Fixed php in html-comment
[phpeclipse.git] / net.sourceforge.phpeclipse / src / net / sourceforge / phpeclipse / phpeditor / php / PHPPartitionScanner.java
index 43970cd..54eaa7d 100644 (file)
-/**********************************************************************
-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: 1.15 $
+ */
 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 = false;
+  private IDocument fDocument = null;
+  private int fOffset = -1;
+  private String fContentType = IPHPPartitionScannerConstants.HTML;
+  private String fPrevContentType;
 
-  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 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 = fOffset - fTokenOffset;
+    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="
+          + fContentType
+          + " fLength="
+          + fLength);
     }
+    return token;
+  }
 
-    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;
-        }
-        if (lineCommentMode || multiLineCommentMode || stringMode) {
-          continue;
+/* (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)
+{
+    this.setRange(document, offset, length);
+    if (DEBUG)
+    {
+        System.out.println(
+            "PartialRange: contentType="
+                + contentType
+                + " partitionOffset="
+                + partitionOffset);
+    }
+       
+       if(offset == partitionOffset)
+       {
+               try
+        {
+            fContentType = fDocument.getContentType(offset);
         }
-
-        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;
-          }
+        catch (BadLocationException e)
+        {
+            //should never happen
         }
-      }
-      scanner.unread();
-      return true;
+       }
+    else if (this.tokens.containsKey(contentType))
+        fContentType = contentType;
+    // TODO Calculate previouse contenttype
+    if (partitionOffset > -1)
+    {
+        partitionBorder = false;
+        fTokenOffset = partitionOffset;
     }
+}
+
+  /* (non-Javadoc)
+   * @see org.eclipse.jface.text.rules.ITokenScanner#getTokenLength()
+   */
+  public int getTokenLength() {
+    return fLength;
   }
-  /**
-   * Detector for empty comments.
+
+  /* (non-Javadoc)
+   * @see org.eclipse.jface.text.rules.ITokenScanner#getTokenOffset()
    */
-  static class EmptyCommentDetector implements IWordDetector {
+  public int getTokenOffset() {
+    return fTokenOffset;
+  }
 
-    /* (non-Javadoc)
-    * Method declared on IWordDetector
-       */
-    public boolean isWordStart(char c) {
-      return (c == '/');
+/* (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();
 
-    /* (non-Javadoc)
-    * Method declared on IWordDetector
-       */
-    public boolean isWordPart(char c) {
-      return (c == '*' || c == '/');
+    if (partitionBorder)
+    {
+        fTokenOffset = fOffset;
+        partitionBorder = false;
     }
-  };
 
-  /**
-   * 
-   */
-  static class WordPredicateRule extends WordRule implements IPredicateRule {
+    while ((c = read()) != ICharacterScanner.EOF)
+    {
+        switch (c)
+        {
+            case '<' :
+                if (fContentType
+                    != IPHPPartitionScannerConstants.PHP_MULTILINE_COMMENT
+                    && checkPattern(new char[] { '?', 'p', 'h', 'p' }, true))
+                {
+                    if (fContentType != IPHPPartitionScannerConstants.PHP
+                        && fOffset - 5 > 0)
+                    {
+                        fOffset -= 5;
+                        IToken token = getToken(fContentType);
+                        // save previouse contenttype
+                        fPrevContentType = fContentType;
 
-    private IToken fSuccessToken;
+                        fContentType = IPHPPartitionScannerConstants.PHP;
 
-    public WordPredicateRule(IToken successToken) {
-      super(new EmptyCommentDetector());
-      fSuccessToken = successToken;
-      addWord("/**/", fSuccessToken);
-    }
+                        return token;
+                    }
+                    else
+                        fContentType = IPHPPartitionScannerConstants.PHP;
+
+                    // remember offset of this partition
+                    fTokenOffset = fOffset - 5;
+                }
+                else if (
+                    fContentType
+                        != IPHPPartitionScannerConstants.PHP_MULTILINE_COMMENT
+                    && checkPattern(new char[] { '?' }, false))
+                {
+                    if (fContentType != IPHPPartitionScannerConstants.PHP
+                        && fOffset - 2 > 0)
+                    {
+                        fOffset -= 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;
+                            }
+                else if (checkPattern(new char[] { '!', '-', '-' }))
+                { // return previouse partition
+                if (
+                    fContentType
+                        != IPHPPartitionScannerConstants.HTML_MULTILINE_COMMENT
+                    && fOffset - 4 > 0)
+                    {
+                        fOffset -= 4;
+                            IToken token = getToken(fContentType);
+                            fContentType =
+                                IPHPPartitionScannerConstants
+                                    .HTML_MULTILINE_COMMENT;
+                            return token;
+                            }
+                    else
+                        fContentType =
+                            IPHPPartitionScannerConstants
+                                .HTML_MULTILINE_COMMENT;
+                            fTokenOffset = fOffset - 4;
+                            }
+                break; case '?' :
+                if (fContentType == IPHPPartitionScannerConstants.PHP)
+                {
+                    if ((c = read()) == '>')
+                    { // TODO Actually calculate the previouse contenttype from the document
+                    if (
+                        fPrevContentType
+                        != null)
+                            fContentType = fPrevContentType; else
+                                fContentType =
+                                    IPHPPartitionScannerConstants.HTML;
+                                    partitionBorder = true;
+                                    return getToken(
+                                        IPHPPartitionScannerConstants.PHP);
+                                    }
+                    else if (c != ICharacterScanner.EOF)
+                        unread(); }
+                break; case '-' :
+                if (fContentType
+                    == IPHPPartitionScannerConstants.HTML_MULTILINE_COMMENT
+                    && checkPattern(new char[] { '-', '>' }))
+                {
+                    fContentType = IPHPPartitionScannerConstants.HTML;
+                        partitionBorder = true;
+                        return getToken(
+                            IPHPPartitionScannerConstants
+                                .HTML_MULTILINE_COMMENT);
+                        }
+                break; case '/' :
+                if ((c = read()) == '*')
+                { // MULTINE COMMENT JAVASCRIPT, CSS, PHP
+                if (
+                    fContentType == IPHPPartitionScannerConstants.PHP
+                    && fOffset - 2 > 0)
+                    {
+                        fOffset -= 2;
+                            IToken token = getToken(fContentType);
+                            fContentType =
+                                IPHPPartitionScannerConstants
+                                    .PHP_MULTILINE_COMMENT;
+                            return token;
+                            }
+                    else if (
+                        fContentType
+                            == IPHPPartitionScannerConstants
+                                .PHP_MULTILINE_COMMENT)
+                    {
 
-    /*
-     * @see org.eclipse.jface.text.rules.IPredicateRule#evaluate(ICharacterScanner, boolean)
-     */
-    public IToken evaluate(ICharacterScanner scanner, boolean resume) {
-      return super.evaluate(scanner);
+                        fTokenOffset = fOffset - 2; }
+
+                }
+                else if (c != ICharacterScanner.EOF)
+                    unread(); break; case '*' :
+                        if ((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 (c != ICharacterScanner.EOF)
+                            unread(); 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);
     }
 
-    /*
-     * @see org.eclipse.jface.text.rules.IPredicateRule#getSuccessToken()
-     */
-    public IToken getSuccessToken() {
-      return fSuccessToken;
+    fDocument = document;
+    fOffset = offset;
+    fTokenOffset = offset;
+    fLength = 0;
+    fEnd = fOffset + length;
+    //partitionBorder = false;
+  }
+
+  private int read() {
+    try {
+      if (fOffset < fEnd) {
+        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;
+  }
+
+  private boolean checkPattern(char[] pattern) {
+    return checkPattern(pattern, false);
+  }
 
   /**
-   * Creates the partitioner and sets up the appropriate rules.
+   * 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 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);
+  private boolean checkPattern(char[] pattern, boolean ignoreCase) {
+    int prevOffset = fOffset;
+    for (int i = 0; i < pattern.length; i++) {
+      int c = read();
+
+      if (c == ICharacterScanner.EOF
+        || !letterEquals(c, pattern[i], ignoreCase)) {
+        fOffset = prevOffset;
+        return false;
+      }
+    }
+
+    return true;
   }
+
+  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;
+  }
+
 }