Changeset 16637 for gli

Timestamp:

2008-08-04T11:43:10+12:00 (16 years ago)

Author:

ak19

Message:

toString() method tries to display the filename in UTF8 if it is UTF8 encoded, otherwise it tries converting the filename to some other common encodings and failing those, returns the original filename (as the string is stored in the OS) for display

File:

• gli/trunk/src/org/greenstone/gatherer/collection/CollectionTreeNode.java (modified) (4 diffs)

gli/trunk/src/org/greenstone/gatherer/collection/CollectionTreeNode.java

@@ -33 +33 @@
 import org.greenstone.gatherer.file.FileNode;
 import org.greenstone.gatherer.util.JarTools;
+import java.util.Set;
+import java.util.Iterator;
+import java.nio.charset.Charset;
 
 
@@ -48 +51 @@
     private boolean is_srcreplaceable = false;
 
+    private String displayFileName = null;
+
     public CollectionTreeNode(File file)
     {
@@ -55 +60 @@
     // To work with replace_srcdoc_with_html.pl
     this.is_srcreplaceable = CollectionDesignManager.plugin_manager.isFileSrcReplaceable(file);
+    
+    displayFileName = calcDisplayString();
+    }
+
+    /** Similar to calcDisplayString, but this version only checks whether the filename
+     * String is UTF8. If so, it converts it into UTF8 and returns it. If the filename
+     * is not UTF8, it is returned unchanged.
+     */ 
+    protected String quickCalcDisplayString() {
+    String filename = super.toString();
+    try{
+        String utf8filename = new String(filename.getBytes(), "UTF8");
+        if(utf8filename.indexOf('\ufffd') == -1) {
+        return utf8filename;
+        } else { // contains the character indicating that it's invalid utf8
+        // return the original string
+        return filename;
+        }
+    } catch(java.io.UnsupportedEncodingException e) {
+        return filename;
+    }
+    }
+
+    /** This method returns a string representation of the filenodes in the Collection
+     * Tree, that can then be displayed in the tree. 
+     * We'll initially assume that the filenames are utf8 encoded and so convert the
+     * filename into utf8 for proper presentation in the Collection tree pane.
+     * If the filenames are not utf8, then the conversion would have introduced funny
+     * characters. Therefore, when converting to utf8, if the converted filename 
+     * contains the special character '\ufffd', then we know the conversion did not work
+     * and we return the original string which may or may not be properly presented by
+     * default.
+     * See http://java.sun.com/j2se/1.4.2/docs/api/java/nio/charset/CharsetDecoder.html
+     * which says "How a decoding error is handled depends upon the action requested for 
+     * that type of error, which is described by an instance of the CodingErrorAction class.
+     * The possible error actions are to ignore the erroneous input, report the error to
+     * the invoker via the returned CoderResult object, or replace the erroneous input with
+     * the current value of the replacement string. The replacement has the initial value 
+     * "\uFFFD"; its value may be changed via the replaceWith method."
+     * The following made me think that String(byte[], String charsetName) constructor may
+     * use the replacement value \uFFFD.
+     * http://www.experts-exchange.com/Programming/Programming_Languages/Java/Q_20512969.html
+     * mentions the following which made me think of this: 
+     * convertedStr = convertedStr.replace('\ufffd', ' ');
+     */
+    protected String calcDisplayString() {
+    String filename = super.toString();
+
+    try{
+        String[] charsets = {"UTF-8", "ISO-8859-1", "US-ASCII", "UTF-16BE", "UTF-16LE", "UTF-16"};
+        for(int i = 0; i < charsets.length; i++) {
+        String charset = charsets[i];
+        String convertedfilename = new String(filename.getBytes(), charset);
+        if(convertedfilename.indexOf('\ufffd') != -1) {
+            // encountered the character that indicates when the conversion is invalid
+            convertedfilename = null;
+            continue; // try to use the next charset to encode the filename as
+        } 
+        else { // valid conversion
+            System.err.println("Filename was encoded in: " + charset);
+            return convertedfilename;
+        }
+        }
+    } catch(java.io.UnsupportedEncodingException e) {
+        return filename;
+    }
+
+    return filename; // attempted conversions all failed
     }
 
@@ -77 +150 @@
     return is_srcreplaceable;
     }
+
+    /** This method returns a string representation of the filenodes in the Collection
+     * Tree, which is what will be displayed in the tree. It tries to convert it into
+     * some common encoding formats. Failing that, the unchanged filepath is returned.
+     * @see displayString
+     */ 
+    public String toString() 
+    {
+    if(displayFileName == null) {    
+      displayFileName = calcDisplayString();
+    }
+    return displayFileName;
+    }
 }