Re: [PATCH] Issue 898 atomic rename, implement svn_fs_rename

cmpilato@collab.net writes:

 I'm wondering if this discussion isn't a bit premature.

It's quite possible.

 I'm thinking that we might come up with a more informed approach
 after we clear the iz #1003 hurdle.

I'm looking at post 1003 code already :)

Here's my working copy

$ svn st ../svn
M S ../svn/subversion/include/svn_fs.h
    S ../svn/subversion/libsvn_fs/bdb/changes-table.c
M S ../svn/subversion/libsvn_fs/dag.c
M S ../svn/subversion/libsvn_fs/dag.h
M S ../svn/subversion/libsvn_fs/fs.h
    S ../svn/subversion/libsvn_fs/structure
M S ../svn/subversion/libsvn_fs/tree.c
    S ../svn/subversion/libsvn_fs/util/fs_skels.c
M S ../svn/subversion/tests/libsvn_fs/changes-test.c
M S ../svn/subversion/tests/libsvn_fs/fs-test.c

I switched the 'S' files from trunk to issue-1003-dev creating a
working copy that is essentially trunk@HEAD plus the 1003 code. Next
I merged the trunk into the working copy and got four conflicting
files. I resolved those manually, and fixed some compiler warnings.
Now it builds and the regression tests pass. I only needed to raise
two new issues during the process...

Do you want me to check in this change? It will mean that the
issue-1003-dev branch will only build if you switch individual files
as I have done above. On the other hand it should simplify the review
and merge. Or would you prefer a patch? It's not that big

$ svn diff ../svn | diffstat
 include/svn_fs.h | 62 ++++++++------
 libsvn_fs/dag.c | 21 ++++
 libsvn_fs/dag.h | 5 +
 libsvn_fs/fs.h | 23 -----
 libsvn_fs/tree.c | 98 +++++++++++++++++++---
 tests/libsvn_fs/changes-test.c | 12 --
 tests/libsvn_fs/fs-test.c | 176 +++++++++++++++++------------------------
 7 files changed, 221 insertions(+), 176 deletions(-)

and about half that consists of svn_error_createf format warnings.

-- 
Philip Martin
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org