1.2.1-beta02

• Checking for poor alignments as part of the hash collision check rather than requiring two nearby hashblock matches, which is faster and more accurate
• Considering --max-penalty-span even when there is a perfect alignment
• Using --ambiguity-penalty as unaligned penalty
• No longer getting stuck if reading queries produces an error
• No longer documenting --no-gapmers because it shouldn't be needed. It's still supported for now, though.

1.2.1-beta01

• Slightly more thorough about finding alignments
• No longer sometimes missing an alignment when indel extension penalty is very low

1.2.0

Rereleasing 1.2.0-beta11 without a qualifier, to indicate stability

1.2.0-beta11

• Adding --version and --help arguments
• No longer getting stuck if setup fails
• Reminding the user about --num-threads if it isn't specified
• Linking to Mapper-Demos

1.2.0-beta10

Writing some cache files atomically

• When calling the api using multiple threads with caching enabled, this can prevent an error where one thread is writing a file and another one is trying to read it

1.2.0-beta09

Api.newDatabase support for cache directory to reuse analyses of the reference

1.2.0-beta08

Fix for --no-output

1.2.0-beta07

fixing off-by-1 error in mutations file deletion start index

1.2.0-beta06

• Copying --out-mutations from QuickVariants, so Mapper can directly generate a file listing the mutations in the queries relative to the reference
• Also copying the vcf-filtering parameters from QuickVariants, so Mapper can be asked to filter out some variants from its .vcf file
• Adding an explanatory header to the top of the vcf and mutations files

1.2.0-beta05

• Enabling the cache more often when using the Api
• SequenceAlignment.formatVerbose now outputting query length rather than alignment length