perfarce bug w/ tip of hg-main

Frank Kingswood frank at kingswood-consulting.co.uk
Thu Sep 9 02:30:03 CDT 2010


On 09/09/10 06:00, Steve Borho wrote:
> When filectx.cmp() was changed to take another fctx rather than a
> string, it broke the perfarce extension:

Thanks for the advance warning. It can be fixed easily in perfarce by 
reverting part of change de0b3a57fad55052f2cd83cc27be93d7a5917e06

              # forget about copies with changes to the data
              forget = []
              for c in cpy:
-                if ctx2[c].data() != ctx1[cpy[c]].data():
+                if ctx2.flags(c) != ctx1.flags(c) or 
ctx1[c].cmp(ctx2[c].data()):
                      forget.append(c)
              for c in forget:
                  del cpy[c]

...but I liked the suggestion of efficiency in the cmp() method.

Frank



More information about the Mercurial-devel mailing list