views:

1964

answers:

3

I am trying to add an existing PDF (created otherwise) to a PDF created with FPDF using FPDI. It seems to work find for most PDFs, but I get the following error:

FPDF error: Unable to find xref table.

After some testing, I figured out that if I use a PDF version of 1.4 or lower (Acrobat v5 or lower) it seems to work. Looking at the file it seems to be because of a different PDF format.

Is there a work around or solution to this? I have probably 10 000+ PDFs uploaded by users, some of the new working, some of them not. It's also annoying that FPDI just dies instead of causing some kind of error or even returning false

The only solution I can think of right now is to parse the first line of the PDF file to determine the version and only add the ones that over versions that will work.

Edit:

I also tried this in Zend_Pdf and it had the following result, using the demo file that comes with the Zend Framework:

Fatal error: Uncaught exception 'Zend_Pdf_Exception' with message 'Cross-reference streams are not supported yet.' in /Volumes/Documents/temp/ZendFramework-1.7.7/library/Zend/Pdf/Parser.php:331
Stack trace:
#0 /Volumes/Documents/temp/ZendFramework-1.7.7/library/Zend/Pdf/Parser.php(455): Zend_Pdf_Parser->_loadXRefTable('116')
#1 /Volumes/Documents/temp/ZendFramework-1.7.7/library/Zend/Pdf.php(297): Zend_Pdf_Parser->__construct('/Volumes/Docume...', Object(Zend_Pdf_ElementFactory_Proxy), true)
#2 /Volumes/Documents/temp/ZendFramework-1.7.7/library/Zend/Pdf.php(250): Zend_Pdf->__construct('/Volumes/Docume...', NULL, true)
#3 /Volumes/Documents/temp/ZendFramework-1.7.7/demos/Zend/Pdf/demo.php(37): Zend_Pdf::load('/Volumes/Docume...')
#4 {main}
  thrown in /Volumes/Documents/temp/ZendFramework-1.7.7/library/Zend/Pdf/Parser.php on line 331

Is it impossible to parse PDFs greater than 1.4?

+3  A: 

You can use FPDI with TCPDF, it will generate and read (with fpdi) version 1.7 pdf.

http://sourceforge.net/projects/tcpdf/

EDIT----------------------------------------------------

I have recently been challenged with the issue of using various versions and formats of PDF with fpdi and tcpdf, and I can confirm that pdf versions above 1.4 are hit and miss, the interestng thing is that I have a lot of working 1.7 pdf, and about 50/50 on 1.5 and 1.6.

I worked around the issue by converting all non workable pdf files to v1.4 using GhostScript which you can exec from the command line. Use ps2pdf14 from the GhostScript lib folder.

+1  A: 

Thanks I used first pdf2ps and then ps2pdf14 as you suggested and it worked perfectly

Erick G
+1  A: 

Setasign is offering a PDF parser as a commercial addon to FPDI. We had no troubles parsing PDF files up to version 1.7 so far using the new parser.

eisbernd