Community technical support mailing list was retired 2010 and replaced with a professional technical support team. For assistance please contact: Pre-sales Technical support via email to sales@march-hare.com.
On Tuesday 12 July 2005 06:51, Gabriel Genellina wrote: > At Monday 11/7/2005 13:46, David Somers wrote: > >I have now released checkcvswrappers (revision 1.5) > > > >It is for the GNU/Linux platform only, but if enough people ask me nicely > > I will port to Windows (its fairly trivial to do so). > > I tried to port it to Windows but failed to get the appropiate libraries to > compile successfully. You had a problem with argv? It should work on Windows... although I think you need to manaully create a config.h for it to be happy. > So I decided to port your program to Python (which appears to me, being > much better suited than "raw" C code for this task...) I'm happier using C than Python. > and then, honoring > Python's "batteries included" philosophy, I had no need for any external > library (and that makes me happy :) ) Um, don't you mean "batteries and the kitchen sink included" ;-) > In the process I noticed the following: > - in CrunchFile, you forgot to initialize the `offset` variable Possibly... automatic variables should really be initialized to 0 by the compiler, but its a bit hard to determine whether gcc does this or not... erring on the side of caution, I've now explicitly initilized this variable. [OT] Normally I'd run code against (Gimpel's )lint to pick up this type of thing, but its only on my Windows box at the moment (and I don't fancy spending $1K for the unix version). Yes, I know on Linux there is also Split (http://www.splint.org) to do static analysis of C source, but its development seems to have stalled... plus it uses a different way of marking up code than that required by Gimpel's lint, so you end up having to do a lot of duplicate work... (one solution would be to use a different markup and then use a processor to convert that into forms recognized by both systems). [/OT] > - the tests in CrunchFile must be reordered, utf32le_bom is never fired > because it's masked by utf16le_bom which comes first. You see, that's what happens when you write code and the caffeione level is too low. Thanks for spotting it. Greetings from Luxembourg, -- David Somers VoIP: FWD 622885 PGP Key = 7E613D4E Fingerprint = 53A0 D84B 7F90 F227 2EAB 4FD7 6278 E2A8 7E61 3D4E