[svn.haxx.se] · SVN Dev · SVN Users · SVN Org · TSVN Dev · TSVN Users · Subclipse Dev · Subclipse Users · this month's index

svnsync tests may fail to FAIL?

From: Neels Janosch Hofmeyr <neels_at_elego.de>
Date: Sat, 23 May 2009 00:49:17 +0200

Hi all,

in svnsync_tests.py, I find these lines in the overall run_test() function:

[[[
  run_sync(dest_sbox.repo_url)
  run_copy_revprops(dest_sbox.repo_url)
]]]

As I was browsing the output of a test, I noticed that the properties were
being synced *twice*. Once by svn_sync(), once by run_copy_revprops().

In fact, if I remove the "run_copy_revprops()", all tests still pass fine.
So I ask epg, who added the line in r29600:

Like this, if a normal sync fails to shovel the props over properly, the
copy_revprops call may mask that by doing it properly afterwards. The test
would fail to FAIL!

Instead, I think we should have either one or two tests specifically for
copy_revprops, or run the relevant or all tests twice, once with and once
without the copy_revprops part.

Any comments?

Thanks,
~Neels

------------------------------------------------------
http://subversion.tigris.org/ds/viewMessage.do?dsForumId=462&dsMessageId=2353055

Received on 2009-05-23 00:49:40 CEST

This is an archived mail posted to the Subversion Dev mailing list.