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 +1,119 @@
+<?php
+
+namespace Tests\Feature;
+
+use App\Models\Recording;
+use App\Models\User;
+use App\Services\TranscriptionService;
+use Illuminate\Foundation\Testing\RefreshDatabase;
+use Illuminate\Support\Facades\Process;
+use Illuminate\Support\Facades\Storage;
+use Tests\TestCase;
+
+class LocalWhisperCurlTransportTest extends TestCase
+{
+    use RefreshDatabase;
+
+    public function test_curl_transport_streams_sse_from_real_http_endpoint(): void
+    {
+        if (! function_exists('curl_init')) {
+            $this->markTestSkipped('curl extension required');
+        }
+
+        Storage::fake('local');
+        Storage::disk('local')->put('recordings/large.mp3', str_repeat('A', 2_000_000));
+
+        $stub = base_path('tests/fixtures/whisper-sse-stub.php');
+        $host = '127.0.0.1';
+        $port = $this->reservePort($host);
+        $docRoot = base_path('tests/fixtures');
+
+        $server = Process::timeout(30)
+            ->path($docRoot)
+            ->start([
+                PHP_BINARY,
+                '-S',
+                "{$host}:{$port}",
+                'whisper-sse-stub.php',
+            ]);
+
+        try {
+            $this->waitForServer($host, $port);
+
+            config([
+                'ai.local_whisper_transport' => 'curl',
+                'ai.providers.local-whisper.url' => "http://{$host}:{$port}/v1",
+                'ai.providers.local-whisper.key' => 'test-key',
+            ]);
+
+            $user = User::factory()->create();
+            $recording = Recording::query()->create([
+                'user_id' => $user->id,
+                'title' => 'Curl transport',
+                'original_filename' => 'large.mp3',
+                'file_path' => 'recordings/large.mp3',
+                'file_size_bytes' => 2_000_000,
+                'duration_seconds' => 40,
+                'transcription_status' => 'processing',
+                'transcription_driver' => 'local',
+            ]);
+
+            $partials = [];
+
+            $text = app(TranscriptionService::class)->transcribe(
+                $recording,
+                function (string $message, int $percent, ?string $partial = null) use (&$partials): void {
+                    if ($partial !== null) {
+                        $partials[] = $partial;
+                    }
+                },
+            );
+
+            $this->assertSame('Hello from whisper.', $text);
+            $this->assertContains('Hello ', $partials);
+            $this->assertContains('Hello from whisper.', $partials);
+        } finally {
+            $server->signal(SIGTERM);
+            $server->wait();
+        }
+    }
+
+    private function reservePort(string $host): int
+    {
+        $socket = stream_socket_server("tcp://{$host}:0");
+
+        if ($socket === false) {
+            $this->fail('Unable to reserve a local TCP port for the Whisper stub.');
+        }
+
+        $name = stream_socket_get_name($socket, false);
+        fclose($socket);
+
+        if (! is_string($name) || ! str_contains($name, ':')) {
+            $this->fail('Unable to determine reserved port.');
+        }
+
+        return (int) substr($name, strrpos($name, ':') + 1);
+    }
+
+    private function waitForServer(string $host, int $port): void
+    {
+        $deadline = microtime(true) + 5;
+
+        while (microtime(true) < $deadline) {
+            $errno = 0;
+            $errstr = '';
+            $fp = @fsockopen($host, $port, $errno, $errstr, 0.2);
+
+            if (is_resource($fp)) {
+                fclose($fp);
+
+                return;
+            }
+
+            usleep(50_000);
+        }
+
+        $this->fail("Whisper stub server did not start on {$host}:{$port}");
+    }
+}