On 3 August 2010 01:37, karl williamson <perlbug-followup@perl.org> wrote: > There is a bug in the last patch I made to charnames which introduces a > regression, when all charnames calls in a program are of the form 'use > charnames ()'. The bug is only in blead. This is nice. Thanks, applied to bleadperl. > This series of commits fixes that bug, and another in the test file, in > which an empty line got added to the end of the .t. git diff doesn't > detect it; which I guess is a bug in that (perhaps fixed in a later > version than I'm running). The blank line got interpreted as > overwriting the name for NULL with the empty string. I don't know if > the blank line is in what I committed, or what is already in blead, but > I added code to skip all such lines. Note to committer: check and > remove any empty line at the end of the file. > > Also, I converted the .t to use srand's seed, instead of computing it > itself, and use fewer tests of algorithmically determined names. >Thread Previous