Some obfuscation rules can be VERY long, and as a result difficult to both read 
and maintain when they aren't generated automatically.  Some other rules can 
also get rather long.

Thus, a simple line continuation facility for rules is suggested.  I believe 
the suggested format is simple enough that it should be quite trivial to 
recognize and handle in parsing, and trivial to glue the rules back together 
when generating the compiled rules.

Allow line continuation with the normal backslash-newline technique.  The 
following line could begin with any amount of whitespace, and the rule 
continuation text would be the first non-whitespace on the line. This second 
line could also be continued in the same manner, etc.  The rule would end with 
a line not containing line continuation.

You could limit this continuation to only inside the regex delimiters, but it 
would be simpler and probably safer to just allow it anywhere in a rule line.  
The various pieces would simply be glued back together by eliminating the line 
continuation characters and any associated whitespace.  (Thus you could not 
break a rule just before whitespace in the regex, but you could perhaps break 
it *after* whitespace in the regex if you wanted to allow that.)