Re: upcase-region changes 'i' to 'P'

[Kenichi Handa]

In article <address@hidden>, Juri Linkov <address@hidden> writes:

> Handling of `i' is broken in the current CVS state.  For example,
> trying to isearch for `i' in etc/HELLO (`C-h h C-s i C-s ...') reveals
> several problems:

> 1. isearch finds `i' on characters from non-Latin scripts, which have
>    nothing in common with `i'.

> 2. not all `i' characters are highlighted in `lazy-highlight' face.

Thank you for finding this bug.  I've just installed a fix.
Please try again.

> 3. isearch becomes very slow while searching for `i'.
>    Every successive `C-s' takes at least 1 sec.

As `i' now can match with dotless-i and dotted-I, we can't
use BM search to search for a text containing `i'.  So,
searching for such a text should become slower.  But, at
least for interactive use, I think the slowness is not that
match after my fix.  Do you still see the slowness?

---
Kenichi Handa
address@hidden