Thanks. Indeed we should improve one step at a time. And this is great to see Pharo getting improved so much. I once discussed with K. Beck and he told me that indeed we often forget how powerful it is to do one little improvement at a time. So let us continue and this is cool that you got that Pharo is yours :) Stef

On 13 Mar 2019, at 18:28, Alistair Grant <akgrant0710@gmail.com> wrote:

Hi Marcus,

On Wed, 13 Mar 2019 at 11:21, Marcus Denker <marcus.denker@inria.fr> wrote:

...

Hi,

The code is barely usable, I am sure it is not complete and wrong in many places.

Hopefully it will be a bit more usable and a little less wrong when I'm finished :-)

I've been able to use it to parse and load mail headers (about 23,000 messages), so it isn't all bad. :-)

Thanks, Alistair

...

...

On 13 Mar 2019, at 11:09, Alistair Grant <akgrant0710@gmail.com> wrote:

Hi Everyone,

I'm playing with parsing and searching mail messages at the moment and am wondering about the interaction of MailMessage and MIMEDocument.

A MailMessage contains a body which is a MIMEDocument, but if I request the parts of the body I get back multiple MailMessages, i.e. the parts of a MIMEDocument are a collection of MailMesssages.

MIME documents are used outside of mail messages, e.g. HTTP, so the parts of a MIMEDocument should also be MIMEDocument's and not MailMessage's. This is also implied by the MIME description at wikipedia: https://en.wikipedia.org/wiki/MIME

Some of the method comments seem to suggest that this was put together quickly, so it is possible that the intention was to do it properly at a later date (which never happened :-)).

Am I missing something, or is it OK for me to modify MIMEDocument to return MIMEDocuments as the parts?

Thanks, Alistair