Understand callstack of subversion
Hi , I am new to programming. I am asked to extend libsvn_fs to create new implementation for sqlite. So far , I have read the code , checked out and built subversion and went through schema of tables mentioned in berkeley DB implementation. Can anyone please explain how to figure out which functions to implement and how to proceed hereonwards. Really appreciate your help , Tanvi
The Berkeley DB implementation is pretty much archaic. Nobody uses anything other than FSFS these days. The new format will be FSX when it's right. You'll need to say more about what the expectations are around using sqlite. FWIW, sqlite is used today in the working copies. It's unlikely to be appropriate for the repositories themselves.
Hi , I had come across a very old link - https://svn.apache.org/repos/asf/subversion/trunk/notes/sql-backend considering svn's library design I thought it should not be very tough to write a database plugin. I am not able to identify whether database will be component parallel to libsvn_fs_fs or it will sit below fsfs layer. I am still not sure whether what I have in mind is feasible or not but the final idea is checkin checkout operations from svn will store files in database
Be forewarned that Subversion repositories currently can take objects in the 10's of GB (or larger) as checkins. I've seen SVN repos at over 1.5TB because of this. With that as a warning, have FUN!
Thanks. I will keep that in mind
1-5 of 5
Reply to this discussion
You cannot edit posts or make replies: You should be logged in before you can post.