feat: make custom server works with stream

src/main/kotlin/cc/unitmesh/devti/llms/custom/CustomLLMProvider.kt

@@ -1,15 +1,24 @@
 package cc.unitmesh.devti.llms.custom
 
 import cc.unitmesh.devti.llms.CodeCopilotProvider
+import cc.unitmesh.devti.llms.azure.SimpleOpenAIBody
 import cc.unitmesh.devti.prompting.model.CustomPromptConfig
 import cc.unitmesh.devti.settings.AutoDevSettingsState
+import com.fasterxml.jackson.databind.ObjectMapper
 import com.intellij.openapi.components.Service
 import com.intellij.openapi.diagnostic.logger
 import com.intellij.openapi.project.Project
+import com.theokanning.openai.completion.chat.ChatCompletionResult
+import com.theokanning.openai.service.SSE
+import io.reactivex.BackpressureStrategy
+import io.reactivex.Flowable
+import io.reactivex.FlowableEmitter
+import kotlinx.coroutines.Dispatchers
 import kotlinx.coroutines.ExperimentalCoroutinesApi
 import kotlinx.coroutines.channels.awaitClose
 import kotlinx.coroutines.flow.Flow
 import kotlinx.coroutines.flow.callbackFlow
+import kotlinx.coroutines.withContext
 import kotlinx.serialization.Serializable
 import kotlinx.serialization.encodeToString
 import kotlinx.serialization.json.Json
@@ -63,30 +72,35 @@ class CustomLLMProvider(val project: Project) : CodeCopilotProvider {
             .post(body)
             .build()
 
-        return callbackFlow {
-            val listener = object : EventSourceListener() {
-                override fun onOpen(eventSource: EventSource, response: Response) {
-                    println("onOpen")
-                }
-
-                override fun onEvent(eventSource: EventSource, id: String?, type: String?, data: String) {
-                    println(data)
-                    trySend(data)
-                }
+        val call = client.newCall(request)
+        val emitDone = false
 
-                override fun onClosed(eventSource: EventSource) {
+        val sseFlowable = Flowable
+            .create({ emitter: FlowableEmitter<SSE> ->
+                call.enqueue(cc.unitmesh.devti.llms.azure.ResponseBodyCallback(emitter, emitDone))
+            }, BackpressureStrategy.BUFFER)
 
-                }
+        try {
+            return callbackFlow {
+                withContext(Dispatchers.IO) {
+                    sseFlowable
+                        .doOnError(Throwable::printStackTrace)
+                        .blockingForEach { sse ->
+                            val result: ChatCompletionResult =
+                                ObjectMapper().readValue(sse!!.data, ChatCompletionResult::class.java)
+                            val completion = result.choices[0].message
+                            if (completion != null && completion.content != null) {
+                                trySend(completion.content)
+                            }
+                        }
 
-                override fun onFailure(eventSource: EventSource, t: Throwable?, response: Response?) {
                     close()
                 }
             }
-
-            val eventSource = EventSources.createFactory(client).newEventSource(request, listener)
-
-            awaitClose {
-                eventSource.cancel()
+        } catch (e: Exception) {
+            logger.error("Failed to stream", e)
+            return callbackFlow {
+                close()
             }
         }
     }