feature request: multiprocessing the validation of a single file

[pierrecamilleri]

Using the frictionless validate --parallel flag at the command line or validate(parallel = True) in the python code does not seem to trigger any parallel treatment.

Performance of the validation on a moderately large csv file (~30 mb) does not change with or without this option. In addition,
monitoring cpu shows that only 1 core seems to be solicited.

In the case of a datapackage, the command does not even run (see #1644)

To reproduce

• URL to data
• URL to schema

time frictionless validate --schema schema.json --parallel data.csv vs time frictionless validate --schema schema.json data.csv

[pierrecamilleri]

After looking into the code, it does not seem to be meant for multiprocessing of a single file (which would really be a neat feature), but instead the parallel processing of files in a datapackage.

I tag feature for the feature request and documentation as there is a clear lack of documentation on this option.