-/**********************************************************************
-Copyright (c) 2000, 2002 IBM Corp. 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 implementation
- Klaus Hartlage - www.eclipseproject.de
-**********************************************************************/
+/**
+ * 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
+ * Created on 05.03.2003
+ *
+ * @author Stefan Langer (musk)
+ * @version $Revision: 1.16 $
+ */
package net.sourceforge.phpeclipse.phpeditor.php;
-import java.io.CharArrayWriter;
-import java.util.ArrayList;
-import java.util.List;
+import java.util.*;
-import org.eclipse.jface.text.rules.ICharacterScanner;
-import org.eclipse.jface.text.rules.IPredicateRule;
-import org.eclipse.jface.text.rules.IToken;
-import org.eclipse.jface.text.rules.IWordDetector;
-import org.eclipse.jface.text.rules.MultiLineRule;
-import org.eclipse.jface.text.rules.RuleBasedPartitionScanner;
-import org.eclipse.jface.text.rules.Token;
-import org.eclipse.jface.text.rules.WordRule;
+import org.eclipse.jface.text.*;
+import org.eclipse.jface.text.rules.*;
/**
- * This scanner recognizes the JavaDoc comments and Java multi line comments.
+ *
*/
-public class PHPPartitionScanner extends RuleBasedPartitionScanner {
+public class PHPPartitionScanner implements IPartitionTokenScanner
+{
+ private static final boolean DEBUG = true;
+ private boolean fInString = false;
+ private boolean fInDoubString = false;
+ private IDocument fDocument = null;
+ private int fOffset = -1;
+ private String fContentType = IPHPPartitionScannerConstants.HTML;
+ private String fPrevContentType = IPHPPartitionScannerConstants.HTML;
- private final static String SKIP = "__skip"; //$NON-NLS-1$
- public final static String JAVA_MULTILINE_COMMENT = "__html_multiline_comment"; //$NON-NLS-1$
- // public final static String JAVA_DOC= "__java_javadoc"; //$NON-NLS-1$
- public final static String PHP = "__php";
- private StringBuffer test;
- public class PHPMultiLineRule extends MultiLineRule {
+ private boolean partitionBorder = false;
+ private int fTokenOffset;
+ private int fEnd = -1;
+ private int fLength;
+ private int fCurrentLength;
+ private Map tokens = new HashMap();
- public PHPMultiLineRule(String startSequence, String endSequence, IToken token) {
- super(startSequence, endSequence, token);
+ public PHPPartitionScanner()
+ {
+ this.tokens.put(
+ IPHPPartitionScannerConstants.PHP,
+ new Token(IPHPPartitionScannerConstants.PHP));
+ this.tokens.put(
+ IPHPPartitionScannerConstants.PHP_MULTILINE_COMMENT,
+ new Token(IPHPPartitionScannerConstants.PHP_MULTILINE_COMMENT));
+ this.tokens.put(
+ IPHPPartitionScannerConstants.HTML,
+ new Token(IPHPPartitionScannerConstants.HTML));
+ this.tokens.put(
+ IPHPPartitionScannerConstants.HTML_MULTILINE_COMMENT,
+ new Token(IPHPPartitionScannerConstants.HTML_MULTILINE_COMMENT));
+ this.tokens.put(
+ IDocument.DEFAULT_CONTENT_TYPE,
+ new Token(IDocument.DEFAULT_CONTENT_TYPE));
}
- public PHPMultiLineRule(String startSequence, String endSequence, IToken token, char escapeCharacter) {
- super(startSequence, endSequence, token, escapeCharacter);
- }
+ private IToken getToken(String type)
+ {
+ fLength = fCurrentLength;
+ if (DEBUG)
+ {
- protected boolean endSequenceDetected(ICharacterScanner scanner) {
- int c;
- int c2;
-
- boolean lineCommentMode = false;
- boolean multiLineCommentMode = false;
- boolean stringMode = false;
-
- char[][] delimiters = scanner.getLegalLineDelimiters();
- while ((c = scanner.read()) != ICharacterScanner.EOF) {
- if (lineCommentMode && (c == '\n')) {
- lineCommentMode = false;
- // read until end of line
- } else if ((!stringMode) && (c == '#')) {
- // read until end of line
- lineCommentMode = true;
- continue;
- } else if ((!stringMode) && (!multiLineCommentMode) && (c == '/')) {
- c2 = scanner.read();
- if (c2 == '/') {
- lineCommentMode = true;
- continue;
- } else if (c2 == '*') {
- multiLineCommentMode = true;
- continue;
- } else {
- scanner.unread();
- }
- } else if (c == '*' && multiLineCommentMode) {
- c2 = scanner.read();
- if (c2 == '/') {
- multiLineCommentMode = false;
- continue;
- } else {
- scanner.unread();
- }
- } else if (c == '\\' && stringMode) {
- c2 = scanner.read();
- if (c2 == '"') {
- continue;
- } else {
- scanner.unread();
- }
- } else if ((!lineCommentMode) && (!multiLineCommentMode) && (c == '"')) {
- if (stringMode) {
- stringMode = false;
- } else {
- stringMode = true;
- }
- continue;
+ try
+ {
+ if (fLength <= 0)
+ {
+ int line = fDocument.getLineOfOffset(fOffset);
+ System.err.println(
+ "Error at "
+ + line
+ + " offset:"
+ + String.valueOf(
+ fOffset - fDocument.getLineOffset(line)));
+ }
+ }
+ catch (BadLocationException e)
+ {
+ // TODO Auto-generated catch block
+ e.printStackTrace();
+ }
}
- if (lineCommentMode || multiLineCommentMode || stringMode) {
- continue;
+ Assert.isTrue(fLength > 0, "Partition length <= 0!");
+ fCurrentLength = 0;
+ // String can never cross partition borders so reset string detection
+ fInString = false;
+ fInDoubString = false;
+ IToken token = (IToken) this.tokens.get(type);
+ Assert.isNotNull(token, "Token for type \"" + type + "\" not found!");
+ if (DEBUG)
+ {
+ System.out.println(
+ "Partition: fTokenOffset="
+ + fTokenOffset
+ + " fContentType="
+ + type
+ + " fLength="
+ + fLength);
}
+ return token;
+ }
- if (c == fEscapeCharacter) {
- // Skip the escaped character.
- scanner.read();
- } else if (fEndSequence.length > 0 && c == fEndSequence[0]) {
- // Check if the specified end sequence has been found.
- if (sequenceDetected(scanner, fEndSequence, true))
- return true;
- } else if (fBreaksOnEOL) {
- // Check for end of line since it can be used to terminate the pattern.
- for (int i = 0; i < delimiters.length; i++) {
- if (c == delimiters[i][0] && sequenceDetected(scanner, delimiters[i], false))
- return true;
- }
+ /* (non-Javadoc)
+ * @see org.eclipse.jface.text.rules.IPartitionTokenScanner#setPartialRange(org.eclipse.jface.text.IDocument, int, int, java.lang.String, int)
+ */
+ public void setPartialRange(
+ IDocument document,
+ int offset,
+ int length,
+ String contentType,
+ int partitionOffset)
+ {
+ if (DEBUG)
+ {
+ System.out.println(
+ "PartialRange: contentType="
+ + contentType
+ + " partitionOffset="
+ + partitionOffset);
+ }
+
+ try
+ {
+ if (partitionOffset > -1)
+ {
+ partitionBorder = false;
+ // because of strings we have to parse the whole partition
+ this.setRange(
+ document,
+ partitionOffset,
+ offset - partitionOffset + length);
+ // sometimes we get a wrong partition so we retrieve the partition
+ // directly from the document
+ fContentType = fDocument.getContentType(partitionOffset);
+ //TODO determine the previouse contenttypes as a stack
+ //if(partitionOffset > 1)
+ // fPrevContentType = fDocument.getContentType(partitionOffset-1);
+ }
+ else
+ this.setRange(document, offset, length);
+
+ }
+ catch (BadLocationException e)
+ {
+ // should never happen
+ // TODO print stack trace to log
+ // fall back just scan the whole document again
+ this.setRange(document, 0, fDocument.getLength());
}
- }
- scanner.unread();
- return true;
+
}
- }
- /**
- * Detector for empty comments.
- */
- static class EmptyCommentDetector implements IWordDetector {
/* (non-Javadoc)
- * Method declared on IWordDetector
- */
- public boolean isWordStart(char c) {
- return (c == '/');
+ * @see org.eclipse.jface.text.rules.ITokenScanner#getTokenLength()
+ */
+ public int getTokenLength()
+ {
+ return fLength;
}
/* (non-Javadoc)
- * Method declared on IWordDetector
- */
- public boolean isWordPart(char c) {
- return (c == '*' || c == '/');
+ * @see org.eclipse.jface.text.rules.ITokenScanner#getTokenOffset()
+ */
+ public int getTokenOffset()
+ {
+ return fTokenOffset;
}
- };
- /**
- *
- */
- static class WordPredicateRule extends WordRule implements IPredicateRule {
+ /* (non-Javadoc)
+ * @see org.eclipse.jface.text.rules.ITokenScanner#nextToken()
+ */
+ public IToken nextToken()
+ {
+ int c;
+
+ // check if we are not allready at the end of the
+ // file
+ if ((c = read()) == ICharacterScanner.EOF)
+ {
+ partitionBorder = false;
+ return Token.EOF;
+ }
+ else
+ unread();
+
+ if (partitionBorder)
+ {
+ fTokenOffset = fOffset;
+ partitionBorder = false;
+ }
+
+ while ((c = read()) != ICharacterScanner.EOF)
+ {
+ switch (c)
+ {
+ case '<' :
+ if (!isInString(IPHPPartitionScannerConstants.PHP)
+ && fContentType
+ != IPHPPartitionScannerConstants.PHP_MULTILINE_COMMENT
+ && checkPattern(new char[] { '?', 'p', 'h', 'p' }, true))
+ {
+ if (fContentType != IPHPPartitionScannerConstants.PHP
+ && fCurrentLength > 5)
+ {
+ unread(5);
+ IToken token = getToken(fContentType);
+ // save previouse contenttype
+ //TODO build stack for previouse contenttype
+ fPrevContentType = fContentType;
+
+ fContentType = IPHPPartitionScannerConstants.PHP;
+
+ return token;
+ }
+ else
+ fContentType = IPHPPartitionScannerConstants.PHP;
+
+ // remember offset of this partition
+ fTokenOffset = fOffset - 5;
+ fCurrentLength = 5;
+ }
+ else if (
+ !isInString(IPHPPartitionScannerConstants.PHP)
+ && fContentType
+ != IPHPPartitionScannerConstants
+ .PHP_MULTILINE_COMMENT
+ && checkPattern(new char[] { '?' }, false))
+ {
+ if (fContentType != IPHPPartitionScannerConstants.PHP
+ && fCurrentLength > 2)
+ {
+ unread(2);
+ IToken token = getToken(fContentType);
+ // save previouse contenttype
+ fPrevContentType = fContentType;
+ fContentType = IPHPPartitionScannerConstants.PHP;
+ return token;
+ }
+ else
+ fContentType = IPHPPartitionScannerConstants.PHP;
+ // remember offset of this partition
+ fTokenOffset = fOffset - 2;
+ fCurrentLength = 2;
+ }
+ else if (
+ !isInString(IPHPPartitionScannerConstants.PHP)
+ && checkPattern(new char[] { '!', '-', '-' }))
+ { // return previouse partition
+ if (fContentType
+ != IPHPPartitionScannerConstants
+ .HTML_MULTILINE_COMMENT
+ && fCurrentLength > 4)
+ {
+ unread(4);
+ IToken token = getToken(fContentType);
+ fContentType =
+ IPHPPartitionScannerConstants
+ .HTML_MULTILINE_COMMENT;
+ return token;
+ }
+ else
+ fContentType =
+ IPHPPartitionScannerConstants
+ .HTML_MULTILINE_COMMENT;
+
+ fTokenOffset = fOffset - 4;
+ fCurrentLength = 4;
+ }
+ break;
+ case '?' :
+ if (!isInString(IPHPPartitionScannerConstants.PHP)
+ && fContentType == IPHPPartitionScannerConstants.PHP)
+ {
+ if ((c = read()) == '>')
+ {
+ if (fPrevContentType != null)
+ fContentType = fPrevContentType;
+ else
+ fContentType =
+ IPHPPartitionScannerConstants.HTML;
+ partitionBorder = true;
+ return getToken(IPHPPartitionScannerConstants.PHP);
+ }
+ else if (c != ICharacterScanner.EOF)
+ unread();
+ }
+ break;
+ case '-' :
+ if (!isInString(IPHPPartitionScannerConstants.PHP)
+ && fContentType
+ == IPHPPartitionScannerConstants
+ .HTML_MULTILINE_COMMENT
+ && checkPattern(new char[] { '-', '>' }))
+ {
+ fContentType = IPHPPartitionScannerConstants.HTML;
+ partitionBorder = true;
+ return getToken(
+ IPHPPartitionScannerConstants
+ .HTML_MULTILINE_COMMENT);
+ }
+ break;
+ case '/' :
+ if (!isInString(IPHPPartitionScannerConstants.PHP) && (c = read()) == '*')
+ { // MULTINE COMMENT JAVASCRIPT, CSS, PHP
+ if (fContentType == IPHPPartitionScannerConstants.PHP
+ && fCurrentLength > 2)
+ {
+ unread(2);
+ IToken token = getToken(fContentType);
+ fContentType =
+ IPHPPartitionScannerConstants
+ .PHP_MULTILINE_COMMENT;
+ return token;
+ }
+ else if (
+ fContentType
+ == IPHPPartitionScannerConstants
+ .PHP_MULTILINE_COMMENT)
+ {
- private IToken fSuccessToken;
+ fTokenOffset = fOffset - 2;
+ fCurrentLength = 2;
+ }
- public WordPredicateRule(IToken successToken) {
- super(new EmptyCommentDetector());
- fSuccessToken = successToken;
- addWord("/**/", fSuccessToken);
+ }
+ else if (!isInString(IPHPPartitionScannerConstants.PHP) && c != ICharacterScanner.EOF)
+ unread();
+ break;
+ case '*' :
+ if (!isInString(IPHPPartitionScannerConstants.PHP) && (c = read()) == '/')
+ {
+ if (fContentType
+ == IPHPPartitionScannerConstants
+ .PHP_MULTILINE_COMMENT)
+ {
+ fContentType = IPHPPartitionScannerConstants.PHP;
+ partitionBorder = true;
+ return getToken(
+ IPHPPartitionScannerConstants
+ .PHP_MULTILINE_COMMENT);
+ }
+ else if (
+ fContentType
+ == IPHPPartitionScannerConstants
+ .CSS_MULTILINE_COMMENT)
+ {
+ }
+ else if (
+ fContentType
+ == IPHPPartitionScannerConstants
+ .JS_MULTILINE_COMMENT)
+ {
+ }
+ }
+ else if (!isInString(IPHPPartitionScannerConstants.PHP) && c != ICharacterScanner.EOF)
+ unread();
+ break;
+ case '\'' :
+ if (!fInDoubString)
+ fInString = !fInString;
+ break;
+ case '"' :
+ // toggle String mode
+ if (!fInString)
+ fInDoubString = !fInDoubString;
+ break;
+ }
+ } // end of file reached but we have to return the
+ // last partition.
+ return getToken(fContentType);
+ }
+ /* (non-Javadoc)
+ * @see org.eclipse.jface.text.rules.ITokenScanner#setRange(org.eclipse.jface.text.IDocument, int, int)
+ */
+ public void setRange(IDocument document, int offset, int length)
+ {
+ if (DEBUG)
+ {
+ System.out.println(
+ "SET RANGE: offset=" + offset + " length=" + length);
+ }
+
+ fDocument = document;
+ fOffset = offset;
+ fTokenOffset = offset;
+ fCurrentLength = 0;
+ fLength = 0;
+ fEnd = fOffset + length;
+ fInString = false;
+ fInDoubString = false;
+ //partitionBorder = false;
+ }
+
+ private int read()
+ {
+ try
+ {
+ if (fOffset < fEnd)
+ {
+ fCurrentLength++;
+ return fDocument.getChar(fOffset++);
+ }
+ return ICharacterScanner.EOF;
+ }
+ catch (BadLocationException e)
+ {
+ // should never happen
+ // TODO write stacktrace to log
+ fOffset = fEnd;
+ return ICharacterScanner.EOF;
+ }
+ }
+
+ private void unread()
+ {
+ --fOffset;
+ --fCurrentLength;
+ }
+
+ private void unread(int num)
+ {
+ fOffset -= num;
+ fCurrentLength -= num;
}
- /*
- * @see org.eclipse.jface.text.rules.IPredicateRule#evaluate(ICharacterScanner, boolean)
+ private boolean checkPattern(char[] pattern)
+ {
+ return checkPattern(pattern, false);
+ }
+
+ /**
+ * Check if next character sequence read from document is equals to
+ * the provided pattern. Pattern is read from left to right until the
+ * first character read doesn't match. If this happens all read characters are
+ * unread.
+ * @param pattern The pattern to check.
+ * @return <code>true</code> if pattern is equals else returns <code>false</code>.
*/
- public IToken evaluate(ICharacterScanner scanner, boolean resume) {
- return super.evaluate(scanner);
+ private boolean checkPattern(char[] pattern, boolean ignoreCase)
+ {
+ int prevOffset = fOffset;
+ int prevLength = fCurrentLength;
+ for (int i = 0; i < pattern.length; i++)
+ {
+ int c = read();
+
+ if (c == ICharacterScanner.EOF
+ || !letterEquals(c, pattern[i], ignoreCase))
+ {
+ fOffset = prevOffset;
+ fCurrentLength = prevLength;
+ return false;
+ }
+ }
+
+ return true;
}
- /*
- * @see org.eclipse.jface.text.rules.IPredicateRule#getSuccessToken()
+ private boolean letterEquals(int test, char letter, boolean ignoreCase)
+ {
+ if (test == letter)
+ return true;
+ else if (
+ ignoreCase
+ && Character.isLowerCase(letter)
+ && test == Character.toUpperCase(letter))
+ return true;
+ else if (
+ ignoreCase
+ && Character.isUpperCase(letter)
+ && test == Character.toLowerCase(letter))
+ return true;
+
+ return false;
+ }
+
+ /**
+ * Checks wether the offset is in a <code>String</code> and the specified
+ * contenttype is the current content type.
+ * Strings are delimited, mutual exclusive, by a " or by a '.
+ *
+ * @param contentType The contenttype to check.
+ * @return <code>true</code> if the current offset is in a string else
+ * returns false.
*/
- public IToken getSuccessToken() {
- return fSuccessToken;
+ private boolean isInString(String contentType)
+ {
+ if(fContentType == contentType)
+ return (fInString || fInDoubString);
+ else
+ return false;
}
- };
-
- /**
- * Creates the partitioner and sets up the appropriate rules.
- */
- public PHPPartitionScanner() {
- super();
-
- // IToken javaDoc= new Token(JAVA_DOC);
- IToken comment = new Token(JAVA_MULTILINE_COMMENT);
- IToken php = new Token(PHP);
-
- List rules = new ArrayList();
-
- // Add rule for single line comments.
- // rules.add(new EndOfLineRule("//", Token.UNDEFINED));
-
- // Add rule for strings and character constants.
- // rules.add(new SingleLineRule("\"", "\"", Token.UNDEFINED, '\\'));
- // rules.add(new SingleLineRule("'", "'", Token.UNDEFINED, '\\'));
-
- // Add special case word rule.
- rules.add(new WordPredicateRule(comment));
-
- // Add rules for multi-line comments and javadoc.
- //rules.add(new MultiLineRule("/**", "*/", javaDoc));
- rules.add(new MultiLineRule("<!--", "-->", comment));
- rules.add(new PHPMultiLineRule("<?\r", "?>", php));
- rules.add(new PHPMultiLineRule("<?\n", "?>", php));
- rules.add(new PHPMultiLineRule("<?\t", "?>", php));
- rules.add(new PHPMultiLineRule("<? ", "?>", php));
- rules.add(new PHPMultiLineRule("<?php", "?>", php));
- rules.add(new PHPMultiLineRule("<?PHP", "?>", php));
- //Add rule for processing instructions
-
- IPredicateRule[] result = new IPredicateRule[rules.size()];
- rules.toArray(result);
- setPredicateRules(result);
- }
+
}
git://secure.phpeclipse.com / phpeclipse.git / blobdiff