I'm writing a lot of Markdown for input to pandoc, so my files contain a lot of Unicode non-breaking spaces (U+00A0). I'd like those to search as whitespace with
isearch-forward, so I tried adding the character to my
search-whitespace-regexp variable. But it doesn't work. I suspect there might be a UTF-8 issue somewhere.
(set (make-local-variable 'search-whitespace-regexp) "[ \t\r\n\f~\u00a0]+")
and I've also tried just hard-coding the character between the square brackets. Neither works. By contrast, I can search the
~ character as white space just fine---so I know something is working.
How can I persuade Emacs that the U+00A0 should be accepted as a space character, for purposes of search?
I'm running Emacs 24.5, a stock Emacs from Debian jessie.
Edited to add: following a comment from steveooo, I've determined that
M-x eval (search-forward-regexp "cise[ \t\r\n\f~\u00a0]6")
correctly finds the string I'm looking for, but
isearch-forward still does not work.
Edited to add: In Emacs 25, nothing seems to have changed.