Am Freitag, 3. Dezember 2004 20:20 schrieb Andre Poenitz:

> Why is  ' < 0 ' equivalent to  ' != npos' ?

The type changed from a signed one to a unsigned one. Based on a 
suggestion by Angus I introduced npos, because we were not sure wether 
values < 0 were used as a special value. The assert checks for valid 
indices, and in this respect ' < 0 ' and ' != npos' are equivalent.


Georg

Reply via email to