bypassing $reverse_alias in $index_format

[Cameron Simpson]

On 22Oct2020 13:08, Cameron Simpson <cs at cskk.id.au> wrote:
>On 29Jul2020 09:13, Cameron Simpson <cs at cskk.id.au> wrote:
>>On 28Jul2020 13:27, Kevin J. McCarthy <kevin at 8t8.us> wrote:
>>>I don't know if this helps, but have you tried a format pipe *just* in
>>>one of the index-format-hook with %f and using the script to extract
>>>the name part of the full address?
>>>
>>>    set index_format="%D %@from_part@ %S %?M?(%M) ?%?H?[%H] ?%s%* %?y?  y? %4c"
>>>    index-format-hook from_part '%f polyname' "~/bin/extractname '%f'|"
>>>    index-format-hook from_part ~A '%-15.15F'
>>
>>I have not.. Might be expensive. I'll see how that goes. Thanks!
>
>Just FYI, I've done this instead:
>    https://bitbucket.org/cameron_simpson/css/commits/9bdab9c16452223a58508939d43b04c23820bdb2

Also, I've just reread Kevin's suggestion above more carefully. And 
tried it. It isn't all that bad, since it only runs the script for 
messages from the target group.

However, I think it uncovers a performance issue with mutt's index 
rendering. I noticed that when there were several candidate addresses 
the scrolling slowed down (scrolling driven by my keyboard autorepeat).

It is as though when the index is scrolling, every displayed line is 
reconsidered instead of just the lines coming into view (or those not 
previously rendered). It might benefit from a cache of rendered index 
lines, niche though this case is, if index-format-hook with a shell 
escape becomes widely used.

I suppose such a cache could be conditional on shelling out to fulfil 
the format...

Just my impressions,
Cameron Simpson <cs at cskk.id.au>