Querying added/removed lines against all commits will stress the server a lot. I'd suggest to clone the repository and run git command against the local copy. It will also be much flexible.

Still hope to have this feature, the problem of pressure I can compensate with a good hardware configuration. I hope you can consider this feature, thanks!

I want to extend this API to a statistics module that can separately count how much code each person has modified and deleted on each branch of each project, and can generate some statistical graphs that I can paginate to query the data stored in the intermediate media, without pulling the data from onedev every time.

To give an example: clone a linux repository takes 5 min on my machine, however getting added/removed lines of all commits takes 2 hours, and eats up to 6G mem since each file in history needs to be read and compared. Also it is not easy to do pagination as git commits are DAG.

Can the built-in contribution stats serve your purpose? An example: https://code.onedev.io/projects/onedev-server/stats/contribs (ignore the main graph, this bug will be fixed in next patch release)

This data is accumulated as commits coming in, and can add a RESTful endpoint for programmatic access.

@hackerdom