You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
ParallelFastxParser now uses all its threads on plain gzip, single-frame zstd and from_reader inputs: one thread decompresses while the workers parse the chunks.
process_parallel now supports ordered batch completion (requires_ordering, Ordered), calls set_thread_id and on_thread_complete once per worker thread, and on_batch_complete after each chunk.
ParallelFastxParser::for_each_with keeps a state per worker thread and returns these states, e.g. for k-mer tables.
ParallelFastxParser::reduce is deprecated in favor of for_each_with, since it creates a state per chunk.
ParallelFastxParser::chunks now applies to slices, mmap, BGZF and multi-frame zstd, not only uncompressed files.