com.oracle.truffle.regex.tregex.parser.ast

Class CalcASTPropsVisitor

java.lang.Object

com.oracle.truffle.regex.tregex.parser.ast.visitors.RegexASTVisitor

com.oracle.truffle.regex.tregex.parser.ast.visitors.DepthFirstTraversalRegexASTVisitor

com.oracle.truffle.regex.tregex.parser.ast.CalcASTPropsVisitor

public class CalcASTPropsVisitor
extends DepthFirstTraversalRegexASTVisitor

This visitor computes various properties of RegexAST and its RegexASTNodes, in two passes.

• RegexASTNode.getMinPath():
  • The minPath of BackReference, PositionAssertion, LookBehindAssertion and LookAheadAssertion nodes is the minimum number of CharacterClass nodes that need to be traversed in order to reach the node.
  • The minPath of MatchFound nodes is undefined (or is always 0). Their minPath is never set by CalcASTPropsVisitor.
  • The minPath of Sequence and Group nodes is the minimum number of CharacterClass nodes that need to be traversed (starting at the AST root) in order to reach the end of the node. The minPath field of Sequence nodes is used as a mutable iteration variable when traversing their children (see CalcASTPropsVisitor.visit(CharacterClass)). The resulting value after the traversal holds the minimum number of CharacterClass nodes that need to be traversed to reach the end of the Sequence. The same holds for Group nodes.
  • The contents of LookAroundAssertions are treated separately, so their minPath values have nothing to do with their parent expression.
• RegexASTNode.getMaxPath() is set analogous to RegexASTNode.getMinPath(), but without taking loops (Group.isLoop()) into account.
• RegexASTNode.isDead(): The following nodes are marked as dead:
  • PositionAssertions whose minimum path is greater than zero (type PositionAssertion.Type.CARET in forward mode and PositionAssertion.Type.DOLLAR in backward mode). Note that this algorithm will e.g. not mark the dollar assertion in /(?=a$)bc/ as dead, since it has a (reverse) minimum path of 0 inside the look-ahead assertion.
  • CharacterClasses that don't match anything (SortedListOfRanges.matchesNothing()).
  • Sequences that contain a dead node.
  • Groups where all alternatives are dead.
  • RegexASTSubtreeRootNodes whose child group is dead.
• RegexASTNode.startsWithCaret()/RegexASTNode.endsWithDollar():
  • Sequences that start with a caret PositionAssertion / end with a dollar PositionAssertion.
  • Groups where all alternatives start with a caret / end with a dollar.
• RegexASTNode.hasCaret()/RegexASTNode.hasDollar():
  • Sequences that contain a caret / dollar PositionAssertion.
  • Groups where any alternatives contains a caret / dollar PositionAssertion.
• RegexAST.getReachableCarets()/RegexAST.getReachableDollars(): all caret/dollar PositionAssertion that are not dead are added to these lists.
• RegexAST.getLookArounds()}: all reachable LookAroundAssertions are added to these lists.

See Also:

RegexAST.getReachableCarets(), RegexAST.getReachableDollars(), RegexAST.getLookArounds(), RegexASTNode.hasCaret(), RegexASTNode.hasDollar(), RegexASTNode.startsWithCaret(), RegexASTNode.endsWithDollar(), RegexASTNode.getMinPath(), RegexASTNode.getMaxPath(), RegexASTNode.isDead(), RegexAST.createPrefix()

Constructor Summary

Constructors

Constructor and Description
CalcASTPropsVisitor(RegexAST ast)

Method Summary

Modifier and Type	Method and Description
protected void	init(RegexASTNode runRoot)
protected void	leave(Group group)
protected void	leave(LookAheadAssertion assertion)
protected void	leave(LookBehindAssertion assertion)
void	leaveLookAroundAssertion(LookAroundAssertion assertion)
static void	run(RegexAST ast)
protected void	visit(BackReference backReference)
protected void	visit(CharacterClass characterClass)
protected void	visit(Group group)
protected void	visit(LookAheadAssertion assertion)
protected void	visit(LookBehindAssertion assertion)
protected void	visit(PositionAssertion assertion)
protected void	visit(Sequence sequence)

Methods inherited from class com.oracle.truffle.regex.tregex.parser.ast.visitors.DepthFirstTraversalRegexASTVisitor

isForward, isReverse, leave, run, runReverse

Methods inherited from class com.oracle.truffle.regex.tregex.parser.ast.visitors.RegexASTVisitor

doLeave, doVisit

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

CalcASTPropsVisitor

public CalcASTPropsVisitor(RegexAST ast)

Method Detail

run

public static void run(RegexAST ast)

init

protected void init(RegexASTNode runRoot)

Overrides:

init in class DepthFirstTraversalRegexASTVisitor

visit

protected void visit(BackReference backReference)

Overrides:

visit in class DepthFirstTraversalRegexASTVisitor

visit

protected void visit(Group group)

Overrides:

visit in class DepthFirstTraversalRegexASTVisitor

leave

protected void leave(Group group)

Overrides:

leave in class DepthFirstTraversalRegexASTVisitor

visit

protected void visit(Sequence sequence)

Overrides:

visit in class DepthFirstTraversalRegexASTVisitor

visit

protected void visit(PositionAssertion assertion)

Overrides:

visit in class DepthFirstTraversalRegexASTVisitor

visit

protected void visit(LookBehindAssertion assertion)

Overrides:

visit in class DepthFirstTraversalRegexASTVisitor

leave

protected void leave(LookBehindAssertion assertion)

Overrides:

leave in class DepthFirstTraversalRegexASTVisitor

visit

protected void visit(LookAheadAssertion assertion)

Overrides:

visit in class DepthFirstTraversalRegexASTVisitor

leave

protected void leave(LookAheadAssertion assertion)

Overrides:

leave in class DepthFirstTraversalRegexASTVisitor

leaveLookAroundAssertion

public void leaveLookAroundAssertion(LookAroundAssertion assertion)

visit

protected void visit(CharacterClass characterClass)

Overrides:

visit in class DepthFirstTraversalRegexASTVisitor