Repository navigation
Trim Galore v2.5.0
What's Changed
- fix(report): accumulate bp_after_cutadapt on the parallel path (#541) by @FelixKrueger in #543
- fix(report): count unpaired rescues where they are written (#544) by @FelixKrueger in #545
- chore(deps): drop the unused memchr and log direct dependencies by @BenjaminDEMAILLE in #550
- chore: remove dead report and clump_only code by @BenjaminDEMAILLE in #563
- fix(cli): reject NaN, infinite and negative -e / --max_n values by @BenjaminDEMAILLE in #556
- fix(demux): sort tied summary rows by barcode by @BenjaminDEMAILLE in #560
- fix(demux): reject duplicate barcodes in the sample sheet by @BenjaminDEMAILLE in #553
- fix(demux): reject sample names that are empty or contain a path separator by @BenjaminDEMAILLE in #554
- fix(clock): label the CAGT count as the pairs that lack it by @BenjaminDEMAILLE in #559
- fix(reader): treat a closed reader channel without the EOF marker as an error by @BenjaminDEMAILLE in #565
- fix(report): cap adapter-table expect and max.err at the adapter length by @BenjaminDEMAILLE in #562
- fix(bam): reject paired uBAM mates whose read names differ by @BenjaminDEMAILLE in #547
- fix(bam): accept SEQ=* records and UInt32 tags on the uBAM paths by @BenjaminDEMAILLE in #548
- fix(cli): accept non-UTF-8 arguments instead of panicking by @BenjaminDEMAILLE in #557
- fix(cli): reject --basename in modes whose output naming ignores it by @BenjaminDEMAILLE in #558
- fix(fastq): reject records that are not one 4-line FASTQ record by @BenjaminDEMAILLE in #564
- fix(demux): compress per-barcode outputs on one thread each by @BenjaminDEMAILLE in #555
- perf: drop per-read allocations in the adapter DP and 5' clip by @BenjaminDEMAILLE in #566
- fix(report): count only too-short pairs on the PE length-cutoff line by @BenjaminDEMAILLE in #561
- fix(report): give --max_length its own line in the paired text report by @FelixKrueger in #567
- feat(ubam): multithreaded uBAM input and output with --cores by @Justype in #529
- refactor(trimmer): share the per-read trim, stats and filter step across all paths by @BenjaminDEMAILLE in #549
- refactor: share the hardtrim clip and the clump bin types by @BenjaminDEMAILLE in #551
- refactor(main): move the clump-only and trim dispatch out of main() by @BenjaminDEMAILLE in #552
- test(ubam): check that a failed parallel sink finish publishes nothing by @BenjaminDEMAILLE in #569
- docs(threading): scope the N+4 thread model to FASTQ input; uBAM input runs 2N+4 by @BenjaminDEMAILLE in #570
New Contributors
Full Changelog: v2.4.0...v2.5.0