v0.6.0

v0.6.0 modifies the gbsketch strategy to reduce the number of calls to the NCBI REST API (necessary due to policy changes, but a better strategy regardless). Rather than downloading all files via the API, we download a single file via the API that contains direct fetch links to all requested accessions. We then download accessions in parallel -- currently limited to 30 simultaneously. It also enables streaming processing for both gbsketch and urlsketch, reducing memory requirements. Finally, it improves batch restart for --keep-fasta by avoiding re-downloading and overwriting of completed downloads.

Note: gbsketch now relies on gzip processing (internal crc32 checks) rather than md5sum checks to ensure we have complete downloads. urlsketch will still check md5sums if they are provided.

What's Changed

• MRG: upd readme: installation, authors by @bluegenes in #200
• Update README.md by @bluegenes in #201
• add more api key docs by @bluegenes in #206
• MRG: fix gbsketch NCBI downloads by using dehydrate-rehydrate approach by @bluegenes in #222
• update some crates by @bluegenes in #229
• Bump openssl from 0.10.69 to 0.10.71 by @dependabot in #205
• MRG: gbsketch streaming processing by @bluegenes in #230
• upd zip to 2.6 by @bluegenes in #231
• MRG: streaming urlsketch processing by @bluegenes in #232
• MRG: Consolidate reused code across gbsketch/urlsketch by @bluegenes in #233
• MRG: optionally avoid re-downloading existing FASTA when using --keep-fasta by @bluegenes in #235
• MRG: upd version to 0.6.0; pin rust to 1.74 by @bluegenes in #234

Full Changelog: v0.5.0...v0.6.0