fix(bluesky): harden video upload per review + retry transient transcode failures

Address review feedback and a transient failure mode seen in production:

- Guard tempnam()/fopen() returning false (bail cleanly instead of a
  TypeError later) and guard the finally-block unlink() with is_string().
- pollVideoJob() now bails on a non-2xx getJobStatus (e.g. expired token)
  instead of sleeping to the timeout and masking the real error.
- resolvePdsEndpoint() maps did:web colon segments to host + path per the
  did:web spec (did:web:example.com:user:alice -> .../user/alice/did.json)
  instead of treating the whole remainder as a host.
- The video service occasionally fails a job transiently
  (JOB_STATE_FAILED "Failed to process video") for valid input, so the
  upload+poll is retried up to VIDEO_UPLOAD_ATTEMPTS times (split into a
  new attemptVideoUpload() helper; the upload service-auth token is minted
  once and reused across attempts).

Adds a test covering retry-after-transient-failure. Full suite (23) green.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
This commit is contained in:
Patrick Schrall 2026-06-23 08:32:27 +02:00
parent 43de7372a6
commit 7f648f8d4d
2 changed files with 172 additions and 44 deletions

View file

@ -20,6 +20,9 @@ class BlueskyPublisher
{
use HasSocialHttpClient;
/** Times to retry a transiently-failing video transcode before giving up. */
private const VIDEO_UPLOAD_ATTEMPTS = 3;
public function publish(PostPlatform $postPlatform): array
{
$this->validateContentLength($postPlatform);
@ -206,6 +209,12 @@ private function uploadVideo(SocialAccount $account, string $service, string $ur
$tempFile = tempnam(sys_get_temp_dir(), 'bsky_video_');
if ($tempFile === false) {
Log::error('Bluesky could not create temp file for video upload', ['url' => $url]);
return null;
}
try {
$downloadResponse = Http::withOptions(['sink' => $tempFile])->timeout(600)->get($url);
@ -240,55 +249,32 @@ private function uploadVideo(SocialAccount $account, string $service, string $ur
// The upload token is consumed by the video service to write the
// blob back to the user's PDS, so its audience is the PDS itself.
// Minted once (valid 30 min) and reused across retries below.
$uploadToken = $this->getServiceAuth($account, $pds, "did:web:{$pdsHost}", BlueskyLexicon::UPLOAD_BLOB);
if ($uploadToken === null) {
return null;
}
$name = bin2hex(random_bytes(8)).'.mp4';
$uploadUrl = "{$videoService}/xrpc/".BlueskyLexicon::VIDEO_UPLOAD
.'?did='.rawurlencode($did).'&name='.rawurlencode($name);
// The transcoder occasionally fails a job transiently
// (JOB_STATE_FAILED "Failed to process video") even for valid input,
// so retry the upload+poll a couple of times before giving up.
for ($attempt = 1; $attempt <= self::VIDEO_UPLOAD_ATTEMPTS; $attempt++) {
$blob = $this->attemptVideoUpload($account, $pds, $videoService, $did, $uploadToken, $tempFile);
$stream = fopen($tempFile, 'r');
if ($blob !== null) {
return $blob;
}
$response = $this->socialHttp()->withToken($uploadToken)
->withHeaders(['Content-Type' => 'video/mp4'])
->withBody($stream, 'video/mp4')
->post($uploadUrl);
if (is_resource($stream)) {
fclose($stream);
if ($attempt < self::VIDEO_UPLOAD_ATTEMPTS) {
Log::warning('Bluesky video upload attempt failed, retrying', [
'attempt' => $attempt,
'url' => $url,
]);
}
}
$jobStatus = data_get($response->json(), 'jobStatus');
// A re-upload of identical bytes returns 409 with the already
// finished job, whose blob we can embed directly.
if ($response->failed() && $response->status() !== 409) {
Log::error('Bluesky video upload failed', [
'status' => $response->status(),
'body' => $this->redactResponseBody($response->body()),
]);
return null;
}
if (data_get($jobStatus, 'blob')) {
return data_get($jobStatus, 'blob');
}
$jobId = data_get($jobStatus, 'jobId');
if (! is_string($jobId) || $jobId === '') {
Log::error('Bluesky video upload returned no jobId', [
'body' => $this->redactResponseBody($response->body()),
]);
return null;
}
return $this->pollVideoJob($account, $pds, $videoService, $jobId);
return null;
} catch (Exception $e) {
Log::error('Bluesky video upload exception', [
'error' => $e->getMessage(),
@ -297,10 +283,72 @@ private function uploadVideo(SocialAccount $account, string $service, string $ur
return null;
} finally {
@unlink($tempFile);
if (is_string($tempFile)) {
@unlink($tempFile);
}
}
}
/**
* A single upload-and-poll attempt against the video service. Returns the
* processed blob, or null if this attempt failed (so the caller can retry).
*/
private function attemptVideoUpload(SocialAccount $account, string $pds, string $videoService, string $did, string $uploadToken, string $tempFile): ?array
{
$stream = fopen($tempFile, 'r');
if ($stream === false) {
Log::error('Bluesky could not open video file for upload', ['file' => $tempFile]);
return null;
}
$name = bin2hex(random_bytes(8)).'.mp4';
$uploadUrl = "{$videoService}/xrpc/".BlueskyLexicon::VIDEO_UPLOAD
.'?did='.rawurlencode($did).'&name='.rawurlencode($name);
$response = $this->socialHttp()->withToken($uploadToken)
->withHeaders(['Content-Type' => 'video/mp4'])
->withBody($stream, 'video/mp4')
->post($uploadUrl);
if (is_resource($stream)) {
fclose($stream);
}
// uploadVideo returns the jobStatus object directly (top-level),
// while getJobStatus wraps it under a `jobStatus` key — accept both.
$body = $response->json();
$jobStatus = data_get($body, 'jobStatus') ?: $body;
// A re-upload of identical bytes returns 409 with the already
// finished job, whose blob we can embed directly.
if ($response->failed() && $response->status() !== 409) {
Log::error('Bluesky video upload failed', [
'status' => $response->status(),
'body' => $this->redactResponseBody($response->body()),
]);
return null;
}
if (data_get($jobStatus, 'blob')) {
return data_get($jobStatus, 'blob');
}
$jobId = data_get($jobStatus, 'jobId');
if (! is_string($jobId) || $jobId === '') {
Log::error('Bluesky video upload returned no jobId', [
'body' => $this->redactResponseBody($response->body()),
]);
return null;
}
return $this->pollVideoJob($account, $pds, $videoService, $jobId);
}
/**
* Poll the video service until the transcode job finishes, then return its
* blob. Returns null if the job fails or never completes in time.
@ -322,7 +370,20 @@ private function pollVideoJob(SocialAccount $account, string $pds, string $video
$response = $this->socialHttp()->withToken($jobToken)
->get($statusUrl, ['jobId' => $jobId]);
$jobStatus = data_get($response->json(), 'jobStatus');
// Bail on a hard error (e.g. expired/invalid token) instead of
// sleeping to the timeout and masking the real failure.
if ($response->failed()) {
Log::error('Bluesky getJobStatus failed', [
'jobId' => $jobId,
'status' => $response->status(),
'body' => $this->redactResponseBody($response->body()),
]);
return null;
}
$body = $response->json();
$jobStatus = data_get($body, 'jobStatus') ?: $body;
$state = data_get($jobStatus, 'state');
if ($state === 'JOB_STATE_COMPLETED' && data_get($jobStatus, 'blob')) {
@ -396,8 +457,12 @@ private function resolvePdsEndpoint(SocialAccount $account, string $service): st
$directory = (string) config('trypost.platforms.bluesky.plc_directory');
$docUrl = "{$directory}/".rawurlencode($did);
} elseif (str_starts_with($did, 'did:web:')) {
$host = substr($did, strlen('did:web:'));
$docUrl = "https://{$host}/.well-known/did.json";
// Per the did:web spec, colon-separated segments map to a host
// plus optional path; a bare host uses /.well-known/did.json.
$segments = array_map('rawurldecode', explode(':', substr($did, strlen('did:web:'))));
$host = array_shift($segments);
$path = $segments === [] ? '/.well-known/did.json' : '/'.implode('/', $segments).'/did.json';
$docUrl = "https://{$host}{$path}";
}
if ($docUrl !== null) {

View file

@ -619,7 +619,8 @@ function fakeBlueskyVideoPipeline(string $jobState = 'JOB_STATE_COMPLETED', bool
}
if (str_contains($url, 'app.bsky.video.uploadVideo')) {
return Http::response(['jobStatus' => ['jobId' => 'job-123', 'state' => 'JOB_STATE_CREATED']], 200);
// uploadVideo returns the jobStatus object directly (not wrapped).
return Http::response(['jobId' => 'job-123', 'did' => 'did:plc:testuser123', 'state' => 'JOB_STATE_CREATED'], 200);
}
if (str_contains($url, 'app.bsky.video.getJobStatus')) {
@ -721,3 +722,65 @@ function fakeBlueskyVideoPipeline(string $jobState = 'JOB_STATE_COMPLETED', bool
&& ! isset($request['record']['embed']);
});
});
test('bluesky publisher retries a transient video transcode failure', function () {
$this->post->update([
'media' => [[
'id' => 'test-video-id',
'path' => 'media/2026-01/test-video.mp4',
'url' => 'https://example.com/media/2026-01/test-video.mp4',
'mime_type' => 'video/mp4',
'original_filename' => 'test.mp4',
]],
]);
$jobCalls = 0;
Http::fake(function ($request) use (&$jobCalls) {
$url = $request->url();
if (str_contains($url, 'plc.directory')) {
return Http::response(['service' => [[
'id' => '#atproto_pds',
'type' => 'AtprotoPersonalDataServer',
'serviceEndpoint' => 'https://pds.example.host',
]]], 200);
}
if (str_contains($url, 'getServiceAuth')) {
return Http::response(['token' => 'service-auth-token'], 200);
}
if (str_contains($url, 'app.bsky.video.uploadVideo')) {
return Http::response(['jobId' => 'job-123', 'state' => 'JOB_STATE_CREATED'], 200);
}
if (str_contains($url, 'app.bsky.video.getJobStatus')) {
$jobCalls++;
// First attempt's job fails transiently; the retry succeeds.
if ($jobCalls === 1) {
return Http::response(['jobStatus' => ['jobId' => 'job-123', 'state' => 'JOB_STATE_FAILED', 'message' => 'transient']], 200);
}
return Http::response(['jobStatus' => [
'jobId' => 'job-123', 'state' => 'JOB_STATE_COMPLETED',
'blob' => ['$type' => 'blob', 'ref' => ['$link' => 'bafretry456'], 'mimeType' => 'video/mp4', 'size' => 2048],
]], 200);
}
if (str_contains($url, 'createRecord')) {
return Http::response(['uri' => 'at://did:plc:testuser123/app.bsky.feed.post/3retry', 'cid' => 'bafretry'], 200);
}
return Http::response(str_repeat('v', 2048), 200);
});
$this->publisher->publish($this->postPlatform);
expect($jobCalls)->toBeGreaterThanOrEqual(2);
Http::assertSent(function ($request) {
if (! str_contains($request->url(), 'createRecord')) {
return false;
}
$embed = $request['record']['embed'] ?? null;
return $embed
&& $embed['$type'] === 'app.bsky.embed.video'
&& data_get($embed, 'video.ref.$link') === 'bafretry456';
});
});