Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Allow to asynchronously generate PDFs #65

Merged
merged 8 commits into from
May 7, 2024
Merged

Allow to asynchronously generate PDFs #65

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
8 commits
Select commit Hold shift + click to select a range
cf6b395
Pdfexport: Use `htmlToPdf()` in `streamPdfFromHtml()`
yhabteab Mar 13, 2024
d7b3897
Pdfexport: Extract merge PDFs into a separate method
yhabteab Mar 13, 2024
4638e76
HeadlessChrome: Introduce `asyncToPdf()` method
yhabteab Mar 12, 2024
bdd2b9e
Pdfexport: Introduce `asyncHtmlToPdf()` method
yhabteab Mar 12, 2024
46d7cc9
HeadlessChrome: Deduplicate `asyncToPdf()` and `toPdf()` methods
yhabteab Mar 13, 2024
4e201da
Fix phpstan errors
yhabteab Mar 13, 2024
bc4f060
Pdfexport: Generate PDF before modifying response headers
yhabteab Mar 15, 2024
8e7ff15
Fix doc block
yhabteab Mar 20, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
228 changes: 144 additions & 84 deletions library/Pdfexport/HeadlessChrome.php
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -12,8 +12,11 @@
use ipl\Html\HtmlString;
use LogicException;
use React\ChildProcess\Process;
use React\EventLoop\Factory;
use React\EventLoop\Loop;
use React\EventLoop\TimerInterface;
use React\Promise;
use React\Promise\ExtendedPromiseInterface;
use Throwable;
use WebSocket\Client;
use WebSocket\ConnectionException;

Expand Down Expand Up @@ -240,111 +243,168 @@ public function fromHtml($html, $asFile = false)
}

/**
* Export to PDF
* Generate a PDF raw string asynchronously.
*
* @return string
* @throws Exception
* @return ExtendedPromiseInterface
*/
public function toPdf()
public function asyncToPdf(): ExtendedPromiseInterface
{
switch (true) {
case $this->remote !== null:
try {
$result = $this->jsonVersion($this->remote[0], $this->remote[1]);
$parts = explode('/', $result['webSocketDebuggerUrl']);
$pdf = $this->printToPDF(
join(':', $this->remote),
end($parts),
! $this->document->isEmpty() ? $this->document->getPrintParameters() : []
);
break;
} catch (Exception $e) {
if ($this->binary === null) {
throw $e;
} else {
$deferred = new Promise\Deferred();
Loop::futureTick(function () use ($deferred) {
switch (true) {
case $this->remote !== null:
try {
$result = $this->jsonVersion($this->remote[0], $this->remote[1]);
if (is_array($result)) {
yhabteab marked this conversation as resolved.
Show resolved Hide resolved
$parts = explode('/', $result['webSocketDebuggerUrl']);
$pdf = $this->printToPDF(
join(':', $this->remote),
end($parts),
! $this->document->isEmpty() ? $this->document->getPrintParameters() : []
);
break;
}
} catch (Exception $e) {
if ($this->binary == null) {
$deferred->reject($e);
return;
}

Logger::warning(
'Failed to connect to remote chrome: %s:%d (%s)',
$this->remote[0],
$this->remote[1],
$e
);
}
}

// Fallback to the local binary if a remote chrome is unavailable
case $this->binary !== null:
$browserHome = $this->getFileStorage()->resolvePath('HOME');
$commandLine = join(' ', [
escapeshellarg($this->getBinary()),
static::renderArgumentList([
'--bwsi',
'--headless',
'--disable-gpu',
'--no-sandbox',
'--no-first-run',
'--disable-dev-shm-usage',
'--remote-debugging-port=0',
'--homedir=' => $browserHome,
'--user-data-dir=' => $browserHome
])
]);

if (Platform::isLinux()) {
Logger::debug('Starting browser process: HOME=%s exec %s', $browserHome, $commandLine);
$chrome = new Process('exec ' . $commandLine, null, ['HOME' => $browserHome]);
} else {
Logger::debug('Starting browser process: %s', $commandLine);
$chrome = new Process($commandLine);
}
// Reject the promise if we didn't get the expected output from the /json/version endpoint.
if ($this->binary === null) {
$deferred->reject(
new Exception('Failed to determine remote chrome version via the /json/version endpoint.')
);
return;
}

$loop = Factory::create();
// Fallback to the local binary if a remote chrome is unavailable
case $this->binary !== null:
$browserHome = $this->getFileStorage()->resolvePath('HOME');
$commandLine = join(' ', [
escapeshellarg($this->getBinary()),
static::renderArgumentList([
'--bwsi',
'--headless',
'--disable-gpu',
'--no-sandbox',
'--no-first-run',
'--disable-dev-shm-usage',
'--remote-debugging-port=0',
'--homedir=' => $browserHome,
'--user-data-dir=' => $browserHome
])
]);

if (Platform::isLinux()) {
Logger::debug('Starting browser process: HOME=%s exec %s', $browserHome, $commandLine);
$chrome = new Process('exec ' . $commandLine, null, ['HOME' => $browserHome]);
} else {
Logger::debug('Starting browser process: %s', $commandLine);
$chrome = new Process($commandLine);
}

$killer = $loop->addTimer(10, function (TimerInterface $timer) use ($chrome) {
$chrome->terminate(6); // SIGABRT
Logger::error(
'Terminated browser process after %d seconds elapsed without the expected output',
$timer->getInterval()
);
});
$killer = Loop::addTimer(10, function (TimerInterface $timer) use ($chrome, $deferred) {
$chrome->terminate(6); // SIGABRT

$chrome->start($loop);
Logger::error(
'Browser timed out after %d seconds without the expected output',
$timer->getInterval()
);

$pdf = null;
$chrome->stderr->on('data', function ($chunk) use (&$pdf, $chrome, $loop, $killer) {
Logger::debug('Caught browser output: %s', $chunk);
$deferred->reject(
new Exception(
'Received empty response or none at all from browser.'
. ' Please check the logs for further details.'
)
);
});

$chrome->start();

$chrome->stderr->on('data', function ($chunk) use ($chrome, $deferred, $killer) {
Logger::debug('Caught browser output: %s', $chunk);

if (preg_match(self::DEBUG_ADDR_PATTERN, trim($chunk), $matches)) {
Loop::cancelTimer($killer);

try {
$pdf = $this->printToPDF(
$matches[1],
$matches[2],
! $this->document->isEmpty() ? $this->document->getPrintParameters() : []
);
} catch (Exception $e) {
Logger::error('Failed to print PDF. An error occurred: %s', $e);
}

$chrome->terminate();

if (! empty($pdf)) {
$deferred->resolve($pdf);
} else {
$deferred->reject(
new Exception(
'Received empty response or none at all from browser.'
. ' Please check the logs for further details.'
)
);
}
}
});

if (preg_match(self::DEBUG_ADDR_PATTERN, trim($chunk), $matches)) {
$loop->cancelTimer($killer);
$chrome->on('exit', function ($exitCode, $signal) use ($killer) {
Loop::cancelTimer($killer);

try {
$pdf = $this->printToPDF(
$matches[1],
$matches[2],
! $this->document->isEmpty() ? $this->document->getPrintParameters() : []
);
} catch (Exception $e) {
Logger::error('Failed to print PDF. An error occurred: %s', $e);
}
Logger::debug('Browser terminated by signal %d and exited with code %d', $signal, $exitCode);

$chrome->terminate();
}
});
// Browser is either timed out (after 10s) and the promise should have already been rejected,
// or it is terminated using its terminate() method, in which case the promise is also already
// resolved/rejected. So, we don't need to resolve/reject the promise here.
});
yhabteab marked this conversation as resolved.
Show resolved Hide resolved

$chrome->on('exit', function ($exitCode, $termSignal) use ($loop, $killer) {
$loop->cancelTimer($killer);
return;
}

Logger::debug('Browser terminated by signal %d and exited with code %d', $termSignal, $exitCode);
});
if (! empty($pdf)) {
$deferred->resolve($pdf);
} else {
$deferred->reject(
new Exception(
'Received empty response or none at all from browser.'
. ' Please check the logs for further details.'
)
);
}
});

$loop->run();
}
return $deferred->promise();
}

if (empty($pdf)) {
throw new Exception(
'Received empty response or none at all from browser.'
. ' Please check the logs for further details.'
);
}
/**
* Export to PDF
*
* @return string
* @throws Exception
*/
public function toPdf()
{
$pdf = '';
// We don't intend to register any then/otherwise handlers, so call done on that promise
// to properly propagate unhandled exceptions to the caller.
$this->asyncToPdf()->done(function (string $newPdf) use (&$pdf) {
$pdf = $newPdf;
});

Loop::run();

return $pdf;
}
Expand Down
57 changes: 39 additions & 18 deletions library/Pdfexport/ProvidedHook/Pdfexport.php
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -14,6 +14,7 @@
use Icinga\Module\Pdfexport\PrintableHtmlDocument;
use iio\libmergepdf\Driver\TcpdiDriver;
use iio\libmergepdf\Merger;
use React\Promise\ExtendedPromiseInterface;

class Pdfexport extends PdfexportHook
{
Expand Down Expand Up @@ -86,43 +87,53 @@ public function htmlToPdf($html)
)
->toPdf();

$merger = new Merger(new TcpdiDriver());
$merger->addRaw($coverPagePdf);
$merger->addRaw($pdf);

$pdf = $merger->merge();
$pdf = $this->mergePdfs($coverPagePdf, $pdf);
}

return $pdf;
}

public function streamPdfFromHtml($html, $filename)
/**
* Transforms the given printable html document/string asynchronously to PDF.
*
* @param PrintableHtmlDocument|string $html
*
* @return ExtendedPromiseInterface
*/
public function asyncHtmlToPdf($html): ExtendedPromiseInterface
{
$filename = basename($filename, '.pdf') . '.pdf';

// Keep reference to the chrome object because it is using temp files which are automatically removed when
// the object is destructed
$chrome = $this->chrome();

$pdf = $chrome->fromHtml($html, static::getForceTempStorage())->toPdf();
$pdfPromise = $chrome->fromHtml($html, static::getForceTempStorage())->asyncToPdf();

if ($html instanceof PrintableHtmlDocument && ($coverPage = $html->getCoverPage()) !== null) {
$coverPagePdf = $chrome
->fromHtml(
/** @var ExtendedPromiseInterface $pdfPromise */
$pdfPromise = $pdfPromise->then(function (string $pdf) use ($chrome, $html, $coverPage) {
return $chrome->fromHtml(
(new PrintableHtmlDocument())
->add($coverPage)
->addAttributes($html->getAttributes())
->removeMargins(),
static::getForceTempStorage()
)
->toPdf();
)->asyncToPdf()->then(
function (string $coverPagePdf) use ($pdf) {
return $this->mergePdfs($coverPagePdf, $pdf);
}
);
});
}

$merger = new Merger(new TcpdiDriver());
$merger->addRaw($coverPagePdf);
$merger->addRaw($pdf);
return $pdfPromise;
}

$pdf = $merger->merge();
}
public function streamPdfFromHtml($html, $filename)
{
$filename = basename($filename, '.pdf') . '.pdf';

// Generate the PDF before changing the response headers to properly handle and display errors in the UI.
$pdf = $this->htmlToPdf($html);

/** @var Web $app */
$app = Icinga::app();
Expand Down Expand Up @@ -151,4 +162,14 @@ protected function chrome()

return $chrome;
}

protected function mergePdfs(string ...$pdfs): string
{
$merger = new Merger(new TcpdiDriver());
foreach ($pdfs as $pdf) {
$merger->addRaw($pdf);
}

return $merger->merge();
}
}
4 changes: 2 additions & 2 deletions phpstan-baseline.neon
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -66,12 +66,12 @@ parameters:
path: library/Pdfexport/HeadlessChrome.php

-
message: "#^Cannot access offset 'webSocketDebuggerUrl' on array\\|bool\\.$#"
message: "#^Cannot call method on\\(\\) on React\\\\Stream\\\\ReadableStreamInterface\\|React\\\\Stream\\\\WritableStreamInterface\\|null\\.$#"
count: 1
path: library/Pdfexport/HeadlessChrome.php

-
message: "#^Cannot call method on\\(\\) on React\\\\Stream\\\\ReadableStreamInterface\\|React\\\\Stream\\\\WritableStreamInterface\\|null\\.$#"
message: "#^Method Icinga\\\\Module\\\\Pdfexport\\\\HeadlessChrome\\:\\:asyncToPdf\\(\\) should return React\\\\Promise\\\\ExtendedPromiseInterface but returns React\\\\Promise\\\\PromiseInterface\\.$#"
count: 1
path: library/Pdfexport/HeadlessChrome.php

Expand Down