API 1.2+4 open ai chat completion fix

src/main/kotlin/com/aitrainer/api/controller/OpenAIController.kt

@@ -1,13 +1,12 @@
 package com.aitrainer.api.controller
 
 import com.aallam.openai.api.BetaOpenAI
-import com.aallam.openai.api.chat.ChatMessage
 import com.aitrainer.api.model.OpenAI
 import com.aitrainer.api.model.OpenAIChat
 import com.aitrainer.api.openai.OpenAIService
-import com.google.gson.Gson
 import kotlinx.coroutines.*
 import org.slf4j.LoggerFactory
+import org.springframework.beans.factory.annotation.Value
 import org.springframework.web.bind.annotation.*
 
 @RestController
@@ -17,9 +16,9 @@ class OpenAIController() {
 
     @OptIn(DelicateCoroutinesApi::class)
     @PostMapping("/openai/completion")
-    fun getOpenAIResponse(@RequestBody question: String) : String {
+    fun getOpenAIResponse(@RequestBody question: String, @Value("\${openai.key}") openaiKey: String) : String {
-        var result = ""
+        var result: String
-        val openAIService = OpenAIService(null, null)
+        val openAIService = OpenAIService(openaiKey, null, null)
         val deferred = GlobalScope.async {
             openAIService.completion(question)
         }
@@ -32,9 +31,9 @@ class OpenAIController() {
 
     @OptIn(DelicateCoroutinesApi::class)
     @PostMapping("/openai/completion_with_model")
-    fun getOpenAIResponseWithModel(@RequestBody openai: OpenAI) : String {
+    fun getOpenAIResponseWithModel(@RequestBody openai: OpenAI, @Value("\${openai.key}") openaiKey: String) : String {
-        var result = ""
+        var result: String
-        val openAIService = OpenAIService(openai.modelName, openai.temperature)
+        val openAIService = OpenAIService(openaiKey, openai.modelName, openai.temperature)
         val deferred = GlobalScope.async {
             openAIService.completion(openai.question)
         }
@@ -47,24 +46,23 @@ class OpenAIController() {
 
     @OptIn(BetaOpenAI::class, DelicateCoroutinesApi::class)
     @PostMapping("/openai/chat_completion")
-    fun getOpenAIChatCompletion(@RequestBody openai: OpenAIChat) : String {
+    fun getOpenAIChatCompletion(@RequestBody openai: OpenAIChat, @Value("\${openai.key}") openaiKey: String, ) : String {
-        var result = ""
+        var result: String
-        val openAIService = OpenAIService(openai.modelName, openai.temperature)
+        val openAIService = OpenAIService(openaiKey, openai.modelName, openai.temperature)
         val deferred = GlobalScope.async {
             openAIService.chatCompletion(openai.messages)
         }
         runBlocking {
             result = deferred.await().toString()
-            println("Result: $result" )
         }
         return result
     }
 
     @OptIn(DelicateCoroutinesApi::class)
     @GetMapping("/openai/list_models")
-    fun getOpenAIModels(): MutableList<String> {
+    fun getOpenAIModels(@Value("\${openai.key}") openaiKey: String): MutableList<String> {
-        var result = mutableListOf<String>()
+        var result: MutableList<String>
-        val openAIService = OpenAIService(null, null)
+        val openAIService = OpenAIService(openaiKey,null, null)
         val deferred = GlobalScope.async {
             openAIService.getModels()
         }

src/main/kotlin/com/aitrainer/api/openai/OpenAIService.kt

@@ -14,31 +14,23 @@ import com.aallam.openai.client.OpenAIConfig
 import com.google.gson.Gson
 import kotlinx.coroutines.Dispatchers
 import kotlinx.coroutines.withContext
-import java.util.Properties
+import org.springframework.beans.factory.annotation.Value
-
+import org.springframework.stereotype.Service
-class OpenAIService(private val modelName: String?, private val temperature: Double?) {
+@Service
-
+class OpenAIService(@Value("\${openai.key}") private val openaiKey: String, private val modelName: String?, private val temperature: Double?) {
 
     private var openAI: OpenAI? = null
     var model: Model? = null
-    private val properties = Properties()
-
-    init {
-        val inputStream = ClassLoader.getSystemResourceAsStream("application.properties")
-        properties.load(inputStream)
-        inputStream?.close()
-    }
 
     private var modelId: ModelId? = null
     private suspend fun connect(modelName: String) {
         val config = OpenAIConfig(
-            token = properties.getProperty("openai.key"),
+            token = openaiKey,
             logLevel = LogLevel.All
         )
         openAI = OpenAI(config)
         modelId = ModelId(modelName)
         model = openAI!!.model(modelId!!)
-
     }
 
 
@@ -112,11 +104,12 @@ class OpenAIService(private val modelName: String?, private val temperature: Dou
 
     suspend fun getModels(): MutableList<String> {
         return withContext(Dispatchers.IO)  {
+            val list: MutableList<String> = mutableListOf()
             if (openAI == null) {
-                openAI = OpenAI(token = properties.getProperty("openai.key"))
+                openAI = OpenAI(token = openaiKey)
             }
 
-            val list: MutableList<String> = mutableListOf()
+
             openAI!!.models().forEach {
                 println(it)
                 list.add(it.id.id)

src/test/kotlin/com/aitrainer/api/test/CompletionTest.kt

@@ -3,15 +3,16 @@ package com.aitrainer.api.test
 import com.aitrainer.api.controller.OpenAIController
 import org.junit.jupiter.api.Test
 import org.junit.jupiter.api.TestInstance
+import org.springframework.beans.factory.annotation.Value
 import org.springframework.boot.test.context.SpringBootTest
 
 @SpringBootTest
 @TestInstance(TestInstance.Lifecycle.PER_CLASS)
 class CompletionTest {
     @Test
-    fun testCompletion() {
+    fun testCompletion(@Value("\${openai.key}") openaiKey: String ) {
         val openAIController = OpenAIController()
-        val response = openAIController.getOpenAIResponse("Mennyi 3 darab közepes tükörtojás kalóriaértéke? Csak egy szám intervallumértéket adj vissza eredményként")
+        val response = openAIController.getOpenAIResponse( "Mennyi 3 darab közepes tükörtojás kalóriaértéke? Csak egy szám intervallumértéket adj vissza eredményként", openaiKey)
         println(response)
        // response = openAIController.getOpenAIResponse("Az utolsó 5 kérdésemre: Mennyi 3 darab tükörtojás kalóriaértéke? nagy szórású intervallumokat adtál vissza. Hogyan tudom pontosítani a kérdést?")
        // println(response)

src/test/kotlin/com/aitrainer/api/test/diet/MealTest.kt

@@ -100,7 +100,7 @@ class MealTest {
         .andExpect(jsonPath("$.quantity").value(90.0))
         .andExpect(jsonPath("$.proteinMax").value(25.0))
 
-        val mvcResult = mockMvc.perform(
+        var mvcResult = mockMvc.perform(
             MockMvcRequestBuilders.post("/api/meal")
                 .contentType(MediaType.APPLICATION_JSON)
                 .header("Authorization", "Bearer $authToken")
@@ -110,8 +110,8 @@ class MealTest {
             .andReturn()
 
         val gson= Gson()
-        val newRawMaterialJson = mvcResult.response.contentAsString
+        val newMealJson = mvcResult.response.contentAsString
-        val newMeal = gson.fromJson(newRawMaterialJson, Meal::class.java)
+        val newMeal = gson.fromJson(newMealJson, Meal::class.java)
 
         mockMvc.perform(
             MockMvcRequestBuilders.post("/api/meal")
@@ -146,9 +146,29 @@ class MealTest {
             .andExpect(jsonPath("$.name").value("Töltötttojás"))
             .andExpect(jsonPath("$.quantity").value(330.0))
 
-        mealRepository.delete(meal)
+
-        mealRepository.delete(meal2)
+        mvcResult = mockMvc.perform(get("/api/meal/${newMeal.id-1}")
-        mealRepository.delete(meal3)
+            .header("Authorization", "Bearer $authToken")
+            .contentType(MediaType.APPLICATION_JSON))
+            .andExpect(status().isOk)
+            . andReturn()
+
+        val newMealJson1 = mvcResult.response.contentAsString
+        val newMeal1 = gson.fromJson(newMealJson1, Meal::class.java)
+
+        mealRepository.delete(newMeal1)
+        mealRepository.delete(newMeal)
+
+        mvcResult = mockMvc.perform(get("/api/meal/${newMeal.id+1}")
+            .header("Authorization", "Bearer $authToken")
+            .contentType(MediaType.APPLICATION_JSON))
+            .andExpect(status().isOk)
+            . andReturn()
+
+        val newMealJson3 = mvcResult.response.contentAsString
+        val newMeal3 = gson.fromJson(newMealJson3, Meal::class.java)
+
+        mealRepository.delete(newMeal3)
     }
 
     private fun toJson(obj: Any): String {

src/test/kotlin/com/aitrainer/api/test/openai/OpenAITest.kt

@@ -63,6 +63,7 @@ class OpenAITest {
         val listMessages = listOf(systemMsg)
         val gson = Gson()
         val messages = gson.toJson(listMessages)
+        println("****** json $messages ********* ")
 
 
         val openai = OpenAIChat(
@@ -71,11 +72,44 @@ class OpenAITest {
             temperature = 0.1
         )
 
+        val json: String = toJson(openai)
+        println(json)
+
         mockMvc.perform(
             MockMvcRequestBuilders.post("/api/openai/chat_completion")
                 .contentType(MediaType.APPLICATION_JSON)
                 .header("Authorization", "Bearer $authToken")
-                .content(toJson(openai))
+                .content(json)
+        )
+        .andExpect(MockMvcResultMatchers.status().isOk)
+
+    }
+
+    @OptIn(BetaOpenAI::class)
+    @Test
+    fun `get a chat message completion no modelname`() {
+        val systemMsg = ChatMessage(
+            role = ChatRole.User,
+            content = "Te a Diet4You applikáció asszisztense vagy. Add meg ennek az ételnek a kalória és tápanyagadatait: 'Sült sütőtök 100g'. A válasz ez az objektum JSON alakított formája legyen: Meal [cal: double, ch: double, fat: double, protein: double, sugar: double, portion: double]. A 'portion' paraméter ezt tartalmazza: hány gramm egy átlagos adag ebből az ételből? "
+        )
+        val listMessages = listOf(systemMsg)
+        val gson = Gson()
+        val messages = gson.toJson(listMessages)
+        println("****** json $messages ********* ")
+
+
+        val openai = OpenAIChat(
+            messages = messages
+        )
+
+        val json: String = toJson(openai)
+        println(json)
+
+        mockMvc.perform(
+            MockMvcRequestBuilders.post("/api/openai/chat_completion")
+                .contentType(MediaType.APPLICATION_JSON)
+                .header("Authorization", "Bearer $authToken")
+                .content(json)
         )
             .andExpect(MockMvcResultMatchers.status().isOk)