I'm very new to Subversion, but I've used other revision control systems like ClearCase for years.
My boss asked me to fix this project so that it could be built with Maven instead of Ant. One of the important things I had to do to was to move src/com
to src/main/java/com
, and move test/com
to src/test/java/com
, which I did using the svn mv
command. I foolishly assumed that since I used Subversion commands to move the directories, that Subversion would then know that things had been moved. And when I merged my branch into the trunk, it appeared to work. But now somebody else just finished work on a branch that he branched off before my work. So we go to merge his stuff into trunk, and basically Subversion appears to think “ok, he made changes to src/com/foo/bar/baz.java
, but that directory doesn’t exist any more, so it’s irrelevant, so discard it” instead of what I expected, which was “ok, he made changes to src/com/foo/bar/baz.java
, but src/com
has been moved, so I need to merge that into src/main/java/com/foo/bar/baz.java
”.
Is there a way to make Subversion do the revision management, or am I going to be manually merging this guy’s changes for the next two days?
To answer your question directly:
Is there a way to make Subversion do the revision management, or am I going to be manually merging this guy’s changes for the next two days?
You should be able to make it a little bit easier on yourself.
One thing that you can do to alleviate some of the pain (assuming a layout as follows)
/branch/foo/src/com
/branch/foo/test/com
/trunk/src/main/java/com
/trunk/src/test/java/com
Before you moved src/com to src/main/java/com and test/com to src/test/java/com you could have done:
cd $TRUNK
svn merge -r N:M http://server/branch/foo .
What you could do now is:
cd $TRUNK
svn merge -r N:M http://server/branch/foo/src/com src/main/java/com
svn merge -r N:M http://server/branch/foo/test/com src/test/java/com
Hope this helps to save you some time.