Open GoogleCodeExporter opened 9 years ago
Hi Gal,
To implement such a feature may be easier than we think. Since all data are
saved to a local folder, you can access that folder directly without going
through Rabbit, for example you could write a small program (which may or may
not be an Eclipse plugin) that observes on that directory (Apache Commons IO
2.0 will be great for this, see org.apache.commons.io.monitor) for changes and
decide which files and when you want to upload to your server.
At the moment when new events are saved to XML files by Rabbit, they are merged
with the older events of the same month into the same file, so working out what
has been changed since last push will be a bit difficult. It will be easier to
push the whole file of that month (the files are small), which means there
won't be much work on the server side either, just replace any existing files.
If you look into the folder where Rabbit saves data to, (the default is
"Rabbit" under you home directory), you will see that there are a number of
folders there, each one contains data from a particular Eclipse workspace (you
can probably tell by the folder names), inside each folder are the event files,
each file contains a single type of events for a particular month, for example,
"fileEvents-2011-01.xml" contains all file events for March 2011.
I hope this helps, I look forward to see more about your idea.
Lae
Original comment by llaec...@gmail.com
on 1 Apr 2011 at 9:06
Original issue reported on code.google.com by
gal.dol...@gmail.com
on 30 Mar 2011 at 3:16