Cannot process pdf with a longer header #868
Replies: 1 comment 4 replies
-
|
I tried recreating this the following way: $targetPath = __DIR__.'/temp/large_header.pdf';
$lorem = file_get_contents('http://loripsum.net/api/html?p=150');
Browsershot::html('<h1>Hello, World!</h1>')
->showBrowserHeaderAndFooter()
->headerHtml($lorem)
->save($targetPath);However, even with a header spanning 150 paragraphs of lorem ipsum text, this finishes in under two seconds, and creates the PDF just fine. Judging by the error you included, what's missing is the command-line argument including the JSON holding the data that gets passed to the browser instance. It looks like it's there, but for some reason, the keys and values (all strings) have two sets of double quotes (") around them. i.e When providing URLs, you should also not provide URLs to files, as indicated by this snippet from the if (str_starts_with(strtolower($url), 'file://')) {
throw FileUrlNotAllowed::make();
}I believe the JSON issue is preventing this exception from throwing, as the string actually begins with |
Beta Was this translation helpful? Give feedback.
-
|
@bluesheep100 Unfortunatly the lorem ipsum API only generates 3 short paragraphs, not 150. Try this: $targetPath = __DIR__.'/temp/large_header.pdf';
$randomString = Str::random(8000); // using Laravel helper, you can use any other method
Browsershot::html('<h1>Hello, World!</h1>')
->showBrowserHeaderAndFooter()
->headerHtml("<p>$randomString</p>")
->save($targetPath);does it still work for you? I was able to generate PDFs by this method with random strings of abount 7800 characters (always with 7805 chars, sometimes with 7806, never with 7807). I have no idea why the printed error has two sets of double quotes, but that's not the point, with smaller headers it works, with longer it doesn't. |
Beta Was this translation helpful? Give feedback.
-
|
For the avoidance of doubt, in any case I am not manually configuring Browsershot to work, but I am using it via the spatie/laravel-pdf package (see this issue). However, the error is directly attributable to Browsershot as demonstrated by the fact that it also occurs with the code written above. |
Beta Was this translation helpful? Give feedback.
-
|
Still works fine for me. I did What version of Puppeteer and Chrome (presumably) are you doing this one? You can try explicitly passing the path to the binaries for Browsershot to use, maybe it's not using the right Chrome version or something. |
Beta Was this translation helpful? Give feedback.
-
|
Very interesting, I tried using Anyway the issue doesn't seem related to the Chrome version: I tried to set a wrong binary path and a small header and I got a "binary not found" error, but when I tried with a 8000 characters long header I get the same error reported above, so I think it is thrown before evaluating the Chrome binaries. Now I will try to generate my complete pdf using WSL, I'll let you know if everything works correctly! |
Beta Was this translation helpful? Give feedback.
-
When the headerTemplate passed as option to Browsershot is too long (this happens for example when the base64 encoding of an image is on average long or by adding multiple images with a short encoding that individually might work) the following error occurs (omitted the full content of headerTemplate to make the error readable):
This prevents me from creating elaborate or multi-image headers or footers, can anyone help me on this?
Beta Was this translation helpful? Give feedback.
All reactions