PerCederberg.Grammatica.Runtime
Class TokenNFA

System.Object
   |
   +--TokenNFA

internal class TokenNFA

extends System.Object

A non-deterministic finite state automaton (NFA) for matching tokens. It supports both fixed strings and simple regular expressions, but should perform similar to a DFA due to highly optimized data structures and tuning. The memory footprint during matching should be near zero, since no heap memory is allocated unless the pre-allocated queues need to be enlarged. The NFA also does not use recursion, but iterates in a loop instead.

Since:

1.5

AddRegExpMatch

public void AddRegExpMatch( string pattern, bool ignoreCase, TokenPattern value );

Adds a regular expression match to this automaton. New states and transitions will be added to extend this automaton to support the specified string. Note that this method only supports a subset of the full regular expression syntax, so a more complete regular expression library must also be provided.

Parameters:

pattern - the regular expression string

ignoreCase - the case-insensitive match flag

value - the match value

Throws:

RegExpException - if the regular expression parsing failed

AddTextMatch

public void AddTextMatch( string str, bool ignoreCase, TokenPattern value );

Adds a string match to this automaton. New states and transitions will be added to extend this automaton to support the specified string.

Parameters:

str - the string to match

ignoreCase - the case-insensitive match flag

value - the match value

Match

public int Match( ReaderBuffer buffer, TokenMatch match );

Checks if this NFA matches the specified input text. The matching will be performed from position zero (0) in the buffer. This method will not read any characters from the stream, just peek ahead.

Parameters:

buffer - the input buffer to check

match - the token match to update

Returns:

the number of characters matched, or zero (0) if no match was found

Throws:

IOException - if an I/O error occurred