[SOLVED] Printing from Flash does not work

[djennings]

Today one of my users complained she could not print a document from a flash application. The job fails in Cups.
The page she was printing from is http://www.senteacher.org/Worksheet/8/Scanning.xhtml
Press the 'Print' button and you will see what I mean.


After some research I discovered that it is a flaw in the Adobe Flash plugin creating a non compliant PostScript document
https://bugzilla.redhat.com/show_bug.cgi?id=773765

I tried gnash instead, but it would not print at all.

I think if I could intercept the print command from the browser I could use sed to edit the file to make it compliant before it went into the cups print queue.
So far I have not been able to work out how to intercept the file. None of the commands in Firefox about:config seem to make any difference to how the job is printed. Any suggestions?

[doktor5000]

Can't you just send the document to a virtual PDF printer instead of a real one?

[djennings]

Can't you just send the document to a virtual PDF printer instead of a real one?

No. Normally when I print from Firefox I get to choose the printer, or to print to file (pdf or postscript), but when printing from flash I am only offered the physical printers.

[Ken-Bergen]

Today one of my users complained she could not print a document from a flash application. The job fails in Cups.
The page she was printing from is http://www.senteacher.org/Worksheet/8/Scanning.xhtml
Press the 'Print' button and you will see what I mean.

What version of Mageia?
What version of Flash?
What printer and how is it connected?

FWIW: I just tried in Magiea 1 and Cauldron, and both printed several different pages perfectly.

[djennings]

What version of Mageia?
What version of Flash?
What printer and how is it connected?

Mageia 2
flash-player-plugin-11.2.202.251-1.mga2.nonfree.x86_64.rpm
HP Officejet-Pro-L7400 using hpcups

If I run cupstestdsc on the job in the print queue I get

Code: Select all

# cupstestdsc d00093-001
d00093-001: FAIL
    Missing or bad %%Pages: comment.
        REF: Page 43, %%Pages:
    Missing or bad %%Page: comments.
        REF: Page 53, %%Page:
    Warning: no %%EndComments comment in file.

If I then edit the print document to add
%%Pages: 1
%%EndComments
%%Page: 1 1

after %%BoundingBox then it prints OK

[djennings]

Here is the same problem described way back in 2009

http://www.cups.org/str.php?L3281

I tried it out in mageia 3, and the problem is still there. If you are not seeing it Ken, then it may be because you are using a different chain of filters in cups.

Clearly it is a fault in the Adobe flash plugin, and clearly Adobe cannot be bothered to fix it. The good news is that the same post shows how to work around the problem by creating a custom Cups filter. I will have a go at making a custom filter today.

[djennings]

Yep. A custom cups filter fixes it.

For the record this is what to do
Create a file /etc/cups/mime.types

Code: Select all

#Custom CUPS mime types
#    broken-postcript mime type is to allow cups to identify postscript generated by Adobe Flash plugin with bad DSC
application/broken-postscript string(0,"%!PS-Adobe-3.0") + !contains(0,4096,"%%Pages:")

Create a file /etc/cups/mime.convs

Code: Select all

# convert broken postscript to good postscript using the filter psfix
application/broken-postscript  application/postscript 10     psfix


Create a file /usr/lib/cups/filter/psfix

Code: Select all

#!/bin/sh
# This is a custom CUPS filter to repair broken postscript documents
# created by Adobe Flash plugin
echo '%!'; exec tail -n +2  "${6-"-"}"


Do not forget to make this file executable

Code: Select all

chmod +x /usr/lib/cups/filter/psfix

Now it is possible to print from within Adobe Flash such as this site
http://www.senteacher.org/Worksheet/8/Scanning.xhtml


If you right click on a completed job in the print job viewer and select 'Attributes' you will see the 'Document type' is shown as application/broken-postscript.

I will raise a bug on it. Since Adobe clearly has no intention of fixing this bug maybe mageia could include this workaround.

[doktor5000]

FWIW, here's the link to that bug for reference: https://bugs.mageia.org/show_bug.cgi?id=8017