RE: new-flex-completion-style

[Drew Adams <drew.adams@oracle.com>]

> Earlier, I forgot to tell you that I had visited your page (this is
> where I learned the "scatter" alias) and for the record I'm
> (re)implementing your "Scatter-Match (Flex) Completion" thing:
> 
>   https://urldefense.proofpoint.com/v2/url?u=https-
> 3A__www.emacswiki.org_emacs_Icicles-5F-2D-5FCompletion-5FMethods-5Fand-
> 5FStyles-
> 23toc8&d=DwIFaQ&c=RoP1YumCXCgaWHvlZYR8PZh8Bv7qIrMUB65eapI_JnE&r=kI3P6lj
> Gv6CTHIKju0jqInF6AOwMCYRDQUmqX22rJ98&m=27ypntXsw7g-l7Nfe2JaeDgtsetIas-
> Qq30T2GdC_58&s=C4LaFFUPksrZ2l6B04Ovt4T-kt8Tz4xwZN2EO-z8iuk&e=
> 
> Tho the equivalent regexp I'm using here is ".*a.*b.*c.*" not "a.*b.*c"
> as your page states.

With Icicles regexp matching for completion you don't need the
initial `.*'.

> Anyway, what kind of scoring, if any, do you use for sorting
> the matches to that particular completion style?

I don't use any scoring for scatter matching.

Icicle users usually have a bunch of sort orders they can use
(which set can depend on the particular command and on user
preferences).  One of them is "by flx score", which uses a
score provided by `flx.el'.  With scatter matching users can
use any sort order allowed for the given command.

This page has a list of most of the sort orders available
by default:

https://www.emacswiki.org/emacs/Icicles_-_Sorting_Candidates

When completing, in the mode-line of `*Completions*' you see
the number of matching candidates (e.g. 925), the match
method (e.g. `scatter completion'), and the sort order (e.g.
`by flx score').

---

[The two Icicles matchings using (1) `fuzzy-match.el' and
(2) Swank fuzzy matching each impose a sort order - users
can't switch sort orders when they use these two kinds of
matching.

These two are the kinds of fuzzy matching available for
"prefix" Icicles completion, i.e., using `TAB'.  The others,
such as scatter and levenshtein, are available for "apropos"
completion, i.e., using `S-TAB'.]