Changeset 7204

Timestamp:

2004-04-19T17:14:16+12:00 (20 years ago)

Author:

mdewsnip

Message:

Changed the processing of the doc.xml files so metadata from bibliographic files isn't completely ignored -- now the values are, but the metadata elements are extracted and added to the extracted metadata set.

File:

• trunk/gli/src/org/greenstone/gatherer/msm/GreenstoneArchiveParser.java (modified) (2 diffs)

trunk/gli/src/org/greenstone/gatherer/msm/GreenstoneArchiveParser.java

@@ -61 +61 @@
 
     private GShell shell;
-
-    static final String ignore_list[] = {"assocfilepath", "gsdl",  "Identifier", "URL"}; //"Source",
+    private String file_path;
+
+    static final String ignore_list[] = { "assocfilepath", "gsdl", "Identifier", "SourceSegment", "URL" };
 
     public GreenstoneArchiveParser(GShellProgressMonitor progress, GShell shell) {
@@ -103 +104 @@
     Document document = Utility.parse(file, false);
 
-    Gatherer.println("Parsed greenstone archive document: " + file.getAbsolutePath());
     // If we successfully parsed the document, then it is time to search through the DOM for the Metadata tags.
-    if(document != null) {
-        String file_path = null;
+    if (document != null) {
+        file_path = null;
         Element archive_element = document.getDocumentElement();
         // Retrieve the initial Section element
         NodeList section_elements = archive_element.getElementsByTagName(StaticStrings.SECTION_ELEMENT);
-        // It may be the case that we have some sort of mutant freak archive file which doesn't have any section elements.
-        if(section_elements.getLength() < 1) {
+        // It may be the case that we have some sort of mutant freak archive file which doesn't have any section elements
+        if (section_elements.getLength() < 1) {
         return count; 
         }
         Element section_element = (Element) section_elements.item(0);       
         section_elements = null;
-        // Retrieve all of the Metadata sections.
+
+        // Retrieve all of the Metadata sections
         NodeList metadata_elements = section_element.getElementsByTagName(StaticStrings.METADATA_ELEMENT);
         section_element = null;
-        // We first zip through the retrieved metadata, and if we encounter the element 'SourceSegment' - a sure sign this collection came from a bibliographic type file - we break out of extracted metadata parsing as no sense could be made of the data extracted anyway (plus we suffer a death of thirty-thousand pointy bits of metadata!)
-        for(int i = 0; i < metadata_elements.getLength(); i++) {
+
+        // Zip through the retrieved metadata checking for SourceSegment elements
+        // These are a good sign of bibliographic files, which we must handle specially
+        boolean ignore_values = false;
+        for (int i = 0; i < metadata_elements.getLength(); i++) {
         Element metadata_element = (Element) metadata_elements.item(i);
         String name = metadata_element.getAttribute(StaticStrings.NAME_ATTRIBUTE);
-        if(name.equalsIgnoreCase(StaticStrings.SOURCESEGMENT_VALUE)) {
-            return 0;
-        }
-        }
+        if (name.equalsIgnoreCase(StaticStrings.SOURCESEGMENT_VALUE)) {
+            ignore_values = true;
+            break;
+        }
+        }
+
         // Now for each Metadata entry retrieved...
-        for(int i = 0; i < metadata_elements.getLength(); i++) {
+        for (int i = 0; i < metadata_elements.getLength(); i++) {
         Element metadata_element = (Element) metadata_elements.item(i);
-        String name = metadata_element.getAttribute(StaticStrings.NAME_ATTRIBUTE);
-        // There is also a special case when the metadata name is gsdlsourcefilename, as we use this to find the FileRecord we want to add metadata to.
-        if(name.equals("gsdlsourcefilename")) {
-            file_path = MSMUtils.getValue(metadata_element);
-        }
-        else {
-            // Check if its name starts with, or is equal to, one of the values in our ignore list, and if so ignore this metadata.
-            boolean ignore = false;
-            for(int j = 0; !ignore && j < ignore_list.length; j++) {
-            ignore = name.startsWith(ignore_list[j]);
-            }
-            // Otherwise ensure the metadata is present in our collection.
-            if(!ignore && file_path != null) {
-            // If we successfully retrieved a record we can continue.
-            if(file_path != null) {
-                // We now retrieve the appropriate element. If no such element exists we create a new one in the greenstone mds. Remember that no element in the greenstone mds has an associated value tree, so it is perfect for metadata elements with a small number of repeated values but where the values have no relation between files (such as encoding, where many files will be iso_8859_1, but if you change one you don't intend to change them all).
-                ElementWrapper element = Gatherer.c_man.msm.getElement(name);
-                if(element == null) {
-                MetadataSet extracted_mds = Gatherer.c_man.msm.getSet(Utility.EXTRACTED_METADATA_NAMESPACE);
-                if(extracted_mds != null) {
-                    element = extracted_mds.addElement(name, Gatherer.config.getLanguage());
-                }
-                }
-                // If we successfully retrieved an element (and we should have) we can continue.
-                // WARNING!! There is one known exception - MARC records. Adding the extracted elements is all good, but adding the extracted metadata causes the whole thing to collapse in a pile of unhappy.
-                if(element != null && !file_path.endsWith(StaticStrings.MARC_EXTENSION) && (element.getNamespace().equals("") || element.getNamespace().equals(Utility.EXTRACTED_METADATA_NAMESPACE))) {
-                // Retrieve the metadata for the current file
-                File target_file = new File(file_path);
-                String value = "";
-                try {
-                    value = Utility.decodeGreenstone(URLDecoder.decode(MSMUtils.getValue(metadata_element), "UTF-8"));
-                }
-                catch(IllegalArgumentException error) { // ****
-                    value = Utility.decodeGreenstone(MSMUtils.getValue(metadata_element));
-                }
-                catch(UnsupportedEncodingException error) {
-                    Gatherer.printStackTrace(error);
-                }
-                // If we successfully retrieved a value we can continue.
-                if(value != null) {
-                    // Create a new metadata object.
-                    GValueModel value_tree = Gatherer.c_man.msm.getValueTree(element);
-                    GValueNode value_node = null;
-                    if(value_tree != null) {
-                    value_node = value_tree.getValue(value);
-                    }
-                    else {
-                    value_node = new GValueNode(element.toString(), value);
-                    }
-                    Metadata metadata = new Metadata(element, value_node);
-                    element.inc();
-                    ///ystem.err.println("Adding extracted metadata: " + metadata);
-                    Gatherer.c_man.getCollection().msm.fireMetadataChanged(new MSMEvent(this, System.currentTimeMillis(), target_file, null, metadata));
-                    count++;
-                    // All done. On to next metadata.
-                }
-                value = null;
-                target_file = null;
-                }
-                else {
-                Gatherer.println("Cannot retrieve metadata element " + name);
-                }
-            }
-            }
-        }
-        }
-    }
+        if (processMetadataElement(metadata_element, ignore_values) == true) {
+            count++;
+        }
+        }
+    }
+
     return count;
     }
+
+
+    private boolean processMetadataElement(Element metadata_element, boolean ignore_values)
+    {
+    String name = metadata_element.getAttribute(StaticStrings.NAME_ATTRIBUTE);
+
+    // There is also a special case when the metadata name is gsdlsourcefilename, as we use this to find the FileRecord we want to add metadata to.
+    if (name.equals("gsdlsourcefilename")) {
+        file_path = MSMUtils.getValue(metadata_element);
+        return false;
+    }
+
+    // Check if its name starts with, or is equal to, one of the values in our ignore list, and if so ignore this metadata
+    for (int j = 0; j < ignore_list.length; j++) {
+        if (name.startsWith(ignore_list[j])) {
+        return false;
+        }
+    }
+
+    if (file_path == null) {
+        return false;
+    }
+
+    // We now retrieve the appropriate element. If no such element exists we create a new one in the greenstone mds. Remember that no element in the greenstone mds has an associated value tree, so it is perfect for metadata elements with a small number of repeated values but where the values have no relation between files (such as encoding, where many files will be iso_8859_1, but if you change one you don't intend to change them all).
+    ElementWrapper element = Gatherer.c_man.msm.getElement(name);
+    if (element == null) {
+        MetadataSet extracted_mds = Gatherer.c_man.msm.getSet(Utility.EXTRACTED_METADATA_NAMESPACE);
+        if (extracted_mds != null) {
+        element = extracted_mds.addElement(name, Gatherer.config.getLanguage());
+        }
+    }
+
+    // If ignore_values is set (bibliographic records) we don't care about the values
+    if (ignore_values == true) {
+        return false;
+    }
+
+    // Retrieve the metadata for the current file
+    File target_file = new File(file_path);
+    String value = "";
+    try {
+        value = Utility.decodeGreenstone(URLDecoder.decode(MSMUtils.getValue(metadata_element), "UTF-8"));
+    }
+    catch (IllegalArgumentException error) { // ****
+        value = Utility.decodeGreenstone(MSMUtils.getValue(metadata_element));
+    }
+    catch (UnsupportedEncodingException error) {
+        Gatherer.printStackTrace(error);
+    }
+
+    if (value == null) {
+        return false;
+    }
+
+    // Create a new metadata object.
+    GValueModel value_tree = Gatherer.c_man.msm.getValueTree(element);
+    GValueNode value_node = null;
+    if (value_tree != null) {
+        value_node = value_tree.getValue(value);
+    }
+    else {
+        value_node = new GValueNode(element.toString(), value);
+    }
+
+    Metadata metadata = new Metadata(element, value_node);
+    element.inc();
+    ///ystem.err.println("Adding extracted metadata: " + metadata);
+    Gatherer.c_man.getCollection().msm.fireMetadataChanged(new MSMEvent(this, System.currentTimeMillis(), target_file, null, metadata));
+
+    return true;
+    }
+
 
     static final String metadata_ignore_list[] = {"assocfilepath", "gsdl", "Identifier","URL"};