The needle is never part of the hay -- it is always a separate string.

(And even if needle and haystack were identical, the routine works fine,
the case does not need special treatment.)

This fixes https://savannah.gnu.org/bugs/?48660.

This fixes https://savannah.gnu.org/bugs/?48578.

This allows the user to specify which other characters, besides the
default alphanumeric ones, should be considered as part of a word, so
that word operations like Ctrl+Left and Ctrl+Right will pass them by.

Using this option overrides the option --wordbounds.

This fulfills https://savannah.gnu.org/bugs/?47283.

So, slightly speed up the functions that check for those.

Invalid multibyte sequences get depicted with the Replacement Character,
and unassigned codepoints are shown as if they were a space.  Both have
a width of one.

Use knowledge of UTF-8 instead of converting to wide characters first.

Any control character is represented by a ^ plus an ASCII character.

Instead of showing the upper control codes like this:

   ^À ^Á ^Â ^Ã ^Ä ^Å ^Æ ^Ç ^È ^É ^Ê ^Ë ^Ì ^Í ^Î ^Ï
   ^Ð ^Ñ ^Ò ^Ó ^Ô ^Õ ^Ö ^× ^Ø ^Ù ^Ú ^Û ^Ü ^Ý ^Þ ^ß

show them like this:

   ^` ^a ^b ^c ^d ^e ^f ^g ^h ^i ^j ^k ^l ^m ^n ^o
   ^p ^q ^r ^s ^t ^u ^v ^w ^x ^y ^z ^{ ^| ^} ^~ ^=

The lower control codes continue to be shown like this:

   ^@ ^A ^B ^C ^D ^E ^F ^G ^H ^I ^J ^K ^L ^M ^N ^O
   ^P ^Q ^R ^S ^T ^U ^V ^W ^X ^Y ^Z ^[ ^\ ^] ^^ ^_

The representation of DEL (0x7F) continues as ^?.

Further, use knowledge of UTF-8 to avoid a roundtrip through
wide characters.

By removing from their main loops a condition that occurs just once.

When running in a non-UTF locale, and when strncasecmp() suffers from
the same defect as strncmp(), make sure not to pass a length with the
high bit set.

The function is_cntrl_mbchar() has always been called successfully before
calling control_mbrep(), so the passed character *is* a valid sequence.

Instead of parsing every multibyte character twice, first with
parse_mbchar() and then with mbtowc(), just let mbtowc() do all
the work.  This makes searching for a fixed string twice as fast.

This also gets rid of four variables and lots of memory allocations.
(And, more importantly: it stops nano messing up the internal state
of the multibyte-to-wide character conversion, and thus would make
the calls to mbtowc_reset() superfluous.)

(The measurable effect (during long searches, for example) is zero, though.)


git-svn-id: svn://svn.savannah.gnu.org/nano/trunk/nano@5773 35c25a1d-7b9e-4130-9fde-d3aeb78583b8

and two other minute tweaks.


git-svn-id: svn://svn.savannah.gnu.org/nano/trunk/nano@5649 35c25a1d-7b9e-4130-9fde-d3aeb78583b8

Apparently there is /some/ state somewhere after all.  Don't have time
now to figure out where exactly.


git-svn-id: svn://svn.savannah.gnu.org/nano/trunk/nano@5369 35c25a1d-7b9e-4130-9fde-d3aeb78583b8

git-svn-id: svn://svn.savannah.gnu.org/nano/trunk/nano@5354 35c25a1d-7b9e-4130-9fde-d3aeb78583b8

git-svn-id: svn://svn.savannah.gnu.org/nano/trunk/nano@5316 35c25a1d-7b9e-4130-9fde-d3aeb78583b8

but only as far back as such a character can possibly be.
Speedup suggested by Mark Majeres.


git-svn-id: svn://svn.savannah.gnu.org/nano/trunk/nano@5147 35c25a1d-7b9e-4130-9fde-d3aeb78583b8

        * src/chars.c (addstrings): Needs to be available even on
          non-utf-8 sustems.
        * nano-regress: Added --disable-utf8 to regression check



git-svn-id: svn://svn.savannah.gnu.org/nano/trunk/nano@4933 35c25a1d-7b9e-4130-9fde-d3aeb78583b8

and handling multibyte characters correctly.
*Patch by Mark Majeres.*


git-svn-id: svn://svn.savannah.gnu.org/nano/trunk/nano@4908 35c25a1d-7b9e-4130-9fde-d3aeb78583b8

git-svn-id: svn://svn.savannah.gnu.org/nano/trunk/nano@4826 35c25a1d-7b9e-4130-9fde-d3aeb78583b8

git-svn-id: svn://svn.savannah.gnu.org/nano/trunk/nano@4772 35c25a1d-7b9e-4130-9fde-d3aeb78583b8

git-svn-id: svn://svn.savannah.gnu.org/nano/trunk/nano@4765 35c25a1d-7b9e-4130-9fde-d3aeb78583b8

git-svn-id: svn://svn.savannah.gnu.org/nano/trunk/nano@4734 35c25a1d-7b9e-4130-9fde-d3aeb78583b8