Stream Whisper uploads with curl so large audio does not OOM.

2026-08-13 20:29:59 +02:00
parent 1898be4def
commit e77a18c49f
7 changed files with 488 additions and 42 deletions
@@ -0,0 +1,120 @@
+<?php
+
+namespace App\Services;
+
+use Closure;
+use CURLFile;
+use RuntimeException;
+
+/**
+ * Streams large audio uploads to Whisper with curl (disk → network),
+ * and delivers response body chunks without buffering the request in PHP.
+ *
+ * Laravel Http's stream => true uses Guzzle's StreamHandler, which casts the
+ * multipart body to a string and OOMs on big recordings.
+ */
+class LocalWhisperCurlClient
+{
+    /**
+     * POST /audio/transcriptions and stream the response body.
+     *
+     * @param  Closure(string): bool  $onBodyChunk  Return false to abort the transfer.
+     * @param  Closure(string): void|null  $onContentType  Invoked when the response Content-Type header arrives.
+     * @return array{status: int, content_type: string}
+     */
+    public function streamTranscription(
+        string $url,
+        string $apiKey,
+        string $path,
+        string $filename,
+        string $mimeType,
+        string $model,
+        int $timeout,
+        Closure $onBodyChunk,
+        ?Closure $onContentType = null,
+    ): array {
+        if (! function_exists('curl_init')) {
+            throw new RuntimeException('The curl PHP extension is required for local Whisper transcription.');
+        }
+
+        $contentType = '';
+        $abort = false;
+
+        $handle = curl_init($url);
+
+        if ($handle === false) {
+            throw new RuntimeException('Unable to initialize curl for local Whisper.');
+        }
+
+        try {
+            curl_setopt_array($handle, [
+                CURLOPT_POST => true,
+                CURLOPT_HTTPHEADER => [
+                    'Authorization: Bearer '.$apiKey,
+                    'Accept: application/json, text/event-stream',
+                ],
+                CURLOPT_POSTFIELDS => [
+                    'file' => new CURLFile($path, $mimeType, $filename),
+                    'model' => $model,
+                    'response_format' => 'verbose_json',
+                    'stream' => 'true',
+                    'without_timestamps' => 'false',
+                    'timestamp_granularities[]' => 'word',
+                ],
+                CURLOPT_RETURNTRANSFER => false,
+                CURLOPT_HEADER => false,
+                CURLOPT_TIMEOUT => max(1, $timeout),
+                CURLOPT_CONNECTTIMEOUT => 10,
+                CURLOPT_HEADERFUNCTION => static function ($ch, string $header) use (&$contentType, $onContentType): int {
+                    if (stripos($header, 'Content-Type:') === 0) {
+                        $contentType = trim(substr($header, strlen('Content-Type:')));
+
+                        if ($onContentType !== null) {
+                            $onContentType($contentType);
+                        }
+                    }
+
+                    return strlen($header);
+                },
+                CURLOPT_WRITEFUNCTION => static function ($ch, string $chunk) use ($onBodyChunk, &$abort): int {
+                    if ($abort) {
+                        return 0;
+                    }
+
+                    if ($onBodyChunk($chunk) === false) {
+                        $abort = true;
+
+                        return 0;
+                    }
+
+                    return strlen($chunk);
+                },
+            ]);
+
+            $ok = curl_exec($handle);
+            $errno = curl_errno($handle);
+            $error = curl_error($handle);
+            $status = (int) curl_getinfo($handle, CURLINFO_RESPONSE_CODE);
+
+            if ($abort) {
+                return [
+                    'status' => $status > 0 ? $status : 499,
+                    'content_type' => $contentType,
+                ];
+            }
+
+            if ($ok === false && $errno !== 0) {
+                throw new RuntimeException(
+                    'Local transcription request failed: '.$error.' (curl '.$errno.')'
+                );
+            }
+
+            return [
+                'status' => $status,
+                'content_type' => $contentType,
+            ];
+        } finally {
+            curl_close($handle);
+        }
+    }
+}