Usage
Wring has various subcommand tools to help keep files small.
Usage:
Options:
--install-completion Install completion for the current shell.
--show-completion Show completion for the current shell, to copy it or
customize the installation.
csv
Crawl a directory and compress csv files into parquet.
Usage:
Options:
--directory <path> [default: .]
--compression <str> compression algorithm [default: zstd]
--compression-level <int> compression level [default: 9]
--clean / --no-clean remove uncompressed files after successful
wringing [default: no-clean]
omx
Convert one or more OMX files using different data types and filters.
Usage:
Options:
glob_pattern file name or glob pattern for the OMX files to convert
[default: *.omx]
complevel compression level [default: 3]
complib compression library [default: blosc2:zstd]
--dtype-shrink <int> maximum bitwidth to use for integer and float types
[default: 32]
--n-processes <int> number of processes to use while converting files
[default: 8]
--out-dir <str> directory name for the resulting reprocessed file(s),
if not provided any '.omx' suffix is stripped from the
original file name(s) and a '.omxz' suffix is added
omx2parquet
Convert the 2-D matrices in one or more OMX files into parquet files.
Each output file is in long format, with the row and column labels of the source matrices in the first two columns and one column per matrix.
Usage:
Options:
glob_pattern file name or glob pattern for the OMX files to
convert [default: *.omx]
--out-dir <str> directory name for the resulting parquet
file(s); by default each file is written
alongside its source with a '.parquet' suffix
--matrix <str> name of a matrix to convert; repeat to select
several, or omit to convert all of them
--row-lookup <str> name of the lookup to use for row labels
--col-lookup <str> name of the lookup to use for column labels
--row-label-name <str> name of the row label column [default:
row_label]
--col-label-name <str> name of the column label column [default:
col_label]
--compression <str> parquet compression algorithm [default: zstd]
--compression-level <int> parquet compression level [default: 9]
--row-group-rows <int> target number of table rows per parquet row
group [default: 1048576]
--dtype-shrink <int> maximum bitwidth to use for integer and float
types [default: 32]
--overwrite / --no-overwrite overwrite existing parquet files [default:
no-overwrite]
tarzst
Compress a directory using the .tar.zst format.
Usage:
Options:
directory the directory to compress [required]
--archive <path> base file name for the resulting archive file(s),
if not provided a '.tar.zst' suffix is added to
the original directory name
-c, --chunk-gigs <float> chunk the resulting archive into parts that are no
larger than this many gigabytes
untarzst
Decompress a directory stored in .tar.zst format.
Usage:
Options: