apexify.js 4.7.6 → 4.7.8

package/lib/ai/ApexAI.ts

@@ -1,25 +1,25 @@
-import {
-  typeWriter,
-  groqAnalyzer,
-  readFile,
-  ApexImagine,
-  ApexText2Speech,
-  ApexChat
+import { 
+  groqAnalyzer, readFile, ApexImagine, ApexText2Speech, ApexListener, ApexChat
 } from "./utils";
 import { 
-  ModalBuilder,
-  TextInputBuilder,
-  TextInputStyle,
-  ActionRowBuilder,
-  Message,
-  PermissionResolvable,
-  TextChannel,
-  EmbedBuilder,
-  AttachmentBuilder
+  ModalBuilder, TextInputBuilder, TextInputStyle, ActionRowBuilder, Message, PermissionResolvable, TextChannel,
+  EmbedBuilder, AttachmentBuilder, MessageActionRowComponentBuilder, ButtonStyle, ButtonBuilder, CommandInteraction,
+  PermissionFlagsBits
 } from "discord.js";
+import { 
+  joinVoiceChannel, createAudioPlayer, createAudioResource, EndBehaviorType,
+   VoiceConnection, DiscordGatewayAdapterCreator, AudioPlayerStatus
+} from "@discordjs/voice";
+
 import { filters } from "./buttons/tools";
 import { imageTools } from "./buttons/drawMenu";
 import { whisper } from "./modals-chat/groq/whisper";
+import { pipeline, Readable } from "stream";
+import prism from "prism-media"
+import path from "path";
+import fs from "fs";
+import { execSync } from "child_process";
+
 
 export interface Options {
   /**
@@ -106,6 +106,17 @@ export interface Options {
           delay?: number;
       };
   };
+
+
+  voiceChannel?: {
+    enable?: boolean;
+    channelId?: string;
+    chatModel?: string;
+    voiceModel?: "elevenlabs" | "myshell-tts" | "deepinfra-tts" | "whisper-large-v3" | "distil-large-v3" | string;
+    joinOn?: {
+      triggeredWords?: string[];
+    }
+  };
   /**
    * Additional configuration options.
    * @param others.messageType Configuration options for message types.
@@ -155,6 +166,39 @@ export interface Options {
 }
 
 
+function createButtonRows(buttons: any[]): ActionRowBuilder<MessageActionRowComponentBuilder>[] {
+  if (!Array.isArray(buttons) || buttons.length === 0) return [];
+
+  const buttonRows: ActionRowBuilder<MessageActionRowComponentBuilder>[] = [];
+  let currentRow = new ActionRowBuilder<MessageActionRowComponentBuilder>();
+  
+  for (const button of buttons) {
+    if (!button.label || !button.style || !button.custom_id) continue;
+
+    const buttonComponent = new ButtonBuilder()
+      .setLabel(button.label)
+      .setStyle(button.style as ButtonStyle)
+      .setCustomId(button.custom_id);
+
+    if (button.emoji) buttonComponent.setEmoji(button.emoji);
+    if (button.url && button.style === ButtonStyle.Link) buttonComponent.setURL(button.url); 
+
+    currentRow.addComponents(buttonComponent);
+
+    if (currentRow.components.length === 5) {
+      buttonRows.push(currentRow);
+      currentRow = new ActionRowBuilder<MessageActionRowComponentBuilder>();
+    }
+  }
+
+  if (currentRow.components.length > 0) {
+    buttonRows.push(currentRow); 
+  }
+
+  return buttonRows.slice(0, 5);
+}
+
+
 export async function ApexAI(message: Message, ApexOptions: Options) {
     if (!message.guild || message.author.bot) return;
   
@@ -168,7 +212,7 @@ export async function ApexAI(message: Message, ApexOptions: Options) {
     await filters(message.client);
   
     let prompt = message.content || "";
-    const { others, chat, voice, imagine } = ApexOptions;
+    const { others, chat, voice, imagine, voiceChannel } = ApexOptions;
     const { permissions, channel, messageType, onMention, loader, buttons } = others ?? {};
     const { Api_Keys, typeWriting } = chat ?? {};
     const { nsfw } = imagine ?? {};
@@ -291,6 +335,18 @@ export async function ApexAI(message: Message, ApexOptions: Options) {
         }
       }
     }
+
+    if (voiceChannel?.enable && voiceChannel.channelId && voiceChannel.joinOn?.triggeredWords) {
+      const { channelId, joinOn, chatModel } = voiceChannel;
+      
+      const triggerDetected = joinOn.triggeredWords?.some(word => 
+          prompt.toLowerCase().startsWith(word.toLowerCase())
+      );
+
+      if (triggerDetected) {
+          return handleVoiceAI(message, channelId, chat);
+      }
+  }
   
 
     let aiResponse: string | Buffer = "";
@@ -322,7 +378,7 @@ export async function ApexAI(message: Message, ApexOptions: Options) {
           
             if (enhancer) {
                 const enhancementRequest = `Rewrite this text below in a more descriptive way, making it clearer to be visualized correctly by an AI image generator. Use stronger words and return only the enhanced prompt—nothing more, nothing less.\n\n`;
-                tempEnhancer = await ApexChat("v3-32k", prompt, { instruction: enhancementRequest });
+                tempEnhancer = await ApexChat("gpt-4-turbo", prompt, { instruction: enhancementRequest });
             }
     
             genImage = await ApexImagine(imagine?.imageModel as string, tempEnhancer || prompt,
@@ -345,7 +401,7 @@ export async function ApexAI(message: Message, ApexOptions: Options) {
           
             if (enhancer) {
                 const enhancementRequest = `Rewrite this text below in a more descriptive way, making it clearer to be visualized correctly by an AI image generator. Use stronger words and return only the enhanced prompt—nothing more, nothing less.\n\n`;
-                tempEnhancer = await ApexChat("v3-32k", prompt, { instruction: enhancementRequest });
+                tempEnhancer = await ApexChat("gpt-4-turbo", prompt, { instruction: enhancementRequest });
             }
     
             genImage = await ApexImagine(imagine?.imageModel as string, tempEnhancer || prompt,
@@ -367,7 +423,7 @@ export async function ApexAI(message: Message, ApexOptions: Options) {
       
         if (enhancer) {
           const enhancementRequest = `Rewrite this text below in a more descriptive way, making it clearer to be visualized correctly by an AI image generator. Use stronger words and return only the enhanced prompt—nothing more, nothing less.\n\n${prompt}`;
-          tempEnhancer = await ApexChat("v3-32k", prompt, { instruction: enhancementRequest });
+          tempEnhancer = await ApexChat("gpt-4-turbo", prompt, { instruction: enhancementRequest });
         }
       
         genImage = await ApexImagine(imagine?.imageModel as string, tempEnhancer || prompt,
@@ -384,6 +440,8 @@ export async function ApexAI(message: Message, ApexOptions: Options) {
     }
 
 
+    const buttonRows = createButtonRows(buttons || []);
+
     const imageAttachments: AttachmentBuilder[] = [];
     if (genImage.length > 0) {
       for (const imageUrl of genImage) {
@@ -413,30 +471,32 @@ export async function ApexAI(message: Message, ApexOptions: Options) {
     const finalText = typeof aiResponse === "string" ? `${initialContent}${aiResponse}` : "";
     const messageChunks = chunkMessage(finalText);
 
-    const sendMessage = async (content: string) => {
-        if (!content.trim()) return;
-        if (messageType?.sendAs === "embed") {
-            const embed = new EmbedBuilder()
-            .setColor(0x0099ff)
-            .setDescription(content)
-            .setFooter({ text: "AI Response" });
-            if (messageType?.type === "send") {
-                return (message.channel as TextChannel).send({ embeds: [embed] });
-            } else {
-                return message.reply({ embeds: [embed], allowedMentions: { repliedUser: false } });
-            }
-        } else {
-            if (messageType?.type === "send") {
-                return (message.channel as TextChannel).send({ content });
-            } else {
-                return message.reply({ content, allowedMentions: { repliedUser: false } });
-            }
-        }
+    const sendMessage = async (content: string, components?: ActionRowBuilder<MessageActionRowComponentBuilder>[]) => {
+      if (!content.trim()) return;
+  
+      if (messageType?.sendAs === "embed") {
+          const embed = new EmbedBuilder()
+              .setColor(0x0099ff)
+              .setDescription(content)
+              .setFooter({ text: "AI Response" });
+  
+          if (messageType?.type === "send") {
+              return (message.channel as TextChannel).send({ embeds: [embed], components });
+          } else {
+              return message.reply({ embeds: [embed], allowedMentions: { repliedUser: false }, components });
+          }
+      } else {
+          if (messageType?.type === "send") {
+              return (message.channel as TextChannel).send({ content, components });
+          } else {
+              return message.reply({ content, allowedMentions: { repliedUser: false }, components });
+          }
+      }
     };
-
-    const sendTypeWritingMessage = async (content: string) => {
+  
+    const sendTypeWritingMessage = async (content: string, components?: ActionRowBuilder<MessageActionRowComponentBuilder>[]) => {
       if (!typeWriting?.enable) {
-          return sendMessage(content);
+          return sendMessage(content, components);
       }
   
       content = content.trimStart(); 
@@ -450,30 +510,237 @@ export async function ApexAI(message: Message, ApexOptions: Options) {
           await typingMessage?.edit({ content: typedSentence });
           i += getStepCount(typeWriting.speed ?? 50);
       }
+  
+      if (components && components.length > 0) {
+          await typingMessage?.edit({ components });
+      }
+  
+      return typingMessage;
     };
   
-  function getStepCount(speed: number): number {
+  
+    function getStepCount(speed: number): number {
       const maxSteps = 120;
       const steps = Math.min(Math.ceil(speed), maxSteps);
       return steps > 0 ? steps : 1;
-  }
+    }
   
-  function sleep(ms: number): Promise<void> {
+    function sleep(ms: number): Promise<void> {
       return new Promise<void>((resolve) => setTimeout(resolve, ms));
-  }
+    }
   
 
     (async () => {
-        if (imageAttachments.length > 0) {
-            await (message.channel as TextChannel).send({ files: imageAttachments });
-        }
-        for (const chunk of messageChunks) {
+      if (imageAttachments.length > 0) {
+          await (message.channel as TextChannel).send({ files: imageAttachments });
+      }
+  
+      let lastSentMessage: Message<boolean> | null = null;
+  
+      for (let i = 0; i < messageChunks.length; i++) {
+        const chunk = messageChunks[i].trim();
+        if (!chunk) continue;
+    
+        const isLastChunk = i === messageChunks.length - 1;
+    
+        if (isLastChunk) {
+            const response = await sendTypeWritingMessage(chunk, buttonRows.length > 0 ? buttonRows : undefined);
+            if (response) lastSentMessage = response as Message<boolean>;
+        } else {
             await sendTypeWritingMessage(chunk);
+        }
+    
+        if (!isLastChunk) {
             await new Promise((resolve) => setTimeout(resolve, chat?.typeWriting?.delay ?? 500));
         }
+      }
+
+      if (voiceAttachment) {
+          await (message.channel as TextChannel).send({ files: [voiceAttachment] });
+      }
+  
+
+      if (lastSentMessage && buttonRows.length > 0) {
+          await lastSentMessage.edit({ components: buttonRows }).catch(() => null);
+      }
+  })();
+  
+
+};
+
+const voiceQueue: { userId: string; text: string }[] = [];
+let isProcessing = false;
+let voiceConnection: VoiceConnection | null = null;
+let activeUser: string | null = null;
+let isRecording = false;
+
+const recordingsDir = path.join(process.cwd(), "recordings");
+if (!fs.existsSync(recordingsDir)) {
+    fs.mkdirSync(recordingsDir, { recursive: true });
+}
+
+try {
+    execSync("ffmpeg -version > nul 2>&1");
+  } catch (err) {
+    console.error("🚨 FFmpeg is NOT installed or not in PATH! Install it first.");
+}
+
+export async function handleVoiceAI(message: any, voiceChannelId: string, chat: any) {
+    const guild = message.guild;
+    if (!guild) return;
+
+    const channel = guild.channels.cache.get(voiceChannelId);
+    if (!channel || channel.type !== 2) {
+      return await message.reply(`🚫 Invalid voice channel ID: ${voiceChannelId}`);
+    }
+
+    const botMember = guild.members.me;
+    if (!botMember) return;
+    const permissions = channel.permissionsFor(botMember);
+
+    if (
+        !permissions?.has(PermissionFlagsBits.Connect) ||
+        !permissions.has(PermissionFlagsBits.Speak) ||
+        !permissions.has(PermissionFlagsBits.Stream) ||
+        !permissions.has(PermissionFlagsBits.UseVAD)
+    ) {
+        return await message.reply("🚫 I don't have the required permissions to join and speak in this voice channel.");
+    }
+
+    if (voiceConnection) {
+      return await message.reply("⚠️ AI is already in a voice channel.");
+    }
+
+    voiceConnection = joinVoiceChannel({
+        channelId: channel.id,
+        guildId: guild.id,
+        adapterCreator: guild.voiceAdapterCreator as any,
+        selfMute: false,
+        selfDeaf: false
+    });
+
+    activeUser = message.author.id;
+
+    captureAudio(voiceConnection, chat);
+}
+
+function captureAudio(connection: VoiceConnection, chat: any) {
+    const receiver = connection.receiver;
+
+    receiver.speaking.on("start", async (userId) => {
+        if (userId !== activeUser || isRecording) return;
+        isRecording = true;
+
+
+        const rawFilePath = path.join(recordingsDir, `${userId}.pcm`);
+        const wavFilePath = path.join(recordingsDir, `${userId}.wav`);
+
+        const opusStream = receiver.subscribe(userId, {
+            end: { behavior: EndBehaviorType.AfterSilence, duration: 2000 } 
+        });
+
+        const pcmStream = new prism.opus.Decoder({
+            frameSize: 960,
+            channels: 1,
+            rate: 48000
+        });
+
+        const writeStream = fs.createWriteStream(rawFilePath);
+        pipeline(opusStream, pcmStream, writeStream, (err) => {
+            isRecording = false;
+            if (err) {
+                console.error("❌ Error writing PCM file:", err);
+                return;
+            }
+
+            convertPCMtoWAV(rawFilePath, wavFilePath, chat);
+        });
+    });
+}
+
+function convertPCMtoWAV(inputPCM: string, outputWAV: string, chat: any) {
+    if (!fs.existsSync(inputPCM) || fs.statSync(inputPCM).size === 0) {
+        return;
+    }
+
+    try {
+        execSync(`ffmpeg -y -f s16le -ar 48000 -ac 1 -i "${inputPCM}" -acodec pcm_s16le "${outputWAV}" > nul 2>&1`);
+        
+        if (fs.existsSync(outputWAV)) {
+          transcribeAudio(outputWAV, chat);
+        }
+
+    } catch (error) {
+        console.error("❌ FFmpeg failed:", error);
+    }
+}
 
-        if (voiceAttachment) {
-            await (message.channel as TextChannel).send({ files: [voiceAttachment] });
+// 🛠 **Transcribe Audio using ApexListener**
+async function transcribeAudio(filePath: string, chat: any) {
+    try {
+        const transcribedText = await ApexListener({
+            filepath: filePath,
+            prompt: "Transcribe what the user said in English.",
+            lang: "en"
+        });
+
+        if (transcribedText.transcribe) {
+            voiceQueue.push({ userId: activeUser || "unknown", text: transcribedText.transcribe });
+            processQueue(chat);
         }
-    })();
-};
+
+        fs.unlinkSync(filePath.replace(".wav", ".pcm"));
+    } catch (error) {
+        console.error("❌ Error in transcription:", error);
+    }
+}
+
+async function processQueue(chat: any) {
+    if (isProcessing || voiceQueue.length === 0) {
+        if (voiceQueue.length === 0) {
+            leaveVoiceChannel();
+        }
+        return;
+    }
+
+    isProcessing = true;
+    const { userId, text } = voiceQueue.shift()!;
+
+    try {
+
+      const aiResponse = await ApexChat(chat?.chatModel as string, text, {
+        instruction: chat.instruction,
+        memory: chat?.memory?.memoryOn,
+        userId: userId,
+        limit: chat?.memory?.limit,
+        threshold: chat?.memory?.threshold
+      });
+      const audioBuffer = await ApexText2Speech({ inputText: aiResponse, modelName: "elevenlabs" });
+
+        if (voiceConnection) {
+            const player = createAudioPlayer();
+            const audioStream = Readable.from(audioBuffer);
+            const resource = createAudioResource(audioStream);
+            voiceConnection.subscribe(player);
+            player.play(resource);
+
+            player.on(AudioPlayerStatus.Idle, () => {
+                isProcessing = false;
+                processQueue(chat);
+            });
+        }
+    } catch (error) {
+        console.error("❌ Error processing AI response:", error);
+        isProcessing = false;
+    }
+}
+
+// 🔄 **Leave Voice Channel When Done**
+function leaveVoiceChannel() {
+    if (voiceConnection) {
+        console.log("👋 AI is leaving the voice channel...");
+        voiceConnection.destroy();
+        voiceConnection = null;
+        activeUser = null;
+    }
+}

package/lib/ai/ApexModules.ts

@@ -795,7 +795,7 @@ async function processChunk(
 }
 
 
-async function ApexListener(options: { filepath: string, model: string, prompt?: string, lang?: string, apiKey?: string }) {
+async function ApexListener(options: { filepath: string, model?: string, prompt?: string, lang?: string, apiKey?: string }) {
 
   const { filepath, model = 'v3', prompt = '', lang = 'en', apiKey = undefined } = options;
 
@@ -807,11 +807,14 @@ async function ApexListener(options: { filepath: string, model: string, prompt?:
 
   const transcribe = await whisper(prompt, filepath, validatedLang, apiKey);
 
-  const enhancer = `Below is the the speech/text that has been said in the voice/audio file.\n\n${transcribe}`
-  const response = await ApexChat(model, enhancer, {});
+  const enhancer = `Below is the the speech/text that has been said in the voice/audio file.\n\n${transcribe}`;
+
+  let response = "";
+  if (model) response = await ApexChat(model, enhancer, {});
+  
 
   return { response, transcribe };
-}
+};
 
 function validateLanguageInput(lang: string): string | null {
   const supportedLanguages = [

package/lib/ai/modals-chat/groq/imageAnalyzer.ts

@@ -40,7 +40,7 @@ function convertBufferToDataUrl(buffer: Buffer): string {
 export async function groqAnalyzer({ img, ApiKey, prompt }: GroqAnalyzerOptions): Promise<string> {
   try {
     const groq = new Groq({
-      apiKey: ApiKey || "your-api-key-here",
+      apiKey: ApiKey || "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",
     });
 
     let imageDataUrl: string;

package/lib/index.ts

@@ -86,4 +86,4 @@ const Apexify = {
     ApexFileReader, ApexImageAnalyzer, ApexVideo, ApexText2Speech, resetHistory
 };
 
-export default Apexify;
+export default Apexify;

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "apexify.js",
-  "version": "4.7.6",
+  "version": "4.7.8",
   "description": "Unlimited AI models and Canvas library. Supports ts & js (supports front/back end).",
   "author": "zenith-79",
   "license": "MIT",
@@ -225,9 +225,11 @@
     "apexify.js"
   ],
   "dependencies": {
+    "@discordjs/opus": "^0.10.0",
+    "@discordjs/voice": "^0.18.0",
     "@google/generative-ai": "^0.22.0",
     "@napi-rs/canvas": "^0.1.53",
-    "apexify.js": "^4.7.5",
+    "apexify.js": "^4.7.7",
     "axios": "^1.7.7",
     "discord.js": "^14.18.0",
     "fluent-ffmpeg": "^2.1.3",
@@ -236,6 +238,7 @@
     "hercai": "^12.4.0",
     "imgur": "^2.4.2",
     "openai": "^4.71.1",
+    "opusscript": "^0.0.8",
     "pdf-parse": "^1.1.1",
     "sharp": "^0.33.5",
     "verse.db": "^2.2.15"