bug#12796: Optimize `ido-completing-read' for larger lists with flex matching enabled

[Dmitry Gutov]

Tags: patch

Currently ido re-filters the full candidates list after every change in 
the minibuffer. With long candidates list and with flex matching enabled 
(like it's often the case with certain third-party packages, namely smex 
and ido-ubiquitous), as soon as ido switches to using flex matching, 
each update takes a noticeable fraction of a second. Even if there's no 
matches anymore for the current input.

If I decide to type quickly but make a typo in one of the first 
characters, I often need to wait a few seconds until I can fix the typo 
or start anew.

This patch adds a simple cache that keeps track of the current matching 
settings (prefix, regexp, or no), and checks the input against a 
previously entered string. If the latter is a prefix of the former (and 
regexp matching is disabled), then we can use the matches from the 
former input as the candidates list for the current one.

Any objections?

ido-speed.diff
Description: Text document