Added
- The read processor's constructor now establishes a root logger if that's
not already been done by a client application.
- API functions to the `logs` module that assist clients in providing and
parsing logging options, and with integrating and using a logging system
in their applications.
Changed
- Previously, an attempt was made to optimize distribution of read chunks
(e.g., by chromosome) by interleaving the read chunk keys (e.g., chromosome
names) according to the genomic size (in base pairs) of the chromosome. This
behavior is now turned off by default, as doing so seems to accelerate real
runtime.
- The name for the function with which to fetch a file that's been registered
with `pararead` changes from `fetch` to the clearer `fetch_file`, to
distinguish this from a function that, say, fetches a block of reads or
genomic positions/regions from a file.
- Simplified use of more information-rich development mode in logging system.
- The name for the function to create and configure a logger changes from
`setup_pararead_logger` to the more general `setup_logger` to reflect the
intention for this to serve as a common infrastructure for other repositories
and projects.