test-convert-cvs failing

Jesse Glick jesse.glick at sun.com
Sat Jan 26 15:41:19 CST 2008


Patrick Mézard wrote:
>> Is anyone else getting failures from test-convert-cvs? [...]
>>
>> I also get _occasional_ failures on test-http.
> 
> http://marc.info/?l=mercurial-devel&m=119888263727709&w=2
> 
> Can you confirm it fixes your issue?

I can hardly verify that it fixes the test-http failures, as these do 
not happen reproducibly.

It does *not* fix the failure in test-convert-cvs for me, though it does 
seem to change the output a bit:

ERROR: test-convert-cvs output changed
--- Expected output
+++ Test output
@@ -71,38 +71,34 @@
  U b/c
  T a
  T b/c
-checking in src/b/c,v
  % convert again
  connecting to cvsrepo
  scanning source...
  sorting...
  converting...
-0 ci2
  a
  a
  c
-d
+c
+c
  % convert again with --filemap
  connecting to cvsrepo
  scanning source...
  sorting...
  converting...
-0 ci2
  c
-d
-4 ci2 files: b/c
+c
+c
  3 ci1 files: b/c
  2 update tags files: .hgtags
  1 ci0 files: b/c
  0 Initial revision files: b/c
-o  5 (branch) ci2 files: b/c
+o  4 () ci1 files: a b/c
  |
-| o  4 () ci1 files: a b/c
-| |
-| o  3 () update tags files: .hgtags
-| |
-| o  2 () ci0 files: b/c
-|/
+o  3 () update tags files: .hgtags
+|
+o  2 () ci0 files: b/c
+|
  | o  1 (INITIAL) import files:
  |/
  o  0 () Initial revision files: a b/c


I normally skip test-convert-svn-source anyway (missing feature: 
subversion python bindings) so I don't know if it helps with that. The 
patch does not cause any other failures.



More information about the Mercurial-devel mailing list