Prepared better HEREDOC support; see comment for bug #1319276
[phpeclipse.git] / net.sourceforge.phpeclipse / src / net / sourceforge / phpeclipse / phpeditor / php / PHPPartitionScanner.java
index 43970cd..9b6420e 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
-**********************************************************************/
+ Copyright (c) 2002  Widespace, OU  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://solareclipse.sourceforge.net/legal/cpl-v10.html
+
+ Contributors:
+ Igor Malinin - initial contribution
+
+ $Id: PHPPartitionScanner.java,v 1.32 2005-10-09 12:35:01 axelcl Exp $
+ **********************************************************************/
 package net.sourceforge.phpeclipse.phpeditor.php;
 
-import java.io.CharArrayWriter;
 import java.util.ArrayList;
-import java.util.List;
+import java.util.HashMap;
+import java.util.Map;
 
+import net.sourceforge.phpdt.internal.compiler.parser.Scanner;
+import net.sourceforge.phpeclipse.ui.text.rules.AbstractPartitioner;
+
+import org.eclipse.jface.text.Assert;
+import org.eclipse.jface.text.BadLocationException;
+import org.eclipse.jface.text.IDocument;
 import org.eclipse.jface.text.rules.ICharacterScanner;
-import org.eclipse.jface.text.rules.IPredicateRule;
+import org.eclipse.jface.text.rules.IPartitionTokenScanner;
 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;
 
 /**
- * This scanner recognizes the JavaDoc comments and Java multi line comments.
+ *
+ *
+ * @author Igor Malinin
  */
-public class PHPPartitionScanner extends RuleBasedPartitionScanner {
-
-  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 {
-
-    public PHPMultiLineRule(String startSequence, String endSequence, IToken token) {
-      super(startSequence, endSequence, token);
-    }
-
-    public PHPMultiLineRule(String startSequence, String endSequence, IToken token, char escapeCharacter) {
-      super(startSequence, endSequence, token, escapeCharacter);
-    }
-
-    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;
-        }
-
-        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;
-          }
-        }
-      }
-      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 == '/');
-    }
-
-    /* (non-Javadoc)
-    * Method declared on IWordDetector
-       */
-    public boolean isWordPart(char c) {
-      return (c == '*' || c == '/');
-    }
-  };
-
-  /**
-   * 
-   */
-  static class WordPredicateRule extends WordRule implements IPredicateRule {
-
-    private IToken fSuccessToken;
-
-    public WordPredicateRule(IToken successToken) {
-      super(new EmptyCommentDetector());
-      fSuccessToken = successToken;
-      addWord("/**/", fSuccessToken);
-    }
-
-    /*
-     * @see org.eclipse.jface.text.rules.IPredicateRule#evaluate(ICharacterScanner, boolean)
-     */
-    public IToken evaluate(ICharacterScanner scanner, boolean resume) {
-      return super.evaluate(scanner);
-    }
-
-    /*
-     * @see org.eclipse.jface.text.rules.IPredicateRule#getSuccessToken()
-     */
-    public IToken getSuccessToken() {
-      return fSuccessToken;
-    }
-  };
-
-  /**
-   * 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);
-  }
-}
+public class PHPPartitionScanner implements IPartitionTokenScanner {
+       public static final String PHP_SCRIPTING_AREA = "__php_scripting_area ";
+
+       public static final int STATE_DEFAULT = 0;
+
+       // public static final int STATE_TAG = 1;
+       // public static final int STATE_SCRIPT = 2;
+
+       private IDocument document;
+
+       // private int begin;
+
+       private int end;
+
+       private int offset;
+
+       private int length;
+
+       private int position;
+
+       // private int state;
+
+       private Map tokens = new HashMap();
+
+       public PHPPartitionScanner() {
+       }
+
+       /*
+        * @see org.eclipse.jface.text.rules.ITokenScanner#nextToken()
+        */
+       public IToken nextToken() {
+               offset += length;
+
+               /*
+                * switch (state) { case STATE_TAG: return nextTagToken(); }
+                */
+
+               switch (read()) {
+               case ICharacterScanner.EOF:
+                       // state = STATE_DEFAULT;
+                       return getToken(null);
+
+               case '<':
+                       switch (read()) {
+                       case ICharacterScanner.EOF:
+                               // state = STATE_DEFAULT;
+                               return getToken(null);
+
+                       case '?': // <?
+                               // int ch = read();
+                               //
+                               // switch (ch) {
+                               // case ICharacterScanner.EOF:
+                               // state = STATE_DEFAULT;
+                               // return getToken(PHP_SCRIPTING_AREA);
+                               // }
+                               return scanUntilPHPEndToken(PHP_SCRIPTING_AREA);
+                       }
+
+                       unread();
+               }
+
+               loop: while (true) {
+                       switch (read()) {
+                       case ICharacterScanner.EOF:
+                               // state = STATE_DEFAULT;
+                               return getToken(null);
+
+                       case '<':
+                               switch (read()) {
+                               case ICharacterScanner.EOF:
+                                       // state = STATE_DEFAULT;
+                                       return getToken(null);
+
+                               case '?':
+                                       unread();
+                                       break;
+
+                               case '<':
+                                       unread();
+
+                               default:
+                                       continue loop;
+                               }
+
+                               unread();
+
+                               // state = STATE_DEFAULT;
+                               return getToken(null);
+                       }
+               }
+       }
+
+       private IToken scanUntilPHPEndToken(String token) {
+               int ch = read();
+               while (true) {
+                       switch (ch) {
+                       case ICharacterScanner.EOF:
+                               // state = STATE_DEFAULT;
+                               return getToken(token);
+                       case '"': // double quoted string
+                               // read until end of double quoted string
+                               if (!readUntilEscapedDQ()) {
+                                       // state = STATE_DEFAULT;
+                                       return getToken(token);
+                               }
+                               break;
+                       case '<': // heredoc string
+                               ch = read();
+                               switch (ch) {
+                               case ICharacterScanner.EOF:
+                                       break;
+                               case '<':
+                                       ch = read();
+                                       switch (ch) {
+                                       case ICharacterScanner.EOF:
+                                               break;
+                                       case '<':
+                                               // read until end of heredoc string
+                                               if (!readUntilEscapedHEREDOC()) {
+                                                       // state = STATE_DEFAULT;
+                                                       return getToken(token);
+                                               }
+                                       }
+                               }
+                               break;
+                       case '\'': // single quoted string
+                               // read until end of single quoted string
+                               if (!readUntilEscapedSQ()) {
+                                       // state = STATE_DEFAULT;
+                                       return getToken(token);
+                               }
+                               break;
+                       case '/': // comment start?
+                               ch = read();
+                               switch (ch) {
+                               case ICharacterScanner.EOF:
+                                       break;
+                               case '/':
+                                       // read until end of line
+                                       if (!readSingleLine()) {
+                                               // state = STATE_DEFAULT;
+                                               return getToken(token);
+                                       }
+                                       break;
+                               case '*':
+                                       // read until end of comment
+                                       if (!readMultiLineComment()) {
+                                               // state = STATE_DEFAULT;
+                                               return getToken(token);
+                                       }
+                                       break;
+                               default:
+                                       continue;
+                               }
+                               break;
+                       case '#': // line comment
+                               // read until end of line
+                               if (!readSingleLine()) {
+                                       // state = STATE_DEFAULT;
+                                       return getToken(token);
+                               }
+                               break;
+                       case '?':
+                               ch = read();
+                               switch (ch) {
+                               case ICharacterScanner.EOF:
+                               case '>':
+                                       // state = STATE_DEFAULT;
+                                       return getToken(token);
+
+                               case '?':
+                                       continue;
+                               default:
+                                       continue;
+                               }
+                       }
+
+                       ch = read();
+               }
+       }
+
+       private IToken getToken(String type) {
+               length = position - offset;
+
+               if (length == 0) {
+                       return Token.EOF;
+               }
+
+               // if (length<0) {
+               // try {
+               // System.out.println("Length<0:"+document.get(offset,5)+""+length);
+               // } catch (BadLocationException e) {
+               // e.printStackTrace();
+               // }
+               // }
+
+               if (type == null) {
+                       return Token.UNDEFINED;
+               }
+
+               IToken token = (IToken) tokens.get(type);
+               if (token == null) {
+                       token = new Token(type);
+                       tokens.put(type, token);
+               }
+
+               return token;
+       }
+
+       private int read() {
+               if (position >= end) {
+                       return ICharacterScanner.EOF;
+               }
+
+               try {
+                       return document.getChar(position++);
+               } catch (BadLocationException e) {
+                       --position;
+                       return ICharacterScanner.EOF;
+               }
+       }
+
+       private boolean readUntilEscapedDQ() {
+               // search last double quoted character
+               try {
+                       char ch;
+                       while (true) {
+                               if (position >= end) {
+                                       return false;
+                               }
+                               ch = document.getChar(position++);
+                               if (ch == '\\') {
+                                       if (position >= end) {
+                                               return false;
+                                       }
+                                       ch = document.getChar(position++); // ignore escaped character
+                               } else if (ch == '"') {
+                                       return true;
+                               }
+                       }
+               } catch (BadLocationException e) {
+                       --position;
+               }
+               return false;
+       }
+
+       private boolean readUntilEscapedSQ() {
+               // search last single quoted character
+               try {
+                       char ch;
+                       while (true) {
+                               if (position >= end) {
+                                       return false;
+                               }
+                               ch = document.getChar(position++);
+                               if (ch == '\\') {
+                                       if (position >= end) {
+                                               return false;
+                                       }
+                                       ch = document.getChar(position++); // ignore escaped character
+                               } else if (ch == '\'') {
+                                       return true;
+                               }
+                       }
+               } catch (BadLocationException e) {
+                       --position;
+               }
+               return false;
+       }
+
+       private boolean readUntilEscapedHEREDOC() {
+               // search until heredoc ends
+               try {
+                       char ch;
+                       StringBuffer buf = new StringBuffer();
+                       char[] heredocIdent;
+                       if (position >= end) {
+                               return false;
+                       }
+                       ch = document.getChar(position++);
+                       if (!Scanner.isPHPIdentifierStart(ch)) {
+                               return false;
+                       }
+                       while (Scanner.isPHPIdentifierPart(ch)) {
+                               buf.append(ch);
+                               if (position >= end) {
+                                       return false;
+                               }
+                               ch = document.getChar(position++);
+                       }
+                       heredocIdent = buf.toString().toCharArray();
+                       while (true) {
+                               if (position >= end) {
+                                       return false;
+                               }
+                               ch = document.getChar(position++);
+                               if (ch == '\n') { // heredoc could end after a newline
+                                       int pos = 0;
+                                       while (true) {
+                                               if (position >= end) {
+                                                       return false;
+                                               }
+                                               if (pos == heredocIdent.length) {
+                                                       return true;
+                                               }
+                                               ch = document.getChar(position++); // ignore escaped character
+                                               if (ch != heredocIdent[pos]) {
+                                                       break;
+                                               }
+                                               pos++;
+                                       }
+                               }
+                       }
+               } catch (BadLocationException e) {
+                       --position;
+               }
+               return false;
+       }
+
+       private boolean readSingleLine() {
+               try {
+                       do {
+                               if (position >= end) {
+                                       return false;
+                               }
+                       } while (document.getChar(position++) != '\n');
+                       return true;
+               } catch (BadLocationException e) {
+                       --position;
+               }
+               return false;
+       }
+
+       private boolean readMultiLineComment() {
+               try {
+                       char ch;
+                       while (true) {
+                               if (position >= end) {
+                                       return false;
+                               }
+                               ch = document.getChar(position++);
+                               if (ch == '*') {
+                                       if (position >= end) {
+                                               return false;
+                                       }
+                                       if (document.getChar(position) == '/') {
+                                               position++;
+                                               return true;
+                                       }
+                               }
+                       }
+               } catch (BadLocationException e) {
+                       --position;
+               }
+               return false;
+       }
+
+       private void unread() {
+               --position;
+       }
+
+       /*
+        * @see org.eclipse.jface.text.rules.ITokenScanner#getTokenOffset()
+        */
+       public int getTokenOffset() {
+               if (AbstractPartitioner.DEBUG) {
+                       Assert.isTrue(offset >= 0, Integer.toString(offset));
+               }
+               return offset;
+       }
+
+       /*
+        * @see org.eclipse.jface.text.rules.ITokenScanner#getTokenLength()
+        */
+       public int getTokenLength() {
+               return length;
+       }
+
+       /*
+        * @see org.eclipse.jface.text.rules.ITokenScanner#setRange(IDocument, int,
+        *      int)
+        */
+       public void setRange(IDocument document, int offset, int length) {
+               this.document = document;
+               // this.begin = offset;
+               this.end = offset + length;
+
+               this.offset = offset;
+               this.position = offset;
+               this.length = 0;
+       }
+
+       /*
+        * @see org.eclipse.jface.text.rules.IPartitionTokenScanner
+        */
+       public void setPartialRange(IDocument document, int offset, int length, String contentType, int partitionOffset) {
+               // state = STATE_DEFAULT;
+               if (partitionOffset > -1) {
+                       int delta = offset - partitionOffset;
+                       if (delta > 0) {
+                               setRange(document, partitionOffset, length + delta);
+                               return;
+                       }
+               }
+               setRange(document, partitionOffset, length);
+       }
+
+       // private boolean isContinuationPartition(IDocument document, int offset) {
+       // try {
+       // String type = document.getContentType(offset - 1);
+       //
+       // if (type != IDocument.DEFAULT_CONTENT_TYPE) {
+       // return true;
+       // }
+       // } catch (BadLocationException e) {}
+       //
+       // return false;
+       // }
+}
\ No newline at end of file