Gitweb can be configured to use pathinfo URLs with the gitweb.conf setting $feature{'pathinfo'}{'default'} = [1]; and appropriate .htaccess changes.

In this configuration, the URLs are prettier. For example, a 'standard' Gitweb URL like this:

https://git.example.com/?p=project.git;a=commit;h=hash_goes_here

Will change to this:

https://git.example.com/project.git/commit/hash_goes_here

SourceGitweb uses RegEx to find the parent hash for a commit. The current RegEx string only works when Gitweb is creating the links on the page with 'standard' URLs. If you try to import commits from a Gitweb instance that is using pathinfo URLs, it only imports one commit (most recent) and stops there because it can't find any parents.

I have modified the RegEx string so it can find the parent hashes in both Gitweb URL modes.