Update static/conv.html
Browse files- static/conv.html +437 -327
static/conv.html
CHANGED
|
@@ -442,361 +442,471 @@
|
|
| 442 |
<!-- Bootstrap JS Bundle -->
|
| 443 |
<script src="https://cdn.jsdelivr.net/npm/[email protected]/dist/js/bootstrap.bundle.min.js"></script>
|
| 444 |
<script>
|
| 445 |
-
|
| 446 |
-
|
| 447 |
-
|
| 448 |
-
|
| 449 |
-
|
| 450 |
-
|
| 451 |
-
|
| 452 |
-
|
| 453 |
-
|
| 454 |
-
|
| 455 |
-
|
| 456 |
-
|
| 457 |
-
|
| 458 |
-
|
| 459 |
-
|
| 460 |
-
|
| 461 |
-
|
| 462 |
-
|
| 463 |
-
|
| 464 |
-
|
| 465 |
-
|
| 466 |
-
|
| 467 |
-
|
| 468 |
-
|
| 469 |
-
|
| 470 |
-
|
| 471 |
-
|
| 472 |
-
|
| 473 |
-
|
| 474 |
-
|
| 475 |
-
|
| 476 |
-
|
| 477 |
-
|
| 478 |
-
|
| 479 |
-
|
| 480 |
-
|
| 481 |
-
|
| 482 |
-
|
| 483 |
-
|
| 484 |
-
|
| 485 |
-
|
| 486 |
-
|
| 487 |
-
|
| 488 |
-
|
| 489 |
-
|
| 490 |
-
|
| 491 |
-
|
| 492 |
-
|
| 493 |
-
|
| 494 |
-
|
| 495 |
-
|
| 496 |
-
|
| 497 |
-
|
| 498 |
-
|
| 499 |
-
|
| 500 |
-
|
| 501 |
-
|
| 502 |
-
const avatar = document.createElement("div");
|
| 503 |
-
avatar.className = "message-avatar";
|
| 504 |
-
avatar.innerHTML = "<i class='bi bi-robot'></i>";
|
| 505 |
-
currentAssistantGroup.appendChild(avatar);
|
| 506 |
-
|
| 507 |
-
conversationArea.appendChild(currentAssistantGroup);
|
| 508 |
-
}
|
| 509 |
-
messageGroup = currentAssistantGroup;
|
| 510 |
-
currentUserGroup = null;
|
| 511 |
-
}
|
| 512 |
-
|
| 513 |
-
// Create message element
|
| 514 |
-
const messageDiv = document.createElement("div");
|
| 515 |
-
messageDiv.className = `message ${sender}`;
|
| 516 |
-
messageDiv.textContent = content;
|
| 517 |
-
|
| 518 |
-
const timestamp = document.createElement("div");
|
| 519 |
-
timestamp.className = "time-stamp";
|
| 520 |
-
timestamp.textContent = timeString;
|
| 521 |
|
| 522 |
-
|
| 523 |
-
|
|
|
|
|
|
|
| 524 |
|
| 525 |
-
|
| 526 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 527 |
|
| 528 |
-
|
| 529 |
-
|
| 530 |
-
|
| 531 |
-
|
| 532 |
-
});
|
| 533 |
-
}
|
| 534 |
-
|
| 535 |
-
function startAnalyzingAudioLevels() {
|
| 536 |
-
if (!audioStream) return;
|
| 537 |
|
| 538 |
-
|
| 539 |
-
|
| 540 |
-
|
| 541 |
-
|
| 542 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 543 |
|
| 544 |
-
|
| 545 |
-
const dataArray = new Uint8Array(bufferLength);
|
| 546 |
|
| 547 |
-
|
| 548 |
-
|
| 549 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
| 550 |
|
| 551 |
-
|
| 552 |
-
|
| 553 |
-
|
|
|
|
| 554 |
|
| 555 |
-
//
|
| 556 |
-
|
| 557 |
-
|
| 558 |
-
sum += dataArray[i];
|
| 559 |
}
|
| 560 |
-
|
| 561 |
-
|
| 562 |
-
|
| 563 |
-
|
| 564 |
-
|
| 565 |
-
|
| 566 |
-
|
| 567 |
-
|
| 568 |
-
|
| 569 |
-
|
| 570 |
-
|
| 571 |
-
|
| 572 |
-
|
| 573 |
-
|
| 574 |
-
|
| 575 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 576 |
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 577 |
|
| 578 |
-
|
| 579 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 580 |
|
| 581 |
try {
|
| 582 |
-
|
| 583 |
-
|
| 584 |
-
audioStream.getTracks().forEach(track => track.stop());
|
| 585 |
-
}
|
| 586 |
|
| 587 |
-
|
| 588 |
-
|
|
|
|
|
|
|
| 589 |
|
| 590 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
| 591 |
|
| 592 |
-
|
| 593 |
-
|
| 594 |
-
|
| 595 |
-
return;
|
| 596 |
-
}
|
| 597 |
|
| 598 |
-
|
|
|
|
|
|
|
| 599 |
|
| 600 |
-
|
| 601 |
-
|
| 602 |
-
|
| 603 |
-
|
| 604 |
-
|
| 605 |
-
|
| 606 |
-
|
| 607 |
-
|
| 608 |
-
|
| 609 |
-
|
| 610 |
-
|
| 611 |
-
method: "POST",
|
| 612 |
-
body: formData
|
| 613 |
-
});
|
| 614 |
-
|
| 615 |
-
if (response.ok) {
|
| 616 |
-
const userMessage = response.headers.get("X-User-Message") || "No user message";
|
| 617 |
-
const llmResponse = response.headers.get("X-LLM-Response") || "No response";
|
| 618 |
-
|
| 619 |
-
// Add messages to chat
|
| 620 |
-
addMessageToChat(userMessage, 'user');
|
| 621 |
-
addMessageToChat(llmResponse, 'assistant');
|
| 622 |
-
|
| 623 |
-
// Get audio response and play it
|
| 624 |
-
const audioData = await response.blob();
|
| 625 |
-
audioPlayer.src = URL.createObjectURL(audioData);
|
| 626 |
-
audioPlayer.play();
|
| 627 |
-
|
| 628 |
-
updateState("idle");
|
| 629 |
-
updateStatus("Ready for next sound input", "bi-broadcast");
|
| 630 |
-
|
| 631 |
-
// If in auto mode, start listening again
|
| 632 |
-
if (isAutoListening) {
|
| 633 |
-
setTimeout(() => {
|
| 634 |
-
startAnalyzingAudioLevels();
|
| 635 |
-
}, 1000);
|
| 636 |
-
}
|
| 637 |
-
} else {
|
| 638 |
-
updateState("idle");
|
| 639 |
-
updateStatus("Error processing audio", "bi-exclamation-triangle");
|
| 640 |
-
}
|
| 641 |
-
} catch (error) {
|
| 642 |
-
console.error("Error:", error);
|
| 643 |
-
updateState("idle");
|
| 644 |
-
updateStatus("Error processing audio", "bi-exclamation-triangle");
|
| 645 |
-
}
|
| 646 |
-
};
|
| 647 |
-
|
| 648 |
-
audioChunks = [];
|
| 649 |
-
mediaRecorder.start();
|
| 650 |
-
|
| 651 |
-
updateState("listening");
|
| 652 |
-
updateStatus("Listening...", "bi-ear");
|
| 653 |
-
|
| 654 |
-
// Set max recording duration (8 seconds)
|
| 655 |
-
setTimeout(() => {
|
| 656 |
-
if (mediaRecorder && mediaRecorder.state === "recording") {
|
| 657 |
-
stopListening();
|
| 658 |
-
}
|
| 659 |
-
}, 8000);
|
| 660 |
-
|
| 661 |
-
isListening = true;
|
| 662 |
} catch (error) {
|
| 663 |
-
console.error("Error
|
| 664 |
-
updateStatus("Microphone access denied", "bi-mic-mute");
|
| 665 |
updateState("idle");
|
|
|
|
| 666 |
}
|
| 667 |
-
}
|
| 668 |
-
|
| 669 |
-
function stopListening() {
|
| 670 |
-
if (!isListening) return;
|
| 671 |
-
|
| 672 |
-
if (mediaRecorder && mediaRecorder.state === "recording") {
|
| 673 |
-
mediaRecorder.stop();
|
| 674 |
-
}
|
| 675 |
-
|
| 676 |
-
isListening = false;
|
| 677 |
-
}
|
| 678 |
-
|
| 679 |
-
function updateState(state) {
|
| 680 |
-
listenBall.classList.remove("listening", "processing");
|
| 681 |
-
|
| 682 |
-
if (state === "listening") {
|
| 683 |
-
listenBall.classList.add("listening");
|
| 684 |
-
listenBall.innerHTML = `
|
| 685 |
-
<div class="sound-wave"></div>
|
| 686 |
-
<div class="sound-wave" style="animation-delay: 0.5s"></div>
|
| 687 |
-
<div class="sound-wave" style="animation-delay: 1s"></div>
|
| 688 |
-
<i class="bi bi-soundwave"></i>
|
| 689 |
-
`;
|
| 690 |
-
} else if (state === "processing") {
|
| 691 |
-
listenBall.classList.add("processing");
|
| 692 |
-
listenBall.innerHTML = `<i class="bi bi-arrow-repeat"></i>`;
|
| 693 |
-
} else {
|
| 694 |
-
listenBall.innerHTML = `<i class="bi bi-soundwave"></i>`;
|
| 695 |
-
}
|
| 696 |
-
}
|
| 697 |
-
|
| 698 |
-
function toggleContinuousListening() {
|
| 699 |
-
isAutoListening = !isAutoListening;
|
| 700 |
|
| 701 |
-
|
| 702 |
-
|
| 703 |
-
|
| 704 |
-
|
| 705 |
-
|
| 706 |
-
clearInterval(silenceDetectionInterval);
|
| 707 |
-
}
|
| 708 |
-
}
|
| 709 |
-
|
| 710 |
-
async function convertWebMToWav(blob) {
|
| 711 |
-
return new Promise((resolve, reject) => {
|
| 712 |
-
try {
|
| 713 |
-
const reader = new FileReader();
|
| 714 |
-
reader.onload = function () {
|
| 715 |
-
const audioContext = new AudioContext();
|
| 716 |
-
audioContext.decodeAudioData(reader.result)
|
| 717 |
-
.then(buffer => {
|
| 718 |
-
const wavBuffer = audioBufferToWav(buffer);
|
| 719 |
-
resolve(new Blob([wavBuffer], { type: "audio/wav" }));
|
| 720 |
-
})
|
| 721 |
-
.catch(error => {
|
| 722 |
-
console.error("Error decoding audio data:", error);
|
| 723 |
-
reject(error);
|
| 724 |
-
});
|
| 725 |
-
};
|
| 726 |
-
reader.readAsArrayBuffer(blob);
|
| 727 |
-
} catch (error) {
|
| 728 |
-
console.error("Error in convertWebMToWav:", error);
|
| 729 |
-
reject(error);
|
| 730 |
-
}
|
| 731 |
-
});
|
| 732 |
-
}
|
| 733 |
-
|
| 734 |
-
function audioBufferToWav(buffer) {
|
| 735 |
-
let numOfChan = buffer.numberOfChannels,
|
| 736 |
-
length = buffer.length * numOfChan * 2 + 44,
|
| 737 |
-
bufferArray = new ArrayBuffer(length),
|
| 738 |
-
view = new DataView(bufferArray),
|
| 739 |
-
channels = [],
|
| 740 |
-
sampleRate = buffer.sampleRate,
|
| 741 |
-
offset = 0,
|
| 742 |
-
pos = 0;
|
| 743 |
-
setUint32(0x46464952); // "RIFF"
|
| 744 |
-
setUint32(length - 8);
|
| 745 |
-
setUint32(0x45564157); // "WAVE"
|
| 746 |
-
setUint32(0x20746d66); // "fmt " chunk
|
| 747 |
-
setUint32(16); // length = 16
|
| 748 |
-
setUint16(1); // PCM (uncompressed)
|
| 749 |
-
setUint16(numOfChan);
|
| 750 |
-
setUint32(sampleRate);
|
| 751 |
-
setUint32(sampleRate * 2 * numOfChan);
|
| 752 |
-
setUint16(numOfChan * 2);
|
| 753 |
-
setUint16(16); // bits per sample
|
| 754 |
-
setUint32(0x61746164); // "data" chunk
|
| 755 |
-
setUint32(length - pos - 4);
|
| 756 |
-
for (let i = 0; i < buffer.numberOfChannels; i++)
|
| 757 |
-
channels.push(buffer.getChannelData(i));
|
| 758 |
-
while (pos < length) {
|
| 759 |
-
for (let i = 0; i < numOfChan; i++) {
|
| 760 |
-
let sample = Math.max(-1, Math.min(1, channels[i][offset]));
|
| 761 |
-
sample = sample < 0 ? sample * 0x8000 : sample * 0x7FFF;
|
| 762 |
-
setUint16(sample);
|
| 763 |
}
|
| 764 |
-
|
| 765 |
-
|
| 766 |
-
function setUint16(data) {
|
| 767 |
-
view.setUint16(pos, data, true);
|
| 768 |
-
pos += 2;
|
| 769 |
-
}
|
| 770 |
-
function setUint32(data) {
|
| 771 |
-
view.setUint32(pos, data, true);
|
| 772 |
-
pos += 4;
|
| 773 |
-
}
|
| 774 |
-
return bufferArray;
|
| 775 |
-
}
|
| 776 |
|
| 777 |
-
|
| 778 |
-
|
| 779 |
-
|
| 780 |
-
|
| 781 |
-
} else if (isAutoListening) {
|
| 782 |
-
toggleContinuousListening();
|
| 783 |
-
} else {
|
| 784 |
-
startListening();
|
| 785 |
-
}
|
| 786 |
-
});
|
| 787 |
|
| 788 |
-
|
|
|
|
| 789 |
|
| 790 |
-
|
| 791 |
-
|
| 792 |
-
|
| 793 |
-
|
| 794 |
-
}, 500);
|
| 795 |
}
|
| 796 |
-
});
|
| 797 |
-
|
| 798 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 799 |
updateStatus("Tap to listen", "bi-info-circle");
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 800 |
</script>
|
| 801 |
</body>
|
| 802 |
</html>
|
|
|
|
| 442 |
<!-- Bootstrap JS Bundle -->
|
| 443 |
<script src="https://cdn.jsdelivr.net/npm/[email protected]/dist/js/bootstrap.bundle.min.js"></script>
|
| 444 |
<script>
|
| 445 |
+
// DOM Elements
|
| 446 |
+
const listenBall = document.getElementById("listenBall");
|
| 447 |
+
const statusMessage = document.getElementById("statusMessage");
|
| 448 |
+
const audioPlayer = document.getElementById("audioPlayer");
|
| 449 |
+
const conversationArea = document.getElementById("conversationArea");
|
| 450 |
+
const emptyState = document.getElementById("emptyState");
|
| 451 |
+
const statusBadge = document.getElementById("statusBadge");
|
| 452 |
+
|
| 453 |
+
// Global variables
|
| 454 |
+
let mediaRecorder;
|
| 455 |
+
let audioChunks = [];
|
| 456 |
+
let audioStream;
|
| 457 |
+
let chatHistory = [];
|
| 458 |
+
let isListening = false;
|
| 459 |
+
let isAutoListening = false;
|
| 460 |
+
let silenceDetectionInterval;
|
| 461 |
+
let activityDetectionInterval;
|
| 462 |
+
let lastAudioLevel = 0;
|
| 463 |
+
let silenceCounter = 0;
|
| 464 |
+
let activityCounter = 0;
|
| 465 |
+
let currentUserGroup = null;
|
| 466 |
+
let currentAssistantGroup = null;
|
| 467 |
+
let audioContext;
|
| 468 |
+
let analyzer;
|
| 469 |
+
let isProcessing = false;
|
| 470 |
+
|
| 471 |
+
// Constants
|
| 472 |
+
const SILENCE_THRESHOLD = 15;
|
| 473 |
+
const ACTIVITY_THRESHOLD = 20;
|
| 474 |
+
const MIN_ACTIVITY_DURATION = 5; // Minimum counts of activity before recording
|
| 475 |
+
const MAX_SILENCE_DURATION = 15; // Maximum counts of silence before stopping
|
| 476 |
+
const MAX_RECORDING_DURATION = 8000; // Maximum recording duration in ms
|
| 477 |
+
const COOLDOWN_PERIOD = 1000; // Cooldown between recordings
|
| 478 |
+
|
| 479 |
+
// Functions
|
| 480 |
+
function updateStatus(message, icon = "bi-info-circle") {
|
| 481 |
+
statusMessage.textContent = message;
|
| 482 |
+
statusBadge.querySelector("i").className = `bi ${icon}`;
|
| 483 |
+
}
|
| 484 |
+
|
| 485 |
+
function addMessageToChat(content, sender) {
|
| 486 |
+
// Hide empty state if it's visible
|
| 487 |
+
if (!emptyState.classList.contains("d-none")) {
|
| 488 |
+
emptyState.classList.add("d-none");
|
| 489 |
+
}
|
| 490 |
+
|
| 491 |
+
const now = new Date();
|
| 492 |
+
const timeString = now.toLocaleTimeString([], { hour: '2-digit', minute: '2-digit' });
|
| 493 |
+
|
| 494 |
+
// Create new message group or use existing one based on sender
|
| 495 |
+
let messageGroup;
|
| 496 |
+
|
| 497 |
+
if (sender === 'user') {
|
| 498 |
+
if (!currentUserGroup || (currentAssistantGroup && currentAssistantGroup.classList.contains("assistant"))) {
|
| 499 |
+
currentUserGroup = document.createElement("div");
|
| 500 |
+
currentUserGroup.className = "message-group user";
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 501 |
|
| 502 |
+
const avatar = document.createElement("div");
|
| 503 |
+
avatar.className = "message-avatar";
|
| 504 |
+
avatar.innerHTML = "<i class='bi bi-person'></i>";
|
| 505 |
+
currentUserGroup.appendChild(avatar);
|
| 506 |
|
| 507 |
+
conversationArea.appendChild(currentUserGroup);
|
| 508 |
+
}
|
| 509 |
+
messageGroup = currentUserGroup;
|
| 510 |
+
currentAssistantGroup = null;
|
| 511 |
+
} else {
|
| 512 |
+
if (!currentAssistantGroup || (currentUserGroup && currentUserGroup.classList.contains("user"))) {
|
| 513 |
+
currentAssistantGroup = document.createElement("div");
|
| 514 |
+
currentAssistantGroup.className = "message-group assistant";
|
| 515 |
|
| 516 |
+
const avatar = document.createElement("div");
|
| 517 |
+
avatar.className = "message-avatar";
|
| 518 |
+
avatar.innerHTML = "<i class='bi bi-robot'></i>";
|
| 519 |
+
currentAssistantGroup.appendChild(avatar);
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 520 |
|
| 521 |
+
conversationArea.appendChild(currentAssistantGroup);
|
| 522 |
+
}
|
| 523 |
+
messageGroup = currentAssistantGroup;
|
| 524 |
+
currentUserGroup = null;
|
| 525 |
+
}
|
| 526 |
+
|
| 527 |
+
// Create message element
|
| 528 |
+
const messageDiv = document.createElement("div");
|
| 529 |
+
messageDiv.className = `message ${sender}`;
|
| 530 |
+
messageDiv.textContent = content;
|
| 531 |
+
|
| 532 |
+
const timestamp = document.createElement("div");
|
| 533 |
+
timestamp.className = "time-stamp";
|
| 534 |
+
timestamp.textContent = timeString;
|
| 535 |
+
|
| 536 |
+
messageGroup.appendChild(messageDiv);
|
| 537 |
+
messageGroup.appendChild(timestamp);
|
| 538 |
+
|
| 539 |
+
// Scroll to bottom
|
| 540 |
+
conversationArea.scrollTop = conversationArea.scrollHeight;
|
| 541 |
+
|
| 542 |
+
// Add to chat history
|
| 543 |
+
chatHistory.push({
|
| 544 |
+
role: sender === 'user' ? 'user' : 'assistant',
|
| 545 |
+
content: content
|
| 546 |
+
});
|
| 547 |
+
}
|
| 548 |
+
|
| 549 |
+
async function setupAudioAnalysis() {
|
| 550 |
+
if (audioContext) {
|
| 551 |
+
audioContext.close();
|
| 552 |
+
}
|
| 553 |
+
|
| 554 |
+
audioContext = new AudioContext();
|
| 555 |
+
const source = audioContext.createMediaStreamSource(audioStream);
|
| 556 |
+
analyzer = audioContext.createAnalyser();
|
| 557 |
+
analyzer.fftSize = 256;
|
| 558 |
+
source.connect(analyzer);
|
| 559 |
+
|
| 560 |
+
const bufferLength = analyzer.frequencyBinCount;
|
| 561 |
+
const dataArray = new Uint8Array(bufferLength);
|
| 562 |
+
|
| 563 |
+
return { analyzer, dataArray, bufferLength };
|
| 564 |
+
}
|
| 565 |
+
|
| 566 |
+
function startContinuousListening() {
|
| 567 |
+
if (!audioStream) return;
|
| 568 |
+
|
| 569 |
+
// Set up audio analysis
|
| 570 |
+
setupAudioAnalysis().then(({ analyzer, dataArray, bufferLength }) => {
|
| 571 |
+
// Start monitoring audio levels
|
| 572 |
+
clearInterval(activityDetectionInterval);
|
| 573 |
+
activityDetectionInterval = setInterval(() => {
|
| 574 |
+
if (isProcessing) return;
|
| 575 |
|
| 576 |
+
analyzer.getByteFrequencyData(dataArray);
|
|
|
|
| 577 |
|
| 578 |
+
// Calculate average audio level
|
| 579 |
+
let sum = 0;
|
| 580 |
+
for (let i = 0; i < bufferLength; i++) {
|
| 581 |
+
sum += dataArray[i];
|
| 582 |
+
}
|
| 583 |
+
const avg = sum / bufferLength;
|
| 584 |
+
lastAudioLevel = avg;
|
| 585 |
|
| 586 |
+
// Detect significant sound
|
| 587 |
+
if (avg > ACTIVITY_THRESHOLD) {
|
| 588 |
+
activityCounter++;
|
| 589 |
+
silenceCounter = 0;
|
| 590 |
|
| 591 |
+
// If we have enough continuous activity and not already listening, start recording
|
| 592 |
+
if (activityCounter >= MIN_ACTIVITY_DURATION && !isListening && !isProcessing) {
|
| 593 |
+
startRecording();
|
|
|
|
| 594 |
}
|
| 595 |
+
} else {
|
| 596 |
+
activityCounter = 0;
|
| 597 |
+
}
|
| 598 |
+
}, 100);
|
| 599 |
+
});
|
| 600 |
+
}
|
| 601 |
+
|
| 602 |
+
function monitorSilenceDuringRecording() {
|
| 603 |
+
if (!audioStream || !isListening) return;
|
| 604 |
+
|
| 605 |
+
clearInterval(silenceDetectionInterval);
|
| 606 |
+
silenceDetectionInterval = setInterval(() => {
|
| 607 |
+
if (!isListening) {
|
| 608 |
+
clearInterval(silenceDetectionInterval);
|
| 609 |
+
return;
|
| 610 |
+
}
|
| 611 |
+
|
| 612 |
+
analyzer.getByteFrequencyData(new Uint8Array(analyzer.frequencyBinCount));
|
| 613 |
+
|
| 614 |
+
// Calculate average audio level
|
| 615 |
+
let sum = 0;
|
| 616 |
+
for (let i = 0; i < analyzer.frequencyBinCount; i++) {
|
| 617 |
+
sum += dataArray[i];
|
| 618 |
+
}
|
| 619 |
+
const avg = sum / analyzer.frequencyBinCount;
|
| 620 |
+
|
| 621 |
+
// If silent, increment counter
|
| 622 |
+
if (avg < SILENCE_THRESHOLD) {
|
| 623 |
+
silenceCounter++;
|
| 624 |
+
if (silenceCounter >= MAX_SILENCE_DURATION) {
|
| 625 |
+
stopRecording();
|
| 626 |
+
}
|
| 627 |
+
} else {
|
| 628 |
+
silenceCounter = 0;
|
| 629 |
}
|
| 630 |
+
}, 100);
|
| 631 |
+
}
|
| 632 |
+
|
| 633 |
+
async function startRecording() {
|
| 634 |
+
if (isListening || isProcessing) return;
|
| 635 |
+
|
| 636 |
+
try {
|
| 637 |
+
// Reset counters
|
| 638 |
+
silenceCounter = 0;
|
| 639 |
+
|
| 640 |
+
// Start recording
|
| 641 |
+
mediaRecorder = new MediaRecorder(audioStream, { mimeType: "audio/webm" });
|
| 642 |
+
audioChunks = [];
|
| 643 |
|
| 644 |
+
mediaRecorder.ondataavailable = event => audioChunks.push(event.data);
|
| 645 |
+
|
| 646 |
+
mediaRecorder.onstop = async () => {
|
| 647 |
+
if (audioChunks.length === 0) {
|
| 648 |
+
updateState("idle");
|
| 649 |
+
isProcessing = false;
|
| 650 |
+
return;
|
| 651 |
+
}
|
| 652 |
+
|
| 653 |
+
isProcessing = true;
|
| 654 |
+
updateState("processing");
|
| 655 |
|
| 656 |
try {
|
| 657 |
+
const audioBlob = new Blob(audioChunks, { type: "audio/webm" });
|
| 658 |
+
const wavBlob = await convertWebMToWav(audioBlob);
|
|
|
|
|
|
|
| 659 |
|
| 660 |
+
// Create form data with the audio and chat history
|
| 661 |
+
const formData = new FormData();
|
| 662 |
+
formData.append("file", wavBlob, "recording.wav");
|
| 663 |
+
formData.append("chat_history", JSON.stringify(chatHistory));
|
| 664 |
|
| 665 |
+
// Send to the continuous-chat endpoint using root-relative path
|
| 666 |
+
const response = await fetch("/continuous-chat/", {
|
| 667 |
+
method: "POST",
|
| 668 |
+
body: formData
|
| 669 |
+
});
|
| 670 |
|
| 671 |
+
if (response.ok) {
|
| 672 |
+
const userMessage = response.headers.get("X-User-Message") || "No user message";
|
| 673 |
+
const llmResponse = response.headers.get("X-LLM-Response") || "No response";
|
|
|
|
|
|
|
| 674 |
|
| 675 |
+
// Add messages to chat
|
| 676 |
+
addMessageToChat(userMessage, 'user');
|
| 677 |
+
addMessageToChat(llmResponse, 'assistant');
|
| 678 |
|
| 679 |
+
// Get audio response and play it
|
| 680 |
+
const audioData = await response.blob();
|
| 681 |
+
audioPlayer.src = URL.createObjectURL(audioData);
|
| 682 |
+
audioPlayer.play();
|
| 683 |
+
|
| 684 |
+
updateState("idle");
|
| 685 |
+
updateStatus("Listening for sound", "bi-broadcast");
|
| 686 |
+
} else {
|
| 687 |
+
updateState("idle");
|
| 688 |
+
updateStatus("Error processing audio", "bi-exclamation-triangle");
|
| 689 |
+
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 690 |
} catch (error) {
|
| 691 |
+
console.error("Error:", error);
|
|
|
|
| 692 |
updateState("idle");
|
| 693 |
+
updateStatus("Error processing audio", "bi-exclamation-triangle");
|
| 694 |
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 695 |
|
| 696 |
+
// Set a cooldown before allowing the next recording
|
| 697 |
+
setTimeout(() => {
|
| 698 |
+
isProcessing = false;
|
| 699 |
+
if (isAutoListening) {
|
| 700 |
+
updateStatus("Listening for sound", "bi-broadcast");
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 701 |
}
|
| 702 |
+
}, COOLDOWN_PERIOD);
|
| 703 |
+
};
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 704 |
|
| 705 |
+
mediaRecorder.start();
|
| 706 |
+
isListening = true;
|
| 707 |
+
updateState("listening");
|
| 708 |
+
updateStatus("Recording...", "bi-ear");
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 709 |
|
| 710 |
+
// Monitor for silence during recording
|
| 711 |
+
monitorSilenceDuringRecording();
|
| 712 |
|
| 713 |
+
// Set max recording duration
|
| 714 |
+
setTimeout(() => {
|
| 715 |
+
if (mediaRecorder && mediaRecorder.state === "recording") {
|
| 716 |
+
stopRecording();
|
|
|
|
| 717 |
}
|
| 718 |
+
}, MAX_RECORDING_DURATION);
|
| 719 |
+
|
| 720 |
+
} catch (error) {
|
| 721 |
+
console.error("Error starting recording:", error);
|
| 722 |
+
updateState("idle");
|
| 723 |
+
updateStatus("Recording error", "bi-exclamation-triangle");
|
| 724 |
+
isListening = false;
|
| 725 |
+
isProcessing = false;
|
| 726 |
+
}
|
| 727 |
+
}
|
| 728 |
+
|
| 729 |
+
function stopRecording() {
|
| 730 |
+
if (!isListening) return;
|
| 731 |
+
|
| 732 |
+
clearInterval(silenceDetectionInterval);
|
| 733 |
+
|
| 734 |
+
if (mediaRecorder && mediaRecorder.state === "recording") {
|
| 735 |
+
mediaRecorder.stop();
|
| 736 |
+
}
|
| 737 |
+
|
| 738 |
+
isListening = false;
|
| 739 |
+
updateStatus("Processing...", "bi-arrow-repeat");
|
| 740 |
+
}
|
| 741 |
+
|
| 742 |
+
function updateState(state) {
|
| 743 |
+
listenBall.classList.remove("listening", "processing");
|
| 744 |
+
|
| 745 |
+
if (state === "listening") {
|
| 746 |
+
listenBall.classList.add("listening");
|
| 747 |
+
listenBall.innerHTML = `
|
| 748 |
+
<div class="sound-wave"></div>
|
| 749 |
+
<div class="sound-wave" style="animation-delay: 0.5s"></div>
|
| 750 |
+
<div class="sound-wave" style="animation-delay: 1s"></div>
|
| 751 |
+
<i class="bi bi-soundwave"></i>
|
| 752 |
+
`;
|
| 753 |
+
} else if (state === "processing") {
|
| 754 |
+
listenBall.classList.add("processing");
|
| 755 |
+
listenBall.innerHTML = `<i class="bi bi-arrow-repeat"></i>`;
|
| 756 |
+
} else {
|
| 757 |
+
listenBall.innerHTML = `<i class="bi bi-soundwave"></i>`;
|
| 758 |
+
}
|
| 759 |
+
}
|
| 760 |
+
|
| 761 |
+
async function toggleContinuousListening() {
|
| 762 |
+
isAutoListening = !isAutoListening;
|
| 763 |
+
|
| 764 |
+
if (isAutoListening) {
|
| 765 |
+
try {
|
| 766 |
+
// Request microphone access if we don't have it
|
| 767 |
+
if (!audioStream) {
|
| 768 |
+
audioStream = await navigator.mediaDevices.getUserMedia({ audio: true });
|
| 769 |
+
}
|
| 770 |
+
|
| 771 |
+
updateStatus("Auto-listening active", "bi-broadcast");
|
| 772 |
+
startContinuousListening();
|
| 773 |
+
} catch (error) {
|
| 774 |
+
console.error("Error accessing microphone:", error);
|
| 775 |
+
updateStatus("Microphone access denied", "bi-mic-mute");
|
| 776 |
+
isAutoListening = false;
|
| 777 |
+
}
|
| 778 |
+
} else {
|
| 779 |
+
// Stop continuous listening
|
| 780 |
+
clearInterval(activityDetectionInterval);
|
| 781 |
+
clearInterval(silenceDetectionInterval);
|
| 782 |
updateStatus("Tap to listen", "bi-info-circle");
|
| 783 |
+
|
| 784 |
+
// If currently recording, stop it
|
| 785 |
+
if (isListening) {
|
| 786 |
+
stopRecording();
|
| 787 |
+
}
|
| 788 |
+
}
|
| 789 |
+
}
|
| 790 |
+
|
| 791 |
+
async function manualListening() {
|
| 792 |
+
if (isListening || isProcessing) return;
|
| 793 |
+
|
| 794 |
+
try {
|
| 795 |
+
// Request microphone access if we don't have it
|
| 796 |
+
if (!audioStream) {
|
| 797 |
+
audioStream = await navigator.mediaDevices.getUserMedia({ audio: true });
|
| 798 |
+
await setupAudioAnalysis();
|
| 799 |
+
}
|
| 800 |
+
|
| 801 |
+
startRecording();
|
| 802 |
+
} catch (error) {
|
| 803 |
+
console.error("Error accessing microphone:", error);
|
| 804 |
+
updateStatus("Microphone access denied", "bi-mic-mute");
|
| 805 |
+
}
|
| 806 |
+
}
|
| 807 |
+
|
| 808 |
+
async function convertWebMToWav(blob) {
|
| 809 |
+
return new Promise((resolve, reject) => {
|
| 810 |
+
try {
|
| 811 |
+
const reader = new FileReader();
|
| 812 |
+
reader.onload = function () {
|
| 813 |
+
const audioContext = new AudioContext();
|
| 814 |
+
audioContext.decodeAudioData(reader.result)
|
| 815 |
+
.then(buffer => {
|
| 816 |
+
const wavBuffer = audioBufferToWav(buffer);
|
| 817 |
+
resolve(new Blob([wavBuffer], { type: "audio/wav" }));
|
| 818 |
+
})
|
| 819 |
+
.catch(error => {
|
| 820 |
+
console.error("Error decoding audio data:", error);
|
| 821 |
+
reject(error);
|
| 822 |
+
});
|
| 823 |
+
};
|
| 824 |
+
reader.readAsArrayBuffer(blob);
|
| 825 |
+
} catch (error) {
|
| 826 |
+
console.error("Error in convertWebMToWav:", error);
|
| 827 |
+
reject(error);
|
| 828 |
+
}
|
| 829 |
+
});
|
| 830 |
+
}
|
| 831 |
+
|
| 832 |
+
function audioBufferToWav(buffer) {
|
| 833 |
+
let numOfChan = buffer.numberOfChannels,
|
| 834 |
+
length = buffer.length * numOfChan * 2 + 44,
|
| 835 |
+
bufferArray = new ArrayBuffer(length),
|
| 836 |
+
view = new DataView(bufferArray),
|
| 837 |
+
channels = [],
|
| 838 |
+
sampleRate = buffer.sampleRate,
|
| 839 |
+
offset = 0,
|
| 840 |
+
pos = 0;
|
| 841 |
+
setUint32(0x46464952); // "RIFF"
|
| 842 |
+
setUint32(length - 8);
|
| 843 |
+
setUint32(0x45564157); // "WAVE"
|
| 844 |
+
setUint32(0x20746d66); // "fmt " chunk
|
| 845 |
+
setUint32(16); // length = 16
|
| 846 |
+
setUint16(1); // PCM (uncompressed)
|
| 847 |
+
setUint16(numOfChan);
|
| 848 |
+
setUint32(sampleRate);
|
| 849 |
+
setUint32(sampleRate * 2 * numOfChan);
|
| 850 |
+
setUint16(numOfChan * 2);
|
| 851 |
+
setUint16(16); // bits per sample
|
| 852 |
+
setUint32(0x61746164); // "data" chunk
|
| 853 |
+
setUint32(length - pos - 4);
|
| 854 |
+
for (let i = 0; i < buffer.numberOfChannels; i++)
|
| 855 |
+
channels.push(buffer.getChannelData(i));
|
| 856 |
+
while (pos < length) {
|
| 857 |
+
for (let i = 0; i < numOfChan; i++) {
|
| 858 |
+
let sample = Math.max(-1, Math.min(1, channels[i][offset]));
|
| 859 |
+
sample = sample < 0 ? sample * 0x8000 : sample * 0x7FFF;
|
| 860 |
+
setUint16(sample);
|
| 861 |
+
}
|
| 862 |
+
offset++;
|
| 863 |
+
}
|
| 864 |
+
function setUint16(data) {
|
| 865 |
+
view.setUint16(pos, data, true);
|
| 866 |
+
pos += 2;
|
| 867 |
+
}
|
| 868 |
+
function setUint32(data) {
|
| 869 |
+
view.setUint32(pos, data, true);
|
| 870 |
+
pos += 4;
|
| 871 |
+
}
|
| 872 |
+
return bufferArray;
|
| 873 |
+
}
|
| 874 |
+
|
| 875 |
+
// Event Listeners
|
| 876 |
+
listenBall.addEventListener("click", () => {
|
| 877 |
+
if (isAutoListening) {
|
| 878 |
+
toggleContinuousListening(); // Turn off auto mode
|
| 879 |
+
} else {
|
| 880 |
+
if (isListening) {
|
| 881 |
+
stopRecording(); // Stop manual recording
|
| 882 |
+
} else {
|
| 883 |
+
manualListening(); // Start manual recording
|
| 884 |
+
}
|
| 885 |
+
}
|
| 886 |
+
});
|
| 887 |
+
|
| 888 |
+
listenBall.addEventListener("dblclick", toggleContinuousListening);
|
| 889 |
+
|
| 890 |
+
audioPlayer.addEventListener("ended", () => {
|
| 891 |
+
if (isAutoListening && !isProcessing) {
|
| 892 |
+
updateStatus("Listening for sound", "bi-broadcast");
|
| 893 |
+
}
|
| 894 |
+
});
|
| 895 |
+
|
| 896 |
+
// Initialize
|
| 897 |
+
updateStatus("Tap to listen, double-tap for auto mode", "bi-info-circle");
|
| 898 |
+
|
| 899 |
+
// Cleanup function for page unload
|
| 900 |
+
window.addEventListener('beforeunload', () => {
|
| 901 |
+
if (audioStream) {
|
| 902 |
+
audioStream.getTracks().forEach(track => track.stop());
|
| 903 |
+
}
|
| 904 |
+
if (audioContext) {
|
| 905 |
+
audioContext.close();
|
| 906 |
+
}
|
| 907 |
+
clearInterval(silenceDetectionInterval);
|
| 908 |
+
clearInterval(activityDetectionInterval);
|
| 909 |
+
});
|
| 910 |
</script>
|
| 911 |
</body>
|
| 912 |
</html>
|