> It seems real difficult to get PL_regprev correct in all cases. Below is > a patch to remove PL_regprev entirely, which doesn't fail any of the > existing tests and which I believe should offer a performance improvement. > I assume there may be problems not currently tested, particularly in the > area of in-place substitutions - can you suggest test cases that fail? > > The remaining mention of PL_regprev is in Porting/findvars - I'm not sure > what that is used for, so I have left it alone. Seems like some relic of olden times. > Hugo > --- thrdvar.h.old Tue Jan 30 18:14:40 2001 > +++ thrdvar.h Sun Apr 29 16:18:07 2001 Thanks, applied. (The PL_regprev being a char would have been broken anyway -- for Unicode. At least a UV would have been needed, but a full SV would have been better (think combining characters and normalization).) -- $jhi++; # http://www.iki.fi/jhi/ # There is this special biologist word we use for 'stable'. # It is 'dead'. -- Jack CohenThread Previous | Thread Next