Improves fallback code in Utf8String::extractWords()

[Sesquipedalian]

Deals with this:

SMF/Sources/Unicode/Utf8String.php

Lines 706 to 713 in 9aa2a6f

	/*
	* This is a sad, weak substitute for the IntlBreakIterator.
	* It works well enough for European languages, but it fails badly
	* for many Asian languages. To improve it will require adding more
	* data to our Unicode data files and then writing code to implement
	* the Unicode word break algorithm.
	* See https://www.unicode.org/reports/tr29/#Word_Boundaries
	*/

Specifically, this PR implements the default Unicode word break algorithm as fallback code for when the IntlBreakIterator class is unavailable. As noted in the Unicode documentation for this algorithm:

It is not possible to provide a uniform set of rules that resolves all issues across languages or that handles all ambiguous situations within a given language. The goal for the specification presented in this annex is to provide a workable default.

In other words, it is still best to use IntlBreakIterator if it is available, because it can adapt to the specific rules of a language, but the default is still a decent option for most languages most of the time.