Changeset 18833 for gli

Timestamp:

2009-03-30T12:58:55+13:00 (15 years ago)

Author:

kjdon

Message:

bug fix 18462 committed to branch - No longer expects ?xml to be the only acceptable xml opening marker, but also accepts DOCTYPE and any rootelement.

File:

• gli/branches/2.81-fixed/src/org/greenstone/gatherer/util/RemoveContentBeforeRootElementXMLReader.java (modified) (1 diff)

gli/branches/2.81-fixed/src/org/greenstone/gatherer/util/RemoveContentBeforeRootElementXMLReader.java

@@ -4 +4 @@
 import java.io.IOException;
 
+import java.io.File;
+import java.io.FileReader;
+
+import java.util.regex.Pattern;
+
 public class RemoveContentBeforeRootElementXMLReader extends Reader {
 
-    static final byte[] xmlIndicator = "<?xml".getBytes();
+  static final Pattern[] xmlIndicators;
+  static Pattern commentStart;
+  static Pattern commentStop;
 
-    Reader ur;
-    int foundBytesReturned = 0;
-    boolean allFoundBytesReturned = false; //redundant, but may help performance
+  static {
+    // generate the xml starting sequences we will try to match against
+    xmlIndicators = new Pattern[3];
+    try{
+      xmlIndicators[0] = Pattern.compile("<\\?xml"); //usual xml declaration
+      xmlIndicators[1] = Pattern.compile("<!DOCTYPE"); //doctype declaration
+      xmlIndicators[2] = Pattern.compile("<[a-zA-Z-]+[ >/]"); // the beginning of a root node
+
+      commentStart = Pattern.compile("<!--");
+      commentStop = Pattern.compile("-->");
+
+    } catch ( java.util.regex.PatternSyntaxException pse ) {
+        System.err.println( "Pattern no good. " + pse );
+        xmlIndicators[0] = xmlIndicators[1] = xmlIndicators[2] = null;
+        commentStart = commentStop = null;
+    }
+  }
+
+  Reader ur;
+
+  String finalBuffer = null;
+  int finalBufferIndex = 0;
+
+  public RemoveContentBeforeRootElementXMLReader( Reader ur ) {
+
+    this.ur = ur;
+
+    int found = -1;
+    boolean inComment = false;
+    StringBuffer buffer = null;
+
+    for ( int c = 0;  c != -1 && found == -1; ) {
+
+      //read a character
+      try {
+        c = ur.read();
+      } catch( Exception e ) {
+        System.err.println( "Exception while reading underlying Reader in RemoveContentBeforeRootElementXMLReader" );
+      }
+
+      //break out if we have reached the end of the input
+      if ( c == -1 ) {
+        break;
+      }
+
+      //we start buffering when we come across the first <
+      //regardless of whether it turns out to be a relevant < or not
+      if ( buffer == null && (char)c == '<' )  {
+        buffer = new StringBuffer();
+      }
+
+      //if not buffering, just display the character and move onto next character
+      if ( buffer == null ) {
+         System.err.print( (char)c );
+        continue;
+      }
+
+      buffer.append( (char)c );
+
+      //check for comment open or close
+      if ( !inComment ) {
+        if ( commentStart.matcher(buffer.toString()).find() ) {
+          inComment = true;
+          System.err.print( buffer.toString() );
+          buffer = new StringBuffer();
+        }
+      } else {
+        if ( commentStop.matcher(buffer.toString()).find() ) {
+          inComment = false;
+          System.err.print( buffer.toString() );
+          buffer = new StringBuffer();
+
+          //skip to reading next character
+          continue;
+        }
+      }
+
+      if ( !inComment ) {
+
+        //check each indicator to see if found
+        for ( int i = 0; i < xmlIndicators.length && found == -1; i++ ) {
+          if ( xmlIndicators[i].matcher(buffer.toString()).find() ) {
+            found = i;
+            String line = buffer.toString();
+            int lastIndex = line.lastIndexOf('<');
+            //flush the previous characters in the buffer to the console
+            System.err.print(line.substring(0, lastIndex));
+            buffer.delete(0, lastIndex);
+            finalBuffer = buffer.toString();
+          }
+        }
+
+      }
+    }
+
+    if ( found == -1 ) {
+      System.err.println( "RemoveContentBeforeRootElementXMLReader:\n" +
+        "The XML being loaded was not valid: couldn't find start of XML input" );
+    }
+  
+  }
+
+  public int read( char[] cbuf, int off, int len ) throws IOException {
+
+    for ( int i=off; i<off+len && i<cbuf.length; i++ ) {
+
+      //read from underlying reader
+      int c = read();
+
+      //catch end of stream
+      if ( c == -1 ) {
+        if ( i == off ) {
+          return -1;
+        }
+        return i - off;
+      }
+
+      //insert character into the array
+      cbuf[i] = (char)c;
+    }
+    return len;
+
+  }
+
+  public int read() throws IOException {
+
+    //flush the buffer containing the opening XML sequence
+    if ( finalBuffer != null && finalBuffer.length() > finalBufferIndex ) {
+      char c = finalBuffer.charAt(finalBufferIndex++);
+      if ( finalBufferIndex == finalBuffer.length() ) {
+        finalBuffer = null;
+      }
+      return c;
+    }
+
+    return ur.read();
+  }
+
+  public void close() throws IOException {
+    ur.close();
+  }
+
+  public static void main ( String[] args ) {
+
+    //init
+    System.out.println( "------------\nWill now initialise the test reader\n------------" );
+    RemoveContentBeforeRootElementXMLReader parser = null;
+    try {
+      parser = new RemoveContentBeforeRootElementXMLReader( 
+        new FileReader( new File("text.xml") ) );
+    } catch ( java.io.FileNotFoundException fnfe ) {
+      System.err.println( "Please create text.xml to test this class" );
+      System.exit(-1);
+    }
+
+    //read the rest of the input
+    System.out.println( "------------\nWill now read the rest of the input\n------------" );
+    try {
+      int c = 0;
+      while ( ( c = parser.read() ) != -1 ) {
+        System.out.print( (char)c );
+      }
+    } catch ( Exception e ) {
+      System.err.println("Exception: " + e);
+    }
+    
+  }
 
 
-    public RemoveContentBeforeRootElementXMLReader( Reader ur ) {
-
-        this.ur = ur;
-
-        //read up to the xml indicator
-        int foundBytes = 0;
-        int c = 0;
-        while ( c != -1 ) {
-            try {
-                c = ur.read();
-            } catch( Exception e ) {
-                System.err.println( "Exception while reading underlying Reader in RemoveContentBeforeRootElementXMLReader" );
-            }
-            
-            if ( c == (int)xmlIndicator[foundBytes] ) {
-                foundBytes++;
-            } else {                
-                if ( foundBytes != 0 ) {
-                for ( int i=0; i<foundBytes; i++ ) {
-                    System.out.print( (char)xmlIndicator[i] );
-                }
-                foundBytes = 0;
-                }
-                if ( c != -1 ) {
-                System.out.print( (char)c );
-                }
-            }
-
-            if ( foundBytes == xmlIndicator.length ) {
-                return;
-            }
-
-        }
-        
-        System.err.println( "RemoveContentBeforeRootElementXMLReader:\n" +
-            "The XML being loaded did not contain the '<?xml' string as expected" );
-    }
-
-    public int read( char[] cbuf, int off, int len ) throws IOException {
-
-        for ( int i=off; i<off+len && i<cbuf.length; i++ ) {
-
-            //read from underlying reader
-            int c = read();
-
-            //catch end of stream
-            if ( c == -1 ) {
-                if ( i == off ) {
-                    return -1;
-                }
-                return i - off;
-            }
-
-            //insert character into the array
-            cbuf[i] = (char)c;
-        }
-        return len;
-
-    }
-
-    public int read() throws IOException {
-
-        if ( allFoundBytesReturned ) {
-            return ur.read();
-        }
-
-        if ( foundBytesReturned < xmlIndicator.length ) {
-            return xmlIndicator[foundBytesReturned++];
-        }
-
-        allFoundBytesReturned = true;
-        return ur.read();
-    }
-
-    public void close() throws IOException {
-        ur.close();
-    }
 }