Re: Specific Handling for text/calendar MIME Type

2019-07-21 Thread Kevin J. McCarthy

On Sun, Jul 21, 2019 at 01:12:19PM -0700, Andrew Marks wrote:

What do you think about implementing something like ~M for
$index_format? %M is already taken for something else.I'd really like
these special emails to be quickly identifiable in the index.


Maybe index-format-hook:
 set index_format="... %1@hascal@ ..."
 index-format-hook  hascal  "~M text/calendar"  "C"

--
Kevin J. McCarthy
GPG Fingerprint: 8975 A9B3 3AA3 7910 385C  5308 ADEF 7684 8031 6BDA


signature.asc
Description: PGP signature


Re: Specific Handling for text/calendar MIME Type

2019-07-21 Thread Andrew Marks
Very Nice! This must mean that there is other attachment counting code
somewhere. 

What do you think about implementing something like ~M for
$index_format? %M is already taken for something else.I'd really like
these special emails to be quickly identifiable in the index.

-Andrew

On Sun, Jul 21, 2019 at 12:15:39PM -0700, Kevin J. McCarthy wrote:
> On Sun, Jul 21, 2019 at 11:40:21AM -0700, Andrew Marks wrote:
> >2. Another option is to implement a new limit, specific for finding
> >emails which contain text/calendar, this could sort of be a duplicate of
> >using attachments with ~X, but perhaps the attachment counting code
> >could be isolated and re-used for some more specific limits to be added?
> 
> You may want to try the ~M pattern.  As you mention, ~X doesn't recurse
> multipart/alternative but I'm not convinced it should.
> 
> -- 
> Kevin J. McCarthy
> GPG Fingerprint: 8975 A9B3 3AA3 7910 385C  5308 ADEF 7684 8031 6BDA




Re: Specific Handling for text/calendar MIME Type

2019-07-21 Thread Kevin J. McCarthy

On Sun, Jul 21, 2019 at 11:40:21AM -0700, Andrew Marks wrote:

2. Another option is to implement a new limit, specific for finding
emails which contain text/calendar, this could sort of be a duplicate of
using attachments with ~X, but perhaps the attachment counting code
could be isolated and re-used for some more specific limits to be added?


You may want to try the ~M pattern.  As you mention, ~X doesn't recurse
multipart/alternative but I'm not convinced it should.

--
Kevin J. McCarthy
GPG Fingerprint: 8975 A9B3 3AA3 7910 385C  5308 ADEF 7684 8031 6BDA


signature.asc
Description: PGP signature


Specific Handling for text/calendar MIME Type

2019-07-21 Thread Andrew Marks
I'll get on my soapbox and state that text/calendar mime attachments to
e-mail is a horrific way to manage events and/or invites. Also this is
currently the bane of my existence because no mail reader makes it easy
to manage the many text/calendar attachments I receive. I think the
RFC's involved could be and should be used by calendaring software
independently of e-mail, the two should never have been combined.

END SOAPBOX

Regardless, I want to make it easier to manage text/calendar mime
attachments in mutt and there are several features I'm considering
working on to do this:

1. The easiest approach is to use attachments, however it's not
currently possible because multipart/alternative is not recursively
counted by the attachments code, and these text/calendar attachments are
typically children of multipart/alternative. To fix this I can make a
setting which toggles this behavior, so I can count all members of
multipart/alternative and change my attachments settings to only count
text/calendar.

2. Another option is to implement a new limit, specific for finding
emails which contain text/calendar, this could sort of be a duplicate of
using attachments with ~X, but perhaps the attachment counting code
could be isolated and re-used for some more specific limits to be added?

3. A third thing I've been considering is adding more specific support
to index_format for this, again, using attachments and %X could
potentially do this, but it could be nice to have a specific option for
finding and highlighting e-mail that has text/calendar mime attachments

In all cases, the bodies of every e-mail must be inspected, so
performance will not be great if your mailbox is not local. I don't see
any way around this unless there is some secret header that I am unaware
of.

If nobody has any better ideas or constructive feedback, I'll probably
start with number 1 and go from there.

-Andrew