Stops with no error indication when low memory limit

[mazhuravlev]

Stops on Parser::parseFile() when file is big and PHP memory_limit is low. No error is indicated.

[aik099]

The actual parsing is done by TCPDF and large file isn't what eats memory, but how file is structured internally. In general it would be a good idea to increase memory limit in any case before doing PDF parsing.

[srakas]

I've got the same error while parsing PDF file of size 115KB, while memory limit is set to 1GB.

After debugging I've located the problem within TCPDF when method getRawObject() is being called. In my case this is happening if there are blank pages in PDF (nothing on the page except maybe footer), so the code enters infinite do-while loop and causes memory leakage.

[keesschepers]

We have a pretty similar problem, even when setting memory limits up to 8GB. We are building a webpage crawler which should also extract PDF documents with your library but especially with big PDF's (like 40mb) it's crashing on Smalot/PdfParser/Object.php on line 143 which is the preg_match_all. I assume that too many occurrences are found and the array consumes too much memory.

Is it possible to split up the content of the document into chunks and then ready every X bytes and cleanup stuff or is there a risk of data corruption in that case?

[lkalinka]

Here is example of PDF that is causing such problems:
http://www.cranendonck.nl/document.php?m=64&fileid=42797&f=6e910448e0030fb42fbfcc848a7e6258&attachment=0&c=33386
its very small PDF but it can eat whole available memory (8GB) when i trying to read(parse) this file, that make server crash.

I speaked with author of TCPDF Nicola Asuni and he fixed that issue as is stated there:(issue was a XREF position that was not correct)
https://sourceforge.net/p/tcpdf/discussion/435311/thread/996c7864/

So its needed to apply the same fix over there.
Its needed to change source of TCPDF_PARSER to this project of the same author:
https://github.com/tecnickcom/tc-lib-pdf-parser

Exactly this commit was a fix for this issue:
tecnickcom/tc-lib-pdf-parser@6e35159

[JonathanZWeber]

This issue has popped up both times I've used this project in the last month - a fix is really essential.

[k00ni]

I know this is super old, but could any of you give #787 a try? Thanks in advance 🙏

[StanProg]

I know this is super old, but could any of you give #787 a try? Thanks in advance 🙏

I tried it, but there was no improvement.

File size: 1454805 bytes (1.4 MB)
Bytes taken from parseContent()/parseFile(): 288030720 (288 MB) + Out of memory

[k00ni]

@StanProg Thank you very much for trying it out.