaeing.sqlanalyzer.logic.promptable.impl.llm.AbstractLLMHandler

client, gson

Constructor Summary

Constructors

Constructor	Description
DeepSeekPromptHandler()	Constructs a new DeepSeekPromptHandler and initializes the required HTTP client and JSON utilities via the superclass.

Method Summary

Modifier and Type	Method	Description
String	prompt(String input, String model, String apiKey, double temperature)	Sends a prompt request to the DeepSeek API and retrieves the generated response.

Methods inherited from class java.lang.Object

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

Constructor Details

DeepSeekPromptHandler

public DeepSeekPromptHandler()

Constructs a new DeepSeekPromptHandler and initializes the required HTTP client and JSON utilities via the superclass.

Method Details

prompt

public String prompt(String input,
 String model,
 String apiKey,
 double temperature)
              throws LLMException

Sends a prompt request to the DeepSeek API and retrieves the generated response.

Builds a JSON request with the given parameters, sends it via HTTP, and parses the API response to extract the first choice's message content.

Parameters:

input - the user prompt to send to the model

model - the model identifier to use

apiKey - the DeepSeek API key for authentication

temperature - the sampling temperature controlling randomness

Returns:

the generated response text from DeepSeek

Throws:

LLMException - if the request fails, returns an error, or the response is malformed