Consider separating words based on /\W/ instead of whitespace.

[lewang]

Consider:

(flx-get-heatmap-str "*group*") ; [84 -2 -3 -4 -5 -6 82]

The first "*" is considered the beginning of the word instead of "g". This should be changed to group non-word non-space characters together as words.

See: #63

[jeancroy]

The problem with \w is that in most implementation it is not unicode friendly.
You'll split on accented letters like é, û, à, and probably any non latin character.

whitelist of a few separator will be more unicode friendly.

[PythonNut]

The problem with \w is that in most implementation it is not unicode friendly.
You'll split on accented letters like é, û, à, and probably any non latin character.

Emacs has excellent Unicode support, and handles non-latin characters correctly.