The behaviour differs depending on whether the target file name already exists or not. It's usually a safety mechanism, and there are at least 3 different cases:
In this case svn mv
should work as follows:
$ svn mv old_file_name new_file_name
A new_file_name
D old_file_name
$ svn stat
A + new_file_name
> moved from old_file_name
D old_file_name
> moved to new_file_name
$ svn commit
Adding new_file_name
Deleting old_file_name
Committing transaction...
In this case, the target file needs to be removed explicitly, before the source file can be renamed. This can be done in the same transaction as follows:
$ svn mv old_file_name new_file_name
svn: E155010: Path 'new_file_name' is not a directory
$ svn rm new_file_name
D new_file_name
$ svn mv old_file_name new_file_name
A new_file_name
D old_file_name
$ svn stat
R + new_file_name
> moved from old_file_name
D old_file_name
> moved to new_file_name
$ svn commit
Replacing new_file_name
Deleting old_file_name
Committing transaction...
In the output of svn stat
, the R
indicates that the file has been replaced, and that the file has a history.
In this case, the content of the local file would be lost. If that's okay, then the file can be removed locally before renaming the existing file.
$ svn mv old_file_name new_file_name
svn: E155010: Path 'new_file_name' is not a directory
$ rm new_file_name
$ svn mv old_file_name new_file_name
A new_file_name
D old_file_name
$ svn stat
A + new_file_name
> moved from old_file_name
D old_file_name
> moved to new_file_name
$ svn commit
Adding new_file_name
Deleting old_file_name
Committing transaction...