Add new option: search_word_boundary

[Mikk3lRo]

Summary

New attempt at #2896 in a single commit.

The default / current word boundary regex ^|\\b|\\s is remarkably bad at actually finding word boundaries in languages that use non-unicode characters. A word boundary is basically detected after any non-ascii character (fx. ü, å, ø and æ to mention just a few - but there are MANY).

I've looked into possibilities, and unfortunately there doesn't seem to be any way to get decent word-boundary detection for anything except ascii in javascripts RegExp implementation... without either using a third-party library or including some 4k+ characters in the string.

Therefore, I don't see any way to reliably detect word boundaries with any pre-set, hardcoded regex.

Turning it into an option means that people can at least set something appropriate for their individual language and / or use case if they care about word boundaries being detected in "weird" places.

Please double-check that:

• All changes were made in CoffeeScript files, not JavaScript files.
• You used Grunt to build the JavaScript files and tested them locally.
• You've updated both the jQuery and Prototype versions.
• You haven't manually updated the version number in package.json.
• If necessary, you've updated the documentation.

References

First partial PR from: #2894

Should solve this issue: #2862