Changelog 2022.6.30.dev1

Jun 30, 2022 • Anders Berkeman, Carl Drougge, and Sofia Hörberg

Highlights

Smart tab mode (-T/--tab-length).
- Longer separation between fields than the tab interval, so that fields are more likely to stay aligned.
- On by default for terminal output, use -T8 for the old behaviour (or use -t<SEPARATOR>).
- Use -T/ to turn it on for file/pipe output without specifying values.
- Configure it with -T<TABLEN>/<FIELDLEN>/<MINLEN>. Everything aligns to TABLEN, fields start at FIELDLEN intervals and each value is separated by at least MINLEN. Default is 8/16/2.
- You can also use NAME=VALUE, e.g. --tab-length min=3/field=24.
New --chain-length and --stop-ds options that work like on ds.chain().
- You can specify them several times if you have several chains (i.e. have specified several datasets to chain from).

Rename --debug option to --keep-temp-files, to reflect the only remaining functionality of that option.

DotDicts (which all dicts in options are) are now OrderedDicts and if you pass an OrderedDict in the call to the method order is preserved. (If you pass a normal dict the keys will be sorted.)
Unset job/dataset values now provide some of the methods and properties a set value would have with empty results. E.g.:
- jobs.foo.datasets() == DatasetList([]) if jobs.foo is unset.
- jobs.foo.files() == set() if jobs.foo is unset.
- jobs.foo.load() == None if jobs.foo is unset.
- datasets.bar.lines[n] == 0 if datasets.bar is unset.
- datasets.bar.columns == {} if datasets.bar is unset.
- Most other properties are None or ''.
job.datasets is a DatasetList, so you can use it like a DatasetChain.
New job.finish_early(result=None) function to end a job early (without running later steps).
DatasetWriter now supports default= on all column types.

Note: Python 2 is still supported, but will most likely be unsupported soon.