bug#35796: 26.1; tabulated-list-sort user-friendliness

bug#35796: 26.1; tabulated-list-sort user-friendliness

[Boruch Baum]

This is a feature request for function `tabulated-list-sort' to make it
more user-friendly.

When the function is called interactively without a prefix-argument, it
should prompt the user (using `completing-read'?) for which column to
sort by (taken from the car of each element of `tabulated-list-format').
When called with a prefix argument, that would be redundant since the
prefix argument *is* the sort column.

The mode should also have a means to directly and explicitly ask for the
direction of sort. Here are a few implementation ideas for that, but any
better one that you can come up would be fine by me:

1) Create explicit functions `tabulated-list-sort-ascending' and
   `tabulated-list-sort-descending'. Allow the user to bind them as 'S'/'s'
   or 'O'/'o' or whatever other binding that desire. ('S' and 'o' are
   currently used by default by the mode).

2) Add code in the existing function to allow for negative prefix
   arguments. Then interpret positive numbers as a request for an
   ascending sort and negative numbers as a request for a descending
   sort.

--
hkp://keys.gnupg.net
CA45 09B5 5351 7C11 A9D1  7286 0036 9E45 1595 8BC0

bug#35796: 26.1; tabulated-list-sort user-friendliness

[Lars Ingebrigtsen]

Boruch Baum <boruch_baum@gmx.com> writes:

> This is a feature request for function `tabulated-list-sort' to make it
> more user-friendly.
>
> When the function is called interactively without a prefix-argument, it
> should prompt the user (using `completing-read'?) for which column to
> sort by (taken from the car of each element of `tabulated-list-format').
> When called with a prefix argument, that would be redundant since the
> prefix argument *is* the sort column.

It sorts based on the current column (I just tried it in a C-x C-b
buffer), and it seems to work fine for me.

> The mode should also have a means to directly and explicitly ask for the
> direction of sort. Here are a few implementation ideas for that, but any
> better one that you can come up would be fine by me:

It reverses the direction when you hit it twice, which seems logical to
me.  So I think this works as it should.
-- 
(domestic pets only, the antidote for overdose, milk.)
   bloggy blog: http://lars.ingebrigtsen.no