Fix: Change int to long long in save_sparse to avoid overflow.

[ErjieWu]

Linked Issue

Fix #6730

What's changed?

• Change int to long long in save_sparse to avoid overflow.

[ErjieWu]

Ready for merge. Another question is the speed for get_S calculation.
In #6460 , we've tested the bottleneck for get_S calculation for systems with not so large atom numbers and found std::map restrict the performance. However, for large systems, I/O is also bottleneck. For R=[0 0 0], the number of non-zero element can be larger than 2^31 and the io for this R is extremely slow (Data write in order in 3 lines). Maybe there are some solutions to this.

[mohanchen]

LGTM

[JTaozhang]

thanks a lot for help. So the problem will be disappeared in the new version?

Best,
Tao