Dnarecords

Performance
* Metadata reader implemented with parallel read ([`da4754f`](https://github.com/amanas/dnarecords/commit/da4754f61719269e1f84f27836029b5d7f3154a2))

Fix
* Fixed dense_shape error on variant_wise writer ([`523a8e6`](https://github.com/amanas/dnarecords/commit/523a8e6cce8779f9363e09533b1d3281481e1599))

Performance
* Use metadata taste when building DNARecordsReader decoder ([`94785a9`](https://github.com/amanas/dnarecords/commit/94785a92cc014b68952dfbea8754e52236163fbe))

Fix
* Added integer to spark<->tf types_dict ([`98e8a2f`](https://github.com/amanas/dnarecords/commit/98e8a2f1adee8aebe47f847db29f093b5115607f))

Fix
* Stop generating property names with chrchr ([`31a9172`](https://github.com/amanas/dnarecords/commit/31a917204422c54a9219f040f9a4f8d7546de66b))

Fix
* Fixed scalability issues related with too many small files generated ([`b4ba095`](https://github.com/amanas/dnarecords/commit/b4ba09581e803f3117f5dba4ec5751ee939c03eb))

Performance
* Added metadata taste functionality to the reader. Improved performace of block transposition ([`26b2353`](https://github.com/amanas/dnarecords/commit/26b2353c28fc5642c84002002c51d0363ec84c85))