On 16 January 2012 06:05, David García Granda <dgra...@gmail.com> wrote: > > r['sport'] value is valid and greater than 0, but the exception he got > makes me think that sport service called wasn't able to find suitable > references for it, not that sport id itself is not valid. So if I am > right, Jason's workaround overcomes the launch of exception but > prevents pytrainer to build correct statistics related to a valid > sport.
Yes, Jason's data must contain corrupt records which reference non-existant sports. > I am not really familiar with sport service, but I guess correct > approach would be to somehow create a new sport reference in it based > on valid sport id received, otherwise future entries with same sport > id will be skipped as well. Ideas from anyone? It is not wise to fix the corrupt data in this way. We need to ensure the bug that introduced the corrupt data has been eliminated and then clean any resulting corrupt data once in an upgrade script. This is basically the same issue discussed in the "release new version" thread on 2011-08-31. The full fix will be done in the activity service which I should complete in the next month. In the mean time I think Jason's workaround is acceptable; even though the totals will not be accurate at least it won't crash on corrupt data. - Nathan ------------------------------------------------------------------------------ RSA(R) Conference 2012 Mar 27 - Feb 2 Save $400 by Jan. 27 Register now! http://p.sf.net/sfu/rsa-sfdev2dev2 _______________________________________________ Pytrainer-devel mailing list Pytrainer-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/pytrainer-devel
