The XWiki syntax parser shouldn't use String#replaceAll

The XWiki 2.0 and 2.1 syntax parser currently uses String#replaceAll to remove "~" from parsed words an spaces. While the performance impact is minor, it is still inefficient as this compiles a regular expression. See also https://medium.com/javarevisited/micro-optimizations-in-java-string-replaceall-c6d0edf2ef6 and https://rules.sonarsource.com/java/tag/regex/RSPEC-5361/