On 2013-03-26 12:25, Daniel Shahaf wrote: > Anil Bakshi wrote on Tue, Mar 26, 2013 at 16:37:25 +0530: >> So I will go with your suggestion: svndumpfilter exclude >> /E_Learning/Development/Project1 < repo.dump > filteredDump.dump >> >> Please correct me. > > You might try this command: > > % grep -a '^Node-path:' < repo.dump | head > > to see what paths inside the dump file look like. > > .oO ( maybe we should have an 'svndumpfilter info' command? I suppose > it could basically cat the dumpfile, except: file reps would be omitted; > dir reps would be omitted (unless we figure out something sensible to > dowith them, eg "5 children"); properties would be omitted (but in the > future we could list propnames or propnames and propvalues). ) >
Here the script normalize-dump.py (in the svn tar file) comes is really handy. From my own experience with big dumps it is ways faster to process the dump with this script and then grep for Node-path/Node-copyfrom-path. Additional hint for Windows users: run svndumpfilter this way: python -u $path/svndumpfilter.py
