The plain text parser shouldn't use regular expressions to detect special characters

The plain text parser uses a regular expression to detect if a character is in a list of characters. This regular expression is used for every single character. When profiling this with the block parser (so creating an XDOM), this regular expression dominated the running time of the plain text parser - 190ms out of 260ms total running time for the "parse" method. While the plain text parser is still very fast, there is absolutely no reason for this relatively slow regular expression, a simple BitSet with the characters to match would be both simpler to use and faster.