From: craigberry@nospam.mac.com   
      
   On 7/3/25 7:24 PM, Arne Vajhøj wrote:   
   > On 7/3/2025 6:31 PM, Craig A. Berry wrote:   
   >> https://github.com/FreeTDS/freetds/blob/Branch-1_5/src/replacements/   
   >> vasprintf.c   
   >>   
   >> does the following:   
   >>   
   >> #if HAVE_PATHS_H   
   >> #include    
   >> #endif /* HAVE_PATHS_H */   
   >> . . .   
   >> #ifndef _PATH_DEVNULL   
   >> #define _PATH_DEVNULL "/dev/null"   
   >> #endif   
   >>   
   >> FreeTDS is relying on autoconf to sort out whether paths.h is available   
   >> and will only include it if HAVE_PATHS_H is defined in config.h, and   
   >> thus only defines _PATH_DEVNULL if it's known not to exist. Obviously   
   >> one should avoid making up one's own reserved identifiers and clobbering   
   >> documented and well-known names, but that's not really the same thing as   
   >> providing an implementation for one that's known to be missing. This   
   >> fallback approach could of course be added to Arne's implementation if   
   >> desired; to me it seems a lot cleaner than defining yet another macro,   
   >> which would need to be defined in terms of _PATH_DEVNULL when it exists   
   >> but otherwise not.   
   >>   
   >> The problem for VMS is that the workaround doesn't work.  Or at least it   
   >> didn't when I was first porting FreeTDS 20+ years ago because the CRTL   
   >> at the time did not recognize '/dev/null' as a valid path.  I believe   
   >> recentish CRTLs (maybe VMS 8.x and later?) do have special case code   
   >> that translates it to the native null device.  But that wasn't   
   >> available, so I added the following line to the template from which   
   >> config.h is generated on VMS:   
   >>   
   >> #define _PATH_DEVNULL "_NLA0:"   
   >>   
   >> The main advantage of this approach is that it worked, and it did so   
   >> with a one-line change to a file I was already maintaining. I didn't   
   >> have to sprinkle '#ifdef __VMS' in files that already existed and were   
   >> already working on other platforms. I didn't have to create a lot of new   
   >> files, the presence of which in the repository I couldn't test without   
   >> autoconf/automake, which I didn't have access to at the time.  All of   
   >> which meant upstream maintainers were far more likely to accept my   
   >> changes.   
   >   
   > So you are saying that for the code to work on *nix then I need   
   > to keep those, because having it defined in config.h is a non-*nix   
   > thing and for *nix it only set HAVE_PATHS_H which requires the   
   > include in this file?   
   >   
   > I will update.   
      
   I think preserving the HAVE_PATHS_H check is probably a good idea.  It   
   makes sure you get the correct, system-specific version of   
   _PATH_DEVNULL, if any.  And it preserves an existing approach to   
   portability that you aren't worried about so you can focus on the   
   vasprintf() guts.   
      
   I haven't done any testing of your code.  I do agree with Dan that   
   reducing the repeated code would be a benefit.  You don't need the first   
   #if because the file won't even be built if HAS_VASPRINTF is detected   
   during configuration. Which means for the remaining 4 cases you've got   
   this in all of them that could be put outside the #ifdefs, and without   
   any nesting:   
      
   va_list cp;   
   va_copy(cp, ap);   
      
   (len < 0) return len;   
      
   *ret = malloc(len + 1);   
   if(!*ret) return -1;   
   return vsprintf(*ret, fmt, cp);   
      
   where something1 is always 1-3 lines and something2 is zero or one line   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)