Releases: markschl/seqtool
v0.4.0-beta.3 (2025-01-05)
Release Notes
Besides dependency updates, the release includes:
- The ability to append records to the output file(s) instead of overwriting
(--append
) (7fe1893). This is especially useful for thesplit
command. - A new
--counts
option in thesplit
command, which returns the record counts
for each file (565b42c). This saves an extrast count
command. - Added some error messages to prevent panics (d91b1fc, 8156c45)
Install seqtool 0.4.0-beta.3
Install prebuilt binaries via shell script
curl --proto '=https' --tlsv1.2 -LsSf https://github.com/markschl/seqtool/releases/download/v0.4.0-beta.3/seqtool-installer.sh | sh
Install prebuilt binaries via powershell script
powershell -ExecutionPolicy ByPass -c "irm https://github.com/markschl/seqtool/releases/download/v0.4.0-beta.3/seqtool-installer.ps1 | iex"
Install prebuilt binaries via Homebrew
brew install seqtool
Download seqtool 0.4.0-beta.3
File | Platform | Checksum |
---|---|---|
seqtool-aarch64-apple-darwin.tar.xz | Apple Silicon macOS | checksum |
seqtool-x86_64-apple-darwin.tar.xz | Intel macOS | checksum |
seqtool-x86_64-pc-windows-msvc.zip | x64 Windows | checksum |
seqtool-x86_64-pc-windows-msvc.msi | x64 Windows | checksum |
seqtool-aarch64-unknown-linux-gnu.tar.xz | ARM64 Linux | checksum |
seqtool-x86_64-unknown-linux-gnu.tar.xz | x64 Linux | checksum |
seqtool-x86_64-unknown-linux-musl.tar.xz | x64 MUSL Linux | checksum |
v0.4.0-beta.2 (pre-release)
Release Notes
This release comes with substantial improvements and lots of internal code improvements
and also some notable bugfixes.
See CHANGELOG for all details.
Download seqtool 0.4.0-beta.2
File | Platform | Checksum |
---|---|---|
seqtool-aarch64-apple-darwin.tar.xz | Apple Silicon macOS | checksum |
seqtool-x86_64-apple-darwin.tar.xz | Intel macOS | checksum |
seqtool-x86_64-pc-windows-msvc.zip | x64 Windows | checksum |
seqtool-x86_64-pc-windows-msvc.msi | x64 Windows | checksum |
seqtool-x86_64-unknown-linux-gnu.tar.xz | x64 Linux | checksum |
v0.3.0
Binary renamed: The binary was renamed from seqtool
to st
, because
typing the rather long name repeatedly can be tiring (-: The documentation
was updated accordingly
New commands:
- view command for viewing sequences in terminal with colored background
Changes / additions:
-
Seqtool now handles quality scores associated with sequences. This includes:
- Support for converting between FASTQ format variants
- The
exp_err
statistics variable, which represents the total number of expected
errors in a sequence, according to the quality scores. This allows filtering
by sequence quality as done byfastq_filter
from USEARCH and VSEARCH - 454-style QUAL files can be read an written
-
--format
renamed to--fmt
, and--outformat
renamed to--to
-
The former
txt
format is nowtsv
(including the option--to-tsv
).
Bug fixes:
seq_io
was updated, containing two small bug fixes. Both bugs were very
unlikely to be hit and of low severity
v0.2.3
New commands:
- concat command
- interleave command
Changes / additions:
-
The find command is now much faster for approximate searching.
The starting position (f:start variable) can be different in some
cases where several alignments with the same distance and end position are
possible.
Note: Pattern length for approximate searching is now limited to 64. -
Added a shortcut
-T
for--read-thread
. Adding this option
(and eventually--write-thread
) can speed up many commands.
Since there are no comprehensive benchmarks, the default
is to read/write in the main thread. -
New dirname variable
Bug fixes:
- Compressed files produced by the split command were still truncated
in some cases (related to bug fixed in v0.2.2). This is fixed now. - Pattern type recognition in find command was not always correct
v0.2.2
Important bugfixes:
- Writing compressed files was partly broken (truncated files) and was fixed
now, along with many unit tests - With
--fields
/--txt
/--csv
, column indices are now correctly interpreted
as 1-based, not 0-based - Fixed handling of NaN values in 'count' command
tail
returned one sequence too much.
Changes to behaviour:
-n
option removed from slice command, use a range (..n
) instead.
New features:
- Added ZSTD (de-)compression which is almost as fast as LZ4, but has a better
compression ratio - Now allows for changing the compression level (
--compr-level
) - Writing attributes to sequences removed by filtering (
--dropped
option of
find
andfilter
) is now possible. - Other small improvements and better testing, almost all options have tests now.