Specify dialect of csv output in frictionless extract

[fjuniorr]

When using the --csv argument of frictionless extract it's possible to control the dialect of the CSV output (eg. use ; as delimiter)?

My use case is for a transforming an excel file to csv, but frictionless transform appears to be too much.

[lwinfree]

Hi @fjuniorr! Does this example help? https://framework.frictionlessdata.io/docs/guides/describing-data/#describing-a-resource

[fjuniorr]

I don't think so @lwinfree, but thanks for the pointer!

[aborruso]

Hi @fjuniorr! Does this example help? https://framework.frictionlessdata.io/docs/guides/describing-data/#describing-a-resource

Hi @lwinfree a note about it.

I have this csv input

a;b
1;3
7;9

and this resource description

dialect:
  delimiter: ;
encoding: utf-8
format: csv
hashing: md5
name: input
path: input.csv
profile: tabular-data-resource
schema:
  fields:
    - name: a
      type: integer
    - name: b
      type: integer
scheme: file

As you can see I have delimiter: ;, but if I run

frictionless extract --type resource input.yml --csv

the output does not match its yaml definition and I have the , as field delimeter.

a,b
1,3
7,9

Probably it could be interesting to have the option, a flag, to force an output that matches the yaml file.

What do you think about?

[lwinfree]

noting this issue, which deals with this question: https://github.com/frictionlessdata/project/issues/719

[roll]

@lwinfree @fjuniorr @aborruso
I think it might be covered by a feature request in the issue above

[aborruso]

I think it might be covered by a feature request in the issue above

Thank you @roll. Do I must change something in my issue?

[roll]

No, I don't think so but I also added this one frictionlessdata/frictionless-py#1012 (as a more broad feature)