Hello Jorrit,
Am Samstag, den 16.08.2014, 00:32 +0200 schrieb Jö Fahlke: > Am Fri, 15. Aug 2014, 15:17:14 +0200 schrieb Jörg Frings-Fürst: > > Am Freitag, den 15.08.2014, 11:37 +0200 schrieb Jö Fahlke: > > > Am Fri, 15. Aug 2014, 01:15:09 +0200 schrieb Jörg Frings-Fürst: > > > > many thanks for your well-written bug report. > > > > [...] > > I had a look at the patch, and also at the PPD specification[1]. The way I > read this, the line > > *DefaultColorModel: Grayscale > > consists of the main keyword "*DefaultColorModel" and the StringValue > "Grayscale ". The specification says about StringValues (page 23): > > • The value is composed of printable 7-bit ASCII byte codes, possibly > separated by spaces and tabs into multiple components. It is terminated > by a newline, or by a slash, in the case of a translation string. > And then: • There is no escape mechanism for forbidden characters. > The part "possibly separated by spaces and tabs into multiple components" > allows some room for interpretation. But in any case this looks like internal > structure of the value, so any non-leading whitespace appears to be part of > the value to me. Sorry for leading you on the wrong path with my idea about > stripping whitespace. > > Also, you unconditionally strip trailing whitespace from all lines. But PPD > knows some types of values that can span multiple lines, and you probably > should not strip trailing whitespace in the middle of such values. That’s right. At all ppd's here I have search for a space or tab at the end. I don't found any. But it is not forbidden. So I moved to the end of the Value reading. So I manipulate every entry only once a time. > > And in any case, this is not sufficient. You could for instance use the > spelling "Greyscale" by mistake, and you'd still get the segfault, since > choice is still dereferenced without checking whether it is NULL. (I checked, > with your patch.) > I think that this work can only done with the upstream maintainer. > By the way, is there a reason you did not reply to the bug report? Feel free > to post this mail there, otherwise. > Sorry for this. I was only remote with a tablet online. I think that I have pressed the wrong menu item. > Regards, > Jö. > > [1] http://partners.adobe.com/public/developer/en/ps/5003.PPD_Spec_v4.3.pdf > CU Jörg -- pgp Fingerprint: 7D13 3C60 0A10 DBE1 51F8 EBCB 422B 44B0 BE58 1B6E pgp Key: BE581B6E CAcert Key S/N: 0E:D4:56 Jörg Frings-Fürst D-54526 Niederkail Threema: SYR8SJXB IRC: [email protected] [email protected]
Description: remove trailing whitespaces from ppd files Remove trailing whitespaces from ppd file to prevent segfault Author: Jörg Frings-Fürst <[email protected]> Origin: <upstream|backport|vendor|other>, <URL, required except if Author is present> Bug: <URL to the upstream bug report if any, implies patch has been forwarded, optional> Bug-Debian: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=758133 Forwarded: <URL|no|not-needed, useless if you have a Bug field, optional> Applied-Upstream: <version|URL|commit, identifies patches merged upstream, optional> Reviewed-by: <name and email of a reviewer, optional> Last-Update: 2014-08-15 --- This patch header follows DEP-3: http://dep.debian.net/deps/dep3/ Index: trunk/options.c =================================================================== --- trunk.orig/options.c 2014-08-17 09:56:36.990949026 +0200 +++ trunk/options.c 2014-08-17 09:59:43.750779956 +0200 @@ -1628,6 +1628,10 @@ /* remove last newline */ dstrremovenewline(value); + /* remove last whitespace */ + /* new 2014-08-17 */ + dstrtrim_right(value); + /* process key/value pairs */ if (strcmp(key, "NickName") == 0) { unhtmlify(printer_model, 256, value->data);
signature.asc
Description: This is a digitally signed message part
