🚀 go-pugleaf

In article <1992Dec23.131946.14537@rdg.dec.com>  
jacksonc@uproar.enet.dec.com (Chris Jackson @ IME) writes:
> 
> My experience seems to suggest that using NXWriteTypes() to archive a 
> 'char *' only works if the 'char *' is not NULL.  If it *is* NULL, then 
> a memory violation results in 'strlen()' in NeXT code when the object 
> concerned is read from the archive.
> 
> Can anyone out there confirm that this is the case and whether or not it 
> is the expected behaviour ?  If it is the expected behaviour, I presume 
> I have to explicitly code around it ?  Next question - is there a 
> recommended way of doing so ?
 
Well, yes... but this is true for all of C, not just NeXT read/write  
calls.

When an object is first instantiated, all ivars are set to 0/NULL. If you  
never assign anything to the pointer, it stays NULL, and reads/writes  
NULL. That is exactly as it should be. The problem is that strlen (and  
str[n]cpy and str[n]cmp), at least as implemented in the NeXT libraries,  
do not guard against NULL in source or dest. Frankly, I think the overhead  
would be worth it, but they chose not to.

Write your own strxxx functions (they're not that complicated) and link  
them in to override the default library implementations, or give yours a  
unique name and use a 

#define strlen mystrlen

to substitute yours for the library copy at compile time.

--
Gregory H. Anderson          | Thus spake the master programmer: 
Master Programmer / Manager  | "Let the programmers be many and  
Anderson Financial Systems   | the managers few; then all will be 
greg@afs.com  (Nextmail OK)  | productive." -- Tao of Programming