Updating field in XFA PDF

[jheasleycrea]

I am attempting to update some fields in an XFA enabled pdf. So far the closest I've come to this is finding /XFA and accessing the dataset portion via xfa[7].get_object().get_data()

This returns the XML that corresponds to all the field values. From there I update the xml with my desired values and then attempt to set it via xfa[7].get_object().set_data(write_xml).

This throws an Exception: pypdf.errors.PdfReadError: Streams encoded with different filter from only FlateDecode is not supported.

I am extremely new to PDF manipulation and wouldn't be surprised if I'm going entirely in the wrong direction, but does anyone have any advice on how to achieve this?

Edit:

import pypdf
def findInDict(needle, haystack):
    for key in haystack.keys():
        try:
            value=haystack[key]
        except:
            continue
        if key==needle:
            return value
        if isinstance(value,dict):            
            x=findInDict(needle,value)            
            if x is not None:
                return x
pdfobject=open('Test.pdf','rb')
pdf=pypdf.PdfReader(pdfobject)
xfa=findInDict('/XFA',pdf.resolved_objects)

xml=xfa[7].get_object().get_data()
write_xml = xml.replace(b'101101101</ce:PartyIdentificationNumberText\n>', b'202202202</ce:PartyIdentificationNumberText\n>')
xfa[7].get_object().set_data(write_xml)


writer = pypdf.PdfWriter(clone_from=pdf)
outputstream = open('write_file.pdf', 'wb')
writer.write(outputstream)

outputstream.close()
pdf.close()

[stefan6419846]

Please provide some complete code example and an example PDF for further analysis.

The error message indicates that the object you are trying to replace does not use the FlateDecode filter. pypdf currently does not know how to encode data with other filters, thus your approach does not work here at the moment.

[jheasleycrea]

write_file.pdf
Here's the file.
The code so far is edited above

[pubpub-zz]

@jheasleycrea
can you send also the original file.thx

[jheasleycrea]

Here's the original
Test.pdf

[jheasleycrea]

To be clear. I originally wiped write_file.pdf before attaching it here. Currently if I don't try and update the data, the PDFWriter successfully spits out a copy of Test.pdf that is identical as far as I can tell

[jheasleycrea]

Figured out the problem. The issue from the fact that the /Filter here was an ArrayObject of ["/FlateDecode"]. in EncodedStreamObject.set_data the line to check Decode status if self.get(SA.FILTER, "") == FT.FLATE_DECODE: returns false.
I was able to work around this by setting /Filter to _base.NameObject("/FlateDecode")
From there the rest of my code ran and the field was updated as expected.

[ljbergmann]

Do you have an example for this? I just looked at the documentation and i can't find anything relating to this topic and i'm a bit lost here.

[pubpub-zz]

have a look here:
https://pypdf.readthedocs.io/en/stable/modules/generic.html

by the way, I've produced PR #2821 that will make set_data compatible with your case

[ljbergmann]

Thank you very much and indeed it is working ;) combined with #2811 I'm able to fill out the pdf. Now only the error message problem is left. I'm trying to identify why acrobat is throwing the error message and maybe i can contribute to the PR / project in finding a solution for this.

[pubpub-zz]

Which error message ?

[ljbergmann]

the one i mentioned in #2811