Create Tabular Data Package with CLI

[rgaiacs]

I can use frictionless describe to generate a datapackage.json for me, for example

$ frictionless describe --type package --json iris.csv
{
  "resources": [
    {
      "path": "iris.csv",
      "name": "iris",
      "profile": "tabular-data-resource",
      "scheme": "file",
      "format": "csv",
      "hashing": "md5",
      "encoding": "utf-8",
      "schema": {
        "fields": [
          {
            "type": "number",
            "name": "Sepal.Length"
          },
          {
            "type": "number",
            "name": "Sepal.Width"
          },
          {
            "type": "number",
            "name": "Petal.Length"
          },
          {
            "type": "number",
            "name": "Petal.Width"
          },
          {
            "type": "string",
            "name": "Species"
          }
        ]
      }
    }
  ],
  "profile": "data-package"
}

How do I change "profile": "data-package" to "profile": "tabular-data-package"?

[aborruso]

How do I change "profile": "data-package" to "profile": "tabular-data-package"?

Does it exist? We do not have it here https://framework.frictionlessdata.io/docs/guides/describing-data/#describe-functions

[rgaiacs]

Hi @aborruso.

Does it exist?

https://specs.frictionlessdata.io/ has Data Package and Tabular Data Package. Tabular Data Package specs says

There MUST be a profile property with the value tabular-data-package

So, I was expecting to to be possible to set the profile to tabular-data-package`.

Is this related to v1 vs v1.1 or v2 of Frictionless Data?

[roll]

Hi @rgaiacs, the framework doesn't add the tabular-data-package profile as it might be too restrcting and we already have resource marked as tabular-data-resource. TBH the concept of Tabular Data Package after intorucing Tabular Data Resources is not fully clear to me

[rgaiacs]

the framework doesn't add the tabular-data-package profile as it might be too restrcting and we already have resource marked as tabular-data-resource.

I understand this part.

TBH the concept of Tabular Data Package after intorucing Tabular Data Resources is not fully clear to me

I'm studying the "reference" implementation in R and it has profile = "tabular-data-package".

Based on your comment, my understanding is that the R implementation could use "profile": "data-package".

[roll]

Yes it could. Tabular Data Package requires all the resources to be a tabular CSV which is quite a strong condition. In real-life majority of datasets has tables AND supporting files like PDFs/etc so in my opinion after introducing Tabular Data Resource concept using Tabular Data Package is not really needed anymore