Package net.oneandone.mork.regexpr

Regular expressions.

See: Description

Class Summary

Class	Description
Action	stores the result from visiting a node
Choice	Left | Right, a little bit more general.
Loop	Loop with <= 1 iterations.
Range	Character ranges.
RegExpr	Regular Expression.
Sequence
Symbol	Symbol constant for a regular expression.
Without	Left | Right, a little bit more general.

Exception Summary

Exception	Description
ActionException

Package net.oneandone.mork.regexpr Description

Regular expressions. For Sequence and Alternative there is a choice between a body of two sub-expression or a list of sub-expressions. I choose the list because it tends to simplify the code, for example be having One and Zero RegExprs as special cases.

TODO
o atoms/ranges of arbitrary bit length
  o default length 16; adjustable
  o atom ::= code [ ":" bits ]
  o an atom is translated into n single-bit transitions
  o FA.mergeTransitions(m)
    o compose m transitions into single transitions
    o caution with end states
  o replace range tests in TableFA by array indexing;
    e.g. scan char by 4x4 transitions
  o caution: reading from an input stream raised byte order issues