Syntax highlighting is changeable.
[phpeclipse.git] / net.sourceforge.phpeclipse / src / net / sourceforge / phpdt / internal / ui / text / JavaHeuristicScanner.java
diff --git a/net.sourceforge.phpeclipse/src/net/sourceforge/phpdt/internal/ui/text/JavaHeuristicScanner.java b/net.sourceforge.phpeclipse/src/net/sourceforge/phpdt/internal/ui/text/JavaHeuristicScanner.java
new file mode 100644 (file)
index 0000000..ebeef70
--- /dev/null
@@ -0,0 +1,801 @@
+/*******************************************************************************
+ * Copyright (c) 2000, 2004 IBM Corporation and others.
+ * All rights reserved. This program and the accompanying materials 
+ * are made available under the terms of the Common Public License v1.0
+ * which accompanies this distribution, and is available at
+ * http://www.eclipse.org/legal/cpl-v10.html
+ * 
+ * Contributors:
+ *     IBM Corporation - initial API and implementation
+ *******************************************************************************/
+package net.sourceforge.phpdt.internal.ui.text;
+
+import java.util.Arrays;
+
+import org.eclipse.jface.text.Assert;
+import org.eclipse.jface.text.BadLocationException;
+import org.eclipse.jface.text.IDocument;
+import org.eclipse.jface.text.IRegion;
+import org.eclipse.jface.text.ITypedRegion;
+import org.eclipse.jface.text.Region;
+import org.eclipse.jface.text.TextUtilities;
+
+/**
+ * Utility methods for heuristic based Java manipulations in an incomplete Java source file.
+ * 
+ * <p>An instance holds some internal position in the document and is therefore not threadsafe.</p>
+ * 
+ * @since 3.0
+ */
+public class JavaHeuristicScanner implements Symbols {
+       /** 
+        * Returned by all methods when the requested position could not be found, or if a 
+        * {@link BadLocationException} was thrown while scanning.
+        */
+       public static final int NOT_FOUND= -1;
+
+       /** 
+        * Special bound parameter that means either -1 (backward scanning) or 
+        * <code>fDocument.getLength()</code> (forward scanning).
+        */
+       public static final int UNBOUND= -2;
+
+
+       /* character constants */
+       private static final char LBRACE= '{';
+       private static final char RBRACE= '}';
+       private static final char LPAREN= '(';
+       private static final char RPAREN= ')';
+       private static final char SEMICOLON= ';';
+       private static final char COLON= ':';
+       private static final char COMMA= ',';
+       private static final char LBRACKET= '[';
+       private static final char RBRACKET= ']';
+       private static final char QUESTIONMARK= '?';
+       private static final char EQUAL= '=';
+
+       /**
+        * Specifies the stop condition, upon which the <code>scanXXX</code> methods will decide whether
+        * to keep scanning or not. This interface may implemented by clients.
+        */
+       public interface StopCondition {
+               /**
+                * Instructs the scanner to return the current position.
+                * 
+                * @param ch the char at the current position
+                * @param position the current position
+                * @param forward the iteration direction 
+                * @return <code>true</code> if the stop condition is met.
+                */
+               boolean stop(char ch, int position, boolean forward);
+       }
+       
+       /**
+        * Stops upon a non-whitespace (as defined by {@link Character#isWhitespace(char)}) character. 
+        */
+       private static class NonWhitespace implements StopCondition {
+               /*
+                * @see org.eclipse.jdt.internal.ui.text.JavaHeuristicScanner.StopCondition#stop(char)
+                */
+               public boolean stop(char ch, int position, boolean forward) {
+                       return !Character.isWhitespace(ch);
+               }
+       }
+       
+       /**
+        * Stops upon a non-whitespace character in the default partition.
+        * 
+        * @see NonWhitespace 
+        */
+       private class NonWhitespaceDefaultPartition extends NonWhitespace {
+               /*
+                * @see org.eclipse.jdt.internal.ui.text.JavaHeuristicScanner.StopCondition#stop(char)
+                */
+               public boolean stop(char ch, int position, boolean forward) {
+                       return super.stop(ch, position, true) && isDefaultPartition(position);
+               }
+       }
+       
+       /**
+        * Stops upon a non-java identifier (as defined by {@link Character#isJavaIdentifierPart(char)}) character. 
+        */
+       private static class NonJavaIdentifierPart implements StopCondition {
+               /*
+                * @see org.eclipse.jdt.internal.ui.text.JavaHeuristicScanner.StopCondition#stop(char)
+                */
+               public boolean stop(char ch, int position, boolean forward) {
+                       return !Character.isJavaIdentifierPart(ch);
+               }
+       }
+       
+       /**
+        * Stops upon a non-java identifier character in the default partition.
+        * 
+        * @see NonJavaIdentifierPart 
+        */
+       private class NonJavaIdentifierPartDefaultPartition extends NonJavaIdentifierPart {
+               /*
+                * @see org.eclipse.jdt.internal.ui.text.JavaHeuristicScanner.StopCondition#stop(char)
+                */
+               public boolean stop(char ch, int position, boolean forward) {
+                       return super.stop(ch, position, true) || !isDefaultPartition(position);
+               }
+       }
+       
+       /**
+        * Stops upon a character in the default partition that matches the given character list. 
+        */
+       private class CharacterMatch implements StopCondition {
+               private final char[] fChars;
+                
+               /**
+                * Creates a new instance.
+                * @param ch the single character to match 
+                */
+               public CharacterMatch(char ch) {
+                       this(new char[] {ch});
+               }
+               
+               /**
+                * Creates a new instance.
+                * @param chars the chars to match.
+                */
+               public CharacterMatch(char[] chars) {
+                       Assert.isNotNull(chars);
+                       Assert.isTrue(chars.length > 0);
+                       fChars= chars;
+                       Arrays.sort(chars);
+               }
+               
+               /*
+                * @see org.eclipse.jdt.internal.ui.text.JavaHeuristicScanner.StopCondition#stop(char, int)
+                */
+               public boolean stop(char ch, int position, boolean forward) {
+                       return Arrays.binarySearch(fChars, ch) >= 0 && isDefaultPartition(position);
+               }
+       }
+       
+       /**
+        * Acts like character match, but skips all scopes introduced by parenthesis, brackets, and 
+        * braces. 
+        */
+       protected class SkippingScopeMatch extends CharacterMatch {
+               private char fOpening, fClosing;
+               private int fDepth= 0;
+               
+               /**
+                * Creates a new instance.
+                * @param ch the single character to match
+                */
+               public SkippingScopeMatch(char ch) {
+                       super(ch);
+               }
+               
+               /**
+                * Creates a new instance.
+                * @param chars the chars to match.
+                */
+               public SkippingScopeMatch(char[] chars) {
+                       super(chars);
+               }
+
+               /*
+                * @see org.eclipse.jdt.internal.ui.text.JavaHeuristicScanner.StopCondition#stop(char, int)
+                */
+               public boolean stop(char ch, int position, boolean forward) {
+                       
+                       if (fDepth == 0 && super.stop(ch, position, true))
+                               return true;
+                       else if (ch == fOpening)
+                               fDepth++;
+                       else if (ch == fClosing) {
+                               fDepth--;
+                               if (fDepth == 0) {
+                                       fOpening= 0;
+                                       fClosing= 0;
+                               }
+                       } else if (fDepth == 0) {
+                               fDepth= 1;
+                               if (forward) {
+                                       
+                                       switch (ch) {
+                                               case LBRACE:
+                                                       fOpening= LBRACE;
+                                                       fClosing= RBRACE;
+                                                       break;
+                                               case LPAREN:
+                                                       fOpening= LPAREN;
+                                                       fClosing= RPAREN;
+                                                       break;
+                                               case LBRACKET:
+                                                       fOpening= LBRACKET;
+                                                       fClosing= RBRACKET;
+                                                       break;
+                                       }
+                                       
+                               } else {
+                                       switch (ch) {
+                                               case RBRACE:
+                                                       fOpening= RBRACE;
+                                                       fClosing= LBRACE;
+                                                       break;
+                                               case RPAREN:
+                                                       fOpening= RPAREN;
+                                                       fClosing= LPAREN;
+                                                       break;
+                                               case RBRACKET:
+                                                       fOpening= RBRACKET;
+                                                       fClosing= LBRACKET;
+                                                       break;
+                                       }
+                                       
+                               }
+                       }
+                       
+                       return false;
+                       
+               }
+
+       }
+       
+       /** The document being scanned. */
+       private IDocument fDocument;
+       /** The partitioning being used for scanning. */
+       private String fPartitioning;
+       /** The partition to scan in. */
+       private String fPartition;
+
+       /* internal scan state */       
+       
+       /** the most recently read character. */
+       private char fChar;
+       /** the most recently read position. */
+       private int fPos;
+       
+       /* preset stop conditions */
+       private final StopCondition fNonWSDefaultPart= new NonWhitespaceDefaultPartition();
+       private final static StopCondition fNonWS= new NonWhitespace();
+       private final StopCondition fNonIdent= new NonJavaIdentifierPartDefaultPartition();
+
+       /**
+        * Creates a new instance.
+        * 
+        * @param document the document to scan
+        * @param partitioning the partitioning to use for scanning
+        * @param partition the partition to scan in
+        */
+       public JavaHeuristicScanner(IDocument document, String partitioning, String partition) {
+               Assert.isNotNull(document);
+               Assert.isNotNull(partitioning);
+               Assert.isNotNull(partition);
+               fDocument= document;
+               fPartitioning= partitioning;
+               fPartition= partition;
+       }
+       
+       /**
+        * Calls <code>this(document, IJavaPartitions.JAVA_PARTITIONING, IDocument.DEFAULT_CONTENT_TYPE)</code>.
+        * 
+        * @param document the document to scan.
+        */
+       public JavaHeuristicScanner(IDocument document) {
+               this(document, IPHPPartitions.PHP_PARTITIONING, IDocument.DEFAULT_CONTENT_TYPE);
+       }
+       
+       /**
+        * Returns the most recent internal scan position.
+        * 
+        * @return the most recent internal scan position.
+        */
+       public int getPosition() {
+               return fPos;
+       }
+       
+       /**
+        * Returns the next token in forward direction, starting at <code>start</code>, and not extending
+        * further than <code>bound</code>. The return value is one of the constants defined in {@link Symbols}.
+        * After a call, {@link #getPosition()} will return the position just after the scanned token
+        * (i.e. the next position that will be scanned). 
+        * 
+        * @param start the first character position in the document to consider
+        * @param bound the first position not to consider any more
+        * @return a constant from {@link Symbols} describing the next token
+        */
+       public int nextToken(int start, int bound) {
+               int pos= scanForward(start, bound, fNonWSDefaultPart);
+               if (pos == NOT_FOUND)
+                       return TokenEOF;
+
+               fPos++;
+                       
+               switch (fChar) {
+                       case LBRACE:
+                               return TokenLBRACE;
+                       case RBRACE:
+                               return TokenRBRACE;
+                       case LBRACKET:
+                               return TokenLBRACKET;
+                       case RBRACKET:
+                               return TokenRBRACKET;
+                       case LPAREN:
+                               return TokenLPAREN;
+                       case RPAREN:
+                               return TokenRPAREN;
+                       case SEMICOLON:
+                               return TokenSEMICOLON;
+                       case COMMA:
+                               return TokenCOMMA;
+                       case QUESTIONMARK:
+                               return TokenQUESTIONMARK;
+                       case EQUAL:
+                               return TokenEQUAL;
+               }
+               
+               // else
+               if (Character.isJavaIdentifierPart(fChar)) {
+                       // assume an ident or keyword
+                       int from= pos, to;
+                       pos= scanForward(pos + 1, bound, fNonIdent);
+                       if (pos == NOT_FOUND)
+                               to= bound == UNBOUND ? fDocument.getLength() : bound;
+                       else
+                               to= pos;
+                       
+                       String identOrKeyword;
+                       try {
+                               identOrKeyword= fDocument.get(from, to - from);
+                       } catch (BadLocationException e) {
+                               return TokenEOF;
+                       }
+                       
+                       return getToken(identOrKeyword);
+                       
+                       
+               } else {
+                       // operators, number literals etc
+                       return TokenOTHER;
+               }
+       }
+       
+       /**
+        * Returns the next token in backward direction, starting at <code>start</code>, and not extending
+        * further than <code>bound</code>. The return value is one of the constants defined in {@link Symbols}.
+        * After a call, {@link #getPosition()} will return the position just before the scanned token
+        * starts (i.e. the next position that will be scanned). 
+        * 
+        * @param start the first character position in the document to consider
+        * @param bound the first position not to consider any more
+        * @return a constant from {@link Symbols} describing the previous token
+        */
+       public int previousToken(int start, int bound) {
+               int pos= scanBackward(start, bound, fNonWSDefaultPart);
+               if (pos == NOT_FOUND)
+                       return TokenEOF;
+               
+               fPos--;
+                       
+               switch (fChar) {
+                       case LBRACE:
+                               return TokenLBRACE;
+                       case RBRACE:
+                               return TokenRBRACE;
+                       case LBRACKET:
+                               return TokenLBRACKET;
+                       case RBRACKET:
+                               return TokenRBRACKET;
+                       case LPAREN:
+                               return TokenLPAREN;
+                       case RPAREN:
+                               return TokenRPAREN;
+                       case SEMICOLON:
+                               return TokenSEMICOLON;
+                       case COLON:
+                               return TokenCOLON;
+                       case COMMA:
+                               return TokenCOMMA;
+                       case QUESTIONMARK:
+                               return TokenQUESTIONMARK;
+                       case EQUAL:
+                               return TokenEQUAL;
+               }
+               
+               // else
+               if (Character.isJavaIdentifierPart(fChar)) {
+                       // assume an ident or keyword
+                       int from, to= pos + 1;
+                       pos= scanBackward(pos - 1, bound, fNonIdent);
+                       if (pos == NOT_FOUND)
+                               from= bound == UNBOUND ? 0 : bound + 1;
+                       else
+                               from= pos + 1;
+                       
+                       String identOrKeyword;
+                       try {
+                               identOrKeyword= fDocument.get(from, to - from);
+                       } catch (BadLocationException e) {
+                               return TokenEOF;
+                       }
+                       
+                       return getToken(identOrKeyword);
+                       
+                       
+               } else {
+                       // operators, number literals etc
+                       return TokenOTHER;
+               }
+               
+       }
+
+       /**
+        * Returns one of the keyword constants or <code>TokenIDENT</code> for a scanned identifier.
+        * 
+        * @param s a scanned identifier
+        * @return one of the constants defined in {@link Symbols}
+        */
+       private int getToken(String s) {
+               Assert.isNotNull(s);
+               
+               switch (s.length()) {
+                       case 2:
+                               if ("if".equals(s)) //$NON-NLS-1$
+                                       return TokenIF;
+                               if ("do".equals(s)) //$NON-NLS-1$
+                                       return TokenDO;
+                               break;
+                       case 3:
+                               if ("for".equals(s)) //$NON-NLS-1$
+                                       return TokenFOR;
+                               if ("try".equals(s)) //$NON-NLS-1$
+                                       return TokenTRY;
+                               if ("new".equals(s)) //$NON-NLS-1$
+                                       return TokenNEW;
+                               break;
+                       case 4:
+                               if ("case".equals(s)) //$NON-NLS-1$
+                                       return TokenCASE;
+                               if ("else".equals(s)) //$NON-NLS-1$
+                                       return TokenELSE;
+                               if ("goto".equals(s)) //$NON-NLS-1$
+                                       return TokenGOTO;
+                               break;
+                       case 5:
+                               if ("break".equals(s)) //$NON-NLS-1$
+                                       return TokenBREAK;
+                               if ("catch".equals(s)) //$NON-NLS-1$
+                                       return TokenCATCH;
+                               if ("while".equals(s)) //$NON-NLS-1$
+                                       return TokenWHILE;
+                               break;
+                       case 6:
+                               if ("return".equals(s)) //$NON-NLS-1$
+                                       return TokenRETURN;
+                               if ("static".equals(s)) //$NON-NLS-1$
+                                       return TokenSTATIC;
+                               if ("switch".equals(s)) //$NON-NLS-1$
+                                       return TokenSWITCH;
+                               break;
+                       case 7:
+                               if ("default".equals(s)) //$NON-NLS-1$
+                                       return TokenDEFAULT;
+                               if ("finally".equals(s)) //$NON-NLS-1$
+                                       return TokenFINALLY;
+                               break;
+                       case 12:
+                               if ("synchronized".equals(s)) //$NON-NLS-1$
+                                       return TokenSYNCHRONIZED;
+                               break;
+               }
+               return TokenIDENT;
+       }
+
+       /**
+        * Returns the position of the closing peer character (forward search). Any scopes introduced by opening peers
+        * are skipped. All peers accounted for must reside in the default partition.
+        * 
+        * <p>Note that <code>start</code> must not point to the opening peer, but to the first
+        * character being searched.</p>
+        * 
+        * @param start the start position
+        * @param openingPeer the opening peer character (e.g. '{')
+        * @param closingPeer the closing peer character (e.g. '}')
+        * @return the matching peer character position, or <code>NOT_FOUND</code>
+        */
+       public int findClosingPeer(int start, final char openingPeer, final char closingPeer) {
+               Assert.isNotNull(fDocument);
+               Assert.isTrue(start >= 0);
+               
+               try {
+                       int depth= 1;
+                       start -= 1;
+                       while (true) {
+                               start= scanForward(start + 1, UNBOUND, new CharacterMatch(new char[] {openingPeer, closingPeer}));
+                               if (start == NOT_FOUND)
+                                       return NOT_FOUND;
+                                       
+                               if (fDocument.getChar(start) == openingPeer)
+                                       depth++;
+                               else
+                                       depth--;
+                               
+                               if (depth == 0)
+                                       return start;
+                       }
+
+               } catch (BadLocationException e) {
+                       return NOT_FOUND;
+               }
+       }
+
+       /**
+        * Returns the position of the opening peer character (backward search). Any scopes introduced by closing peers
+        * are skipped. All peers accounted for must reside in the default partition.
+        * 
+        * <p>Note that <code>start</code> must not point to the closing peer, but to the first
+        * character being searched.</p>
+        * 
+        * @param start the start position
+        * @param openingPeer the opening peer character (e.g. '{')
+        * @param closingPeer the closing peer character (e.g. '}')
+        * @return the matching peer character position, or <code>NOT_FOUND</code>
+        */
+       public int findOpeningPeer(int start, char openingPeer, char closingPeer) {
+               Assert.isTrue(start < fDocument.getLength());
+
+               try {
+                       int depth= 1;
+                       start += 1;
+                       while (true) {
+                               start= scanBackward(start - 1, UNBOUND, new CharacterMatch(new char[] {openingPeer, closingPeer}));
+                               if (start == NOT_FOUND)
+                                       return NOT_FOUND;
+                                       
+                               if (fDocument.getChar(start) == closingPeer)
+                                       depth++;
+                               else
+                                       depth--;
+                               
+                               if (depth == 0)
+                                       return start;
+                       }
+
+               } catch (BadLocationException e) {
+                       return NOT_FOUND;
+               }
+       }
+
+       /**
+        * Computes the surrounding block around <code>offset</code>. The search is started at the
+        * beginning of <code>offset</code>, i.e. an opening brace at <code>offset</code> will not be
+        * part of the surrounding block, but a closing brace will.
+        * 
+        * @param offset the offset for which the surrounding block is computed
+        * @return a region describing the surrounding block, or <code>null</code> if none can be found
+        */
+       public IRegion findSurroundingBlock(int offset) {
+               if (offset < 1 || offset >= fDocument.getLength())
+                       return null;
+                       
+               int begin= findOpeningPeer(offset - 1, LBRACE, RBRACE);
+               int end= findClosingPeer(offset, LBRACE, RBRACE);
+               if (begin == NOT_FOUND || end == NOT_FOUND)
+                       return null;
+               return new Region(begin, end + 1 - begin);
+       }
+
+       /**
+        * Finds the smallest position in <code>fDocument</code> such that the position is &gt;= <code>position</code>
+        * and &lt; <code>bound</code> and <code>Character.isWhitespace(fDocument.getChar(pos))</code> evaluates to <code>false</code>
+        * and the position is in the default partition.   
+        * 
+        * @param position the first character position in <code>fDocument</code> to be considered
+        * @param bound the first position in <code>fDocument</code> to not consider any more, with <code>bound</code> &gt; <code>position</code>, or <code>UNBOUND</code>
+        * @return the smallest position of a non-whitespace character in [<code>position</code>, <code>bound</code>) that resides in a Java partition, or <code>NOT_FOUND</code> if none can be found
+        */
+       public int findNonWhitespaceForward(int position, int bound) {
+               return scanForward(position, bound, fNonWSDefaultPart);
+       }
+
+       /**
+        * Finds the smallest position in <code>fDocument</code> such that the position is &gt;= <code>position</code>
+        * and &lt; <code>bound</code> and <code>Character.isWhitespace(fDocument.getChar(pos))</code> evaluates to <code>false</code>.   
+        * 
+        * @param position the first character position in <code>fDocument</code> to be considered
+        * @param bound the first position in <code>fDocument</code> to not consider any more, with <code>bound</code> &gt; <code>position</code>, or <code>UNBOUND</code>
+        * @return the smallest position of a non-whitespace character in [<code>position</code>, <code>bound</code>), or <code>NOT_FOUND</code> if none can be found
+        */
+       public int findNonWhitespaceForwardInAnyPartition(int position, int bound) {
+               return scanForward(position, bound, fNonWS);
+       }
+
+       /**
+        * Finds the highest position in <code>fDocument</code> such that the position is &lt;= <code>position</code>
+        * and &gt; <code>bound</code> and <code>Character.isWhitespace(fDocument.getChar(pos))</code> evaluates to <code>false</code>
+        * and the position is in the default partition.   
+        * 
+        * @param position the first character position in <code>fDocument</code> to be considered
+        * @param bound the first position in <code>fDocument</code> to not consider any more, with <code>bound</code> &lt; <code>position</code>, or <code>UNBOUND</code>
+        * @return the highest position of a non-whitespace character in (<code>bound</code>, <code>position</code>] that resides in a Java partition, or <code>NOT_FOUND</code> if none can be found
+        */
+       public int findNonWhitespaceBackward(int position, int bound) {         
+               return scanBackward(position, bound, fNonWSDefaultPart);
+       }
+
+       /**
+        * Finds the lowest position <code>p</code> in <code>fDocument</code> such that <code>start</code> &lt;= p &lt;
+        * <code>bound</code> and <code>condition.stop(fDocument.getChar(p), p)</code> evaluates to <code>true</code>.
+        * 
+        * @param start the first character position in <code>fDocument</code> to be considered
+        * @param bound the first position in <code>fDocument</code> to not consider any more, with <code>bound</code> &gt; <code>start</code>, or <code>UNBOUND</code>
+        * @param condition the <code>StopCondition</code> to check
+        * @return the lowest position in [<code>start</code>, <code>bound</code>) for which <code>condition</code> holds, or <code>NOT_FOUND</code> if none can be found
+        */
+       public int scanForward(int start, int bound, StopCondition condition) {
+               Assert.isTrue(start >= 0);
+
+               if (bound == UNBOUND)
+                       bound= fDocument.getLength();
+               
+               Assert.isTrue(bound <= fDocument.getLength());
+               
+               try {
+                       fPos= start;
+                       while (fPos < bound) {
+
+                               fChar= fDocument.getChar(fPos);
+                               if (condition.stop(fChar, fPos, true))
+                                       return fPos;
+
+                               fPos++;
+                       }
+               } catch (BadLocationException e) {
+               }
+               return NOT_FOUND;
+       }
+       
+
+       /**
+        * Finds the lowest position in <code>fDocument</code> such that the position is &gt;= <code>position</code>
+        * and &lt; <code>bound</code> and <code>fDocument.getChar(position) == ch</code> evaluates to <code>true</code>
+        * and the position is in the default partition.   
+        * 
+        * @param position the first character position in <code>fDocument</code> to be considered
+        * @param bound the first position in <code>fDocument</code> to not consider any more, with <code>bound</code> &gt; <code>position</code>, or <code>UNBOUND</code>
+        * @param ch the <code>char</code> to search for
+        * @return the lowest position of <code>ch</code> in (<code>bound</code>, <code>position</code>] that resides in a Java partition, or <code>NOT_FOUND</code> if none can be found
+        */
+       public int scanForward(int position, int bound, char ch) {
+               return scanForward(position, bound, new CharacterMatch(ch));
+       }
+
+       /**
+        * Finds the lowest position in <code>fDocument</code> such that the position is &gt;= <code>position</code>
+        * and &lt; <code>bound</code> and <code>fDocument.getChar(position) == ch</code> evaluates to <code>true</code> for at least one
+        * ch in <code>chars</code> and the position is in the default partition.   
+        * 
+        * @param position the first character position in <code>fDocument</code> to be considered
+        * @param bound the first position in <code>fDocument</code> to not consider any more, with <code>bound</code> &gt; <code>position</code>, or <code>UNBOUND</code>
+        * @param chars an array of <code>char</code> to search for
+        * @return the lowest position of a non-whitespace character in [<code>position</code>, <code>bound</code>) that resides in a Java partition, or <code>NOT_FOUND</code> if none can be found
+        */
+       public int scanForward(int position, int bound, char[] chars) {
+               return scanForward(position, bound, new CharacterMatch(chars));
+       }
+       
+       /**
+        * Finds the highest position <code>p</code> in <code>fDocument</code> such that <code>bound</code> &lt; <code>p</code> &lt;= <code>start</code>
+        * and <code>condition.stop(fDocument.getChar(p), p)</code> evaluates to <code>true</code>.
+        * 
+        * @param start the first character position in <code>fDocument</code> to be considered
+        * @param bound the first position in <code>fDocument</code> to not consider any more, with <code>bound</code> &lt; <code>start</code>, or <code>UNBOUND</code>
+        * @param condition the <code>StopCondition</code> to check
+        * @return the highest position in (<code>bound</code>, <code>start</code> for which <code>condition</code> holds, or <code>NOT_FOUND</code> if none can be found
+        */
+       public int scanBackward(int start, int bound, StopCondition condition) {
+               if (bound == UNBOUND)
+                       bound= -1;
+               
+               Assert.isTrue(bound >= -1);
+               Assert.isTrue(start < fDocument.getLength() );
+               
+               try {
+                       fPos= start;
+                       while (fPos > bound) {
+                               
+                               fChar= fDocument.getChar(fPos);
+                               if (condition.stop(fChar, fPos, false))
+                                       return fPos;
+
+                               fPos--;
+                       }
+               } catch (BadLocationException e) {
+               }
+               return NOT_FOUND;
+       }
+       
+       /**
+        * Finds the highest position in <code>fDocument</code> such that the position is &lt;= <code>position</code>
+        * and &gt; <code>bound</code> and <code>fDocument.getChar(position) == ch</code> evaluates to <code>true</code> for at least one
+        * ch in <code>chars</code> and the position is in the default partition.   
+        * 
+        * @param position the first character position in <code>fDocument</code> to be considered
+        * @param bound the first position in <code>fDocument</code> to not consider any more, with <code>bound</code> &lt; <code>position</code>, or <code>UNBOUND</code>
+        * @param ch the <code>char</code> to search for
+        * @return the highest position of one element in <code>chars</code> in (<code>bound</code>, <code>position</code>] that resides in a Java partition, or <code>NOT_FOUND</code> if none can be found
+        */
+       public int scanBackward(int position, int bound, char ch) {
+               return scanBackward(position, bound, new CharacterMatch(ch));
+       }
+       
+       /**
+        * Finds the highest position in <code>fDocument</code> such that the position is &lt;= <code>position</code>
+        * and &gt; <code>bound</code> and <code>fDocument.getChar(position) == ch</code> evaluates to <code>true</code> for at least one
+        * ch in <code>chars</code> and the position is in the default partition.   
+        * 
+        * @param position the first character position in <code>fDocument</code> to be considered
+        * @param bound the first position in <code>fDocument</code> to not consider any more, with <code>bound</code> &lt; <code>position</code>, or <code>UNBOUND</code>
+        * @param chars an array of <code>char</code> to search for
+        * @return the highest position of one element in <code>chars</code> in (<code>bound</code>, <code>position</code>] that resides in a Java partition, or <code>NOT_FOUND</code> if none can be found
+        */
+       public int scanBackward(int position, int bound, char[] chars) {
+               return scanBackward(position, bound, new CharacterMatch(chars));
+       }
+       
+       /**
+        * Checks whether <code>position</code> resides in a default (Java) partition of <code>fDocument</code>.
+        * 
+        * @param position the position to be checked
+        * @return <code>true</code> if <code>position</code> is in the default partition of <code>fDocument</code>, <code>false</code> otherwise
+        */
+       public boolean isDefaultPartition(int position) {
+               Assert.isTrue(position >= 0);
+               Assert.isTrue(position <= fDocument.getLength());
+               
+               try {
+                       ITypedRegion region= TextUtilities.getPartition(fDocument, fPartitioning, position, false);
+                       return region.getType().equals(fPartition);
+                       
+               } catch (BadLocationException e) {
+               }
+               
+               return false;
+       }
+
+       /**
+        * Checks if the line seems to be an open condition not followed by a block (i.e. an if, while, 
+        * or for statement with just one following statement, see example below). 
+        * 
+        * <pre>
+        * if (condition)
+        *     doStuff();
+        * </pre>
+        * 
+        * <p>Algorithm: if the last non-WS, non-Comment code on the line is an if (condition), while (condition),
+        * for( expression), do, else, and there is no statement after that </p> 
+        * 
+        * @param position the insert position of the new character
+        * @param bound the lowest position to consider
+        * @return <code>true</code> if the code is a conditional statement or loop without a block, <code>false</code> otherwise
+        */
+       public boolean isBracelessBlockStart(int position, int bound) {
+               if (position < 1)
+                       return false;
+               
+               switch (previousToken(position, bound)) {
+                       case TokenDO:
+                       case TokenELSE:
+                               return true;
+                       case TokenRPAREN:
+                               position= findOpeningPeer(fPos, LPAREN, RPAREN);
+                               if (position > 0) {
+                                       switch (previousToken(position - 1, bound)) {
+                                               case TokenIF:
+                                               case TokenFOR:
+                                               case TokenWHILE:
+                                                       return true;
+                                       }
+                               }
+               }
+               
+               return false;
+       }
+}