IGD Search is returning 0 hits in some test cases

[donaldcampbelljr]

When using the dev branch of IGD, I notice a bug where I'm creating an .igd from a directory and my search is returning 0 hits.

However, if I create an igd from a single source file and query that source file against the igd, it appears to be working fine.

Originally posted by @donaldcampbelljr in #44 (comment)

[sanghoonio]

Findings from debugging session today:

1. some bed files result in negative avg size when making a .igd
2. for the issue involving index out of bounds: involves gData length in get_overlaps()

Test bed files for making the igd are here: https://myuva-my.sharepoint.com/:u:/g/personal/sp5fd_virginia_edu/ESZDBr8_mjBDjl9Zr8xAkt8BJKVjoMecpdrGZE3Dqg6Bog?e=BKxNR8

[sanghoonio]

Testing with small bed files shows some consistent differences between the Rust and C versions of IGD search (at least with these test files):

1. Rust version does not capture hits at index 0
2. Rust version does not match the number of hits in C version or GenomicRanges::countOverlaps(), and often just returns 1 hit per index