Currently, I'm using rows to import a CSV file like this:

LAST NAME, FIRST NAME;LANGUAGE;Address 1;Address 2;RATE;Mobile;Home;Fax;email
Fontes, Bernardo;Portuguese, English;0, Rua dos Bobos;São Paulo, RJ 0101010;$10;1234-4321;1234-4321;1234-4321;[email protected]

But is acceptable to accept an entry only with the name and email values like the following:

LAST NAME, FIRST NAME;LANGUAGE;Address 1;Address 2;RATE;Mobile;Home;Fax;email
Fontes, Bernardo, Ali;;;;;;;;[email protected]

When trying to import this CSV, the lib gets confused and stops use the ; as delimiter and starts to use ,. So, it would be nice to have a way to overwrite the delimiter parameter to be passed to csv.reader here.

I know that I can do that by passing a custom Dialect object within the dialect parameter, but, in my opinion, this is less straightforward than following Python API.