Suggestion

[BenjaminHofstetter]

I like your approach. Just as a suggestion. You can actually add semantics to your CSV. Like what is a Row. CSV files often only provide the file name or column headers to determine the content of a row. But you can add this explicit.

e.g.

 {
            "virtual": true,
            "propertyUrl": "rdf:type",
            "valueUrl": "schema:Person"
 }

You can do the same with properties. E.g. use schema.org properties if possible.

{
  "@context": ["http://www.w3.org/ns/csvw", {
    "schema": "http://schema.org/",
    "rdf": "http://www.w3.org/1999/02/22-rdf-syntax-ns#",
    "ex": "http://example.org/vocab#"
  }],
  "tables": [
    {
      "url": "contacts.csv",
      "tableSchema": {
        "aboutUrl": "http://example.org/person/{email}",
        "columns": [
          {
            "name": "name",
            "titles": "Name",
            "datatype": "string",
            "required": true,
            "propertyUrl": "schema:name"
          },
          {
            "name": "email",
            "titles": "Email",
            "datatype": "string",
            "required": true,
            "propertyUrl": "schema:email"
          },
          {
            "name": "age",
            "titles": "Age",
            "datatype": "integer",
            "propertyUrl": "schema:age"
          },
          {
            "name": "subscribed",
            "titles": "Subscribed",
            "datatype": "boolean",
            "propertyUrl": "ex:newsletterSubscribed"
          },
          {
            "virtual": true,
            "propertyUrl": "rdf:type",
            "valueUrl": "schema:Person"
          }
        ]
      }
    }
  ]
}

As a side effect of your Editor isn't just an editor. It can communicate the semantics of the content.