samba - May 2003 - Modified Cups Backed pdf writer to keep the original document name.

Hi All,
I've got a samba/cups pdf writer going thanks to PeteFwee and Wolfram Q and
a few others...  the cups pdf backed i used has a default of
username-date.pdf which I didn't like and wanted the name of the job that
got send in the printer q on the winslows client.

The only way I could see in my very limited experience was to grab the
%%Title out of the postscript file and make sure that it didn't have any bad
characters.

I also did a few extra lines to send the .pdf to the person that sent it
(thankfully that was already passed though).

Here it is.  I didn't use awk when i probably should of because i'm crap
at
regular expressions.  Real programmers might have a laugh and my error
checking is non-existant, but it worked for me and may be useful to some
newbie out there somewhere....

thanks goes to Michael Goffioul for writing the orginal script which i got
from
http://printing.kde.org/downloads/pdfdistiller and the how to posted
originally to samba by Tom.    i'd be interested in any feedback about how i
should have used awk so i can learn.  :)


Regards
Matthew

#!/bin/sh
#
# This script is intended to be used as a CUPS backend, to create
# PDF file on-the-fly. Just create a printer using the device uri
# pdf:/path/to/dir/. When printing to this printer, a PDF file
# will be generated in the directory specified. The file name will
# be either "<jobname>.pdf" or "unknown.pdf",
depending whether the
# jobname is empty or not.
#
# To use it, simply copy this script to your backend directory, and
# create a printer with the correct URI. That's it.
#
# Copyright (C) Michael Goffioul (goffioul@imec.be) 2001
LOGFILE=/var/log/cups/pdf.log
PDFBIN=`which ps2pdf`
FILENAME# this is borrowed from printpdf script for the filename
PRINTTIME=`date +%b%d-%H%M%S`
DOMAIN="@yourdomain.com.au"
WEBSERVER="http://yourwebserver/pdf/"

#mail stuff
DATE=`date +%d%b%Y-%H:%M:%S`
MAILFILE=/tmp/mailtemp
HEADER=/tmp/header


# case of no argument, prints available URIs
if [ $# -eq 0 ]; then
	if [ ! -x "$PDFBIN" ]; then
		exit 0
	fi
	echo "direct pdf \"Unknown\" \"PDF Writing\""
	exit 0
fi

# case of wrong number of arguments
if [ $# -ne 5 -a $# -ne 6 ]; then
	echo "Usage: pdf job-id user title copies options [file]"
	exit 1
fi

# get PDF directory from device URI, and check write status
PDFDIR=${DEVICE_URI#pdf:}
if [ ! -d "$PDFDIR" -o ! -w "$PDFDIR" ]; then
	echo "ERROR: directory $PDFDIR not writable"
	exit 1
fi

echo "PDF directory: $PDFDIR" >> $LOGFILE


# Start of Matthew's modified stuff.

echo " " >> $LOGFILE
echo "Executable: $PDFBIN" >> $LOGFILE
echo "Arguments: |$1|$2|$3|$4|$5|$6|" >> $LOGFILE
echo $# $PRINTTIME >> $LOGFILE
echo "PS File: $6" >> $LOGFILE
FULLTITLE=`grep Title $6`
TITLE=${FULLTITLE:9}

#replace all not allowed characters with a _  - could do it with a regular
expression but too hard_
#      \ / : * ? " < > |

TITLE=${TITLE//:/_}
TITLE=${TITLE//\//_}
TITLE=${TITLE//</_}
TITLE=${TITLE//>/_}
TITLE=${TITLE//|/_}


echo "fulltitle: "$FULLTITLE >> $LOGFILE
echo "chopped: "$TITLE >> $LOGFILE



# generate output filename
OUTPUTFILENAMEif [ "$3" = "" ]; then
	OUTPUTFILENAME="$PDFDIR/unknown.pdf"
else
	# OUTPUTFILENAME="$PDFDIR/${3//[^[:alnum:]]/_}.pdf"
	# I changed this to user name, and the printtime to track down who
	# printed the PDF and when, samba printing just uses nobody

	#	default is username and date.  Matt modified to read the postscript
title.
	# OUTPUTFILENAME="$PDFDIR/$2-$PRINTTIME.pdf"
	OUTPUTFILENAME="$PDFDIR/$TITLE.pdf"

	echo "PDF file: $OUTPUTFILENAME placed in: $PDFDIR" >> $LOGFILE
fi

echo "Output file name: $OUTPUTFILENAME" >> $LOGFILE

# run ghostscript
if [ $# -eq 6 ]; then
	$PDFBIN  -sPAPERSIZE=a4 $6 "$OUTPUTFILENAME"
#	$PDFBIN  $6 "$OUTPUTFILENAME"
#>& /dev/null
else
	$PDFBIN -sPAPERSIZE=a4 - "$OUTPUTFILENAME" >& /dev/null
fi

# modify ownership and permissions on the file
#  - world readable
#  - owns to user specified in argument
chmod a+r "$OUTPUTFILENAME"
if [ "$2" != "" ]; then
	chown $2 "$OUTPUTFILENAME"
fi


#Sendmail to the owner.
echo "Sending mail: "$2$DOMAIN >> $LOGFILE

TO=$2$DOMAIN
SUBJECT='PDF Printer Job'
BODY='PDF Created :'$DATE
#echo PDF Created: $DATE >> BODY


echo From: PDFPrinter$DOMAIN >> $HEADER
echo To: $TO >> $HEADER
#echo Cc: $CC >> $HEADER
#echo Bcc: $BCC >> $HEADER
echo Subject: $SUBJECT >> $HEADER
echo "" >> $HEADER
echo $BODY >> $HEADER
echo "" >> $HEADER
echo "" >> $HEADER
echo "If your PDF job is not attached to this email please forward this
message \
to the help desk for assistance." >> $HEADER
echo "------------------" >> $HEADER
echo Filename: $OUTPUTFILENAME >> $HEADER


cat $HEADER > $MAILFILE
#uuencode "$OUTPUTFILENAME" $2-$PRINTTIME.pdf >> $MAILFILE

uuencode "$OUTPUTFILENAME" "$TITLE.pdf" >> $MAILFILE
cat $MAILFILE | /usr/lib/sendmail -t

rm $MAILFILE
rm $HEADER


exit 0

Matthew Western mwestern at affairs.net.au wrote on Samba-Digest
> Fri May 16 14:33:29 GMT 2003
> 
> 
> Hi All,
> I've got a samba/cups pdf writer going thanks to PeteFwee and Wolfram Q
and
> a few others...  the cups pdf backed i used has a default of
> username-date.pdf which I didn't like and wanted the name of the job
that
> got send in the printer q on the winslows client.
> 
> The only way I could see in my very limited experience was to grab the
> %%Title out of the postscript file and make sure that it didn't have
any bad
> characters.
Hi Matthew,

the job title is passed to all filters and backends as $3 by CUPS. It is
my experience that it is more reliable to use this, because the %%Title
is more often not present than $3. The complete list of parameters provided
by CUPS to all filters and backends is this:

   $0  --   printername
   $1  --   job-ID
   $2  --   username
   $3  --   job-title
   $4  --   printjob-options
   $5  --   requested copies
   $6  --   <filename>; empty if taken from 'stdin'

They may be picked up by self-rolled custom backends and filters to
use for their purposes.x
> I also did a few extra lines to send the .pdf to the person that sent it
> (thankfully that was already passed though).
> 
> Here it is.  I didn't use awk when i probably should of because i'm
crap at
> regular expressions.  Real programmers might have a laugh and my error
> checking is non-existant, but it worked for me and may be useful to some
> newbie out there somewhere....
> 
> thanks goes to Michael Goffioul for writing the orginal script which i got
> from
> http://printing.kde.org/downloads/pdfdistiller and the how to posted
> originally to samba by Tom.    i'd be interested in any feedback about
how i
> should have used awk so i can learn.  :)
Sorry for giving feedback on a not-asked question...  ;-)
> Regards
> Matthew
> 
Cheers,
Kurt