Initial

Author: a-d

foundation-models/openai/src/main/java/com/sap/ai/sdk/foundationmodels/openai/OpenAiClient.java

@@ -37,7 +37,7 @@
 /** Client for interacting with OpenAI models. */
 @Slf4j
 @RequiredArgsConstructor(access = AccessLevel.PRIVATE)
-public final class OpenAiClient {
+public final class OpenAiClient implements OpenAiClientWithSystemPrompt {
   private static final String DEFAULT_API_VERSION = "2024-02-01";
   static final ObjectMapper JACKSON = getDefaultObjectMapper();
   @Nullable private String systemPrompt = null;
@@ -113,18 +113,12 @@ public static OpenAiClient withCustomDestination(@Nonnull final Destination dest
    * @return the client
    */
   @Nonnull
-  public OpenAiClient withSystemPrompt(@Nonnull final String systemPrompt) {
+  public OpenAiClientWithSystemPrompt withSystemPrompt(@Nonnull final String systemPrompt) {
     this.systemPrompt = systemPrompt;
     return this;
   }
 
-  /**
-   * Generate a completion for the given user prompt.
-   *
-   * @param prompt a text message.
-   * @return the completion output
-   * @throws OpenAiClientException if the request fails
-   */
+  @Override
   @Nonnull
   public OpenAiChatCompletionOutput chatCompletion(@Nonnull final String prompt)
       throws OpenAiClientException {
@@ -146,35 +140,10 @@ public OpenAiChatCompletionOutput chatCompletion(@Nonnull final String prompt)
   @Nonnull
   public OpenAiChatCompletionOutput chatCompletion(
       @Nonnull final OpenAiChatCompletionParameters parameters) throws OpenAiClientException {
-    warnIfUnsupportedUsage();
     return execute("/chat/completions", parameters, OpenAiChatCompletionOutput.class);
   }
 
-  /**
-   * Stream a completion for the given prompt. Returns a <b>lazily</b> populated stream of text
-   * chunks. To access more details about the individual chunks, use {@link
-   * #streamChatCompletionDeltas(OpenAiChatCompletionParameters)}.
-   *
-   * <p>The stream should be consumed using a try-with-resources block to ensure that the underlying
-   * HTTP connection is closed.
-   *
-   * <p>Example:
-   *
-   * <pre>{@code
-   * try (var stream = client.streamChatCompletion("...")) {
-   *       stream.forEach(System.out::println);
-   * }
-   * }</pre>
-   *
-   * <p>Please keep in mind that using a terminal stream operation like {@link Stream#forEach} will
-   * block until all chunks are consumed. Also, for obvious reasons, invoking {@link
-   * Stream#parallel()} on this stream is not supported.
-   *
-   * @param prompt a text message.
-   * @return A stream of message deltas
-   * @throws OpenAiClientException if the request fails or if the finish reason is content_filter
-   * @see #streamChatCompletionDeltas(OpenAiChatCompletionParameters)
-   */
+  @Override
   @Nonnull
   public Stream<String> streamChatCompletion(@Nonnull final String prompt)
       throws OpenAiClientException {
@@ -225,18 +194,10 @@ private static void throwOnContentFilter(@Nonnull final OpenAiChatCompletionDelt
   @Nonnull
   public Stream<OpenAiChatCompletionDelta> streamChatCompletionDeltas(
       @Nonnull final OpenAiChatCompletionParameters parameters) throws OpenAiClientException {
-    warnIfUnsupportedUsage();
     parameters.enableStreaming();
     return executeStream("/chat/completions", parameters, OpenAiChatCompletionDelta.class);
   }
 
-  private void warnIfUnsupportedUsage() {
-    if (systemPrompt != null) {
-      log.warn(
-          "Previously set messages will be ignored, set it as an argument of this method instead.");
-    }
-  }
-
   /**
    * Get a vector representation of a given input that can be easily consumed by machine learning
    * models and algorithms.

foundation-models/openai/src/main/java/com/sap/ai/sdk/foundationmodels/openai/OpenAiClientWithSystemPrompt.java

@@ -0,0 +1,49 @@
+package com.sap.ai.sdk.foundationmodels.openai;
+
+import com.sap.ai.sdk.foundationmodels.openai.model.OpenAiChatCompletionOutput;
+import com.sap.ai.sdk.foundationmodels.openai.model.OpenAiChatCompletionParameters;
+import java.util.stream.Stream;
+import javax.annotation.Nonnull;
+
+/** Client for interacting with OpenAI models. Allows for convenient string prompts only. */
+public interface OpenAiClientWithSystemPrompt {
+
+  /**
+   * Generate a completion for the given user prompt.
+   *
+   * @param prompt a text message.
+   * @return the completion output
+   * @throws OpenAiClientException if the request fails
+   */
+  @Nonnull
+  OpenAiChatCompletionOutput chatCompletion(@Nonnull final String prompt)
+      throws OpenAiClientException;
+
+  /**
+   * Stream a completion for the given prompt. Returns a <b>lazily</b> populated stream of text
+   * chunks. To access more details about the individual chunks, use {@link
+   * OpenAiClient#streamChatCompletionDeltas(OpenAiChatCompletionParameters)}.
+   *
+   * <p>The stream should be consumed using a try-with-resources block to ensure that the underlying
+   * HTTP connection is closed.
+   *
+   * <p>Example:
+   *
+   * <pre>{@code
+   * try (var stream = client.streamChatCompletion("...")) {
+   *       stream.forEach(System.out::println);
+   * }
+   * }</pre>
+   *
+   * <p>Please keep in mind that using a terminal stream operation like {@link Stream#forEach} will
+   * block until all chunks are consumed. Also, for obvious reasons, invoking {@link
+   * Stream#parallel()} on this stream is not supported.
+   *
+   * @param prompt a text message.
+   * @return A stream of message deltas
+   * @throws OpenAiClientException if the request fails or if the finish reason is content_filter
+   * @see OpenAiClient#streamChatCompletionDeltas(OpenAiChatCompletionParameters)
+   */
+  @Nonnull
+  Stream<String> streamChatCompletion(@Nonnull final String prompt) throws OpenAiClientException;
+}