Changeset 5040

Timestamp:

2003-07-25T16:44:43+12:00 (21 years ago)

Author:

jmt12

Message:

Modified searching process so that a valid metadata.xml can be read in, even if it isn't part of a current collection. This functionality was an unitentional side-effect of searching for required hierarchy files. Note that if reading in a so-called 'wild metadata' file, then the profile actions generated will be matched against that files parent folders only (whereas it would be against a particular collection's name in the normal case).

File:

• trunk/gli/src/org/greenstone/gatherer/msm/parsers/GreenstoneMetadataParser.java (modified) (19 diffs)

trunk/gli/src/org/greenstone/gatherer/msm/parsers/GreenstoneMetadataParser.java

@@ -1 @@
-package org.greenstone.gatherer.msm.parsers;
 /**
  *#########################################################################
@@ -7 +6 @@
  * University of Waikato, New Zealand.
  *
- * <BR><BR>
- *
  * Author: John Thompson, Greenstone Digital Library, University of Waikato
  *
- * <BR><BR>
- *
  * Copyright (C) 1999 New Zealand Digital Library Project
- *
- * <BR><BR>
  *
  * This program is free software; you can redistribute it and/or modify
@@ -22 +15 @@
  * (at your option) any later version.
  *
- * <BR><BR>
- *
  * This program is distributed in the hope that it will be useful,
  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  * GNU General Public License for more details.
- *
- * <BR><BR>
  *
  * You should have received a copy of the GNU General Public License
@@ -36 +25 @@
  *########################################################################
  */
+package org.greenstone.gatherer.msm.parsers;
+/**************************************************************************************
+ * Written:      ??/??/02
+ * Revised:      ??/??/02 - Commented
+ *               25/07/03 - Fix to allow any valid greenstone metadata.xml to be imported from, not just those that occur within a collection. This functionality is accidental and caused by GLI attempting to find a collect.cfg to extract hierarchy file information from, and failing.
+ **************************************************************************************/
 import java.io.*;
 import java.lang.ref.*;
@@ -41 +36 @@
 import java.util.*;
 import java.util.regex.*;
-import javax.swing.JOptionPane;
+import javax.swing.*;
 import javax.swing.tree.*;
 import org.greenstone.gatherer.Gatherer;
@@ -55 +50 @@
 import org.greenstone.gatherer.valuetree.GValueModel;
 import org.greenstone.gatherer.valuetree.GValueNode;
-import org.w3c.dom.Document;
-import org.w3c.dom.Element;
-import org.w3c.dom.Node;
-import org.w3c.dom.NodeList;
+import org.w3c.dom.*;
 /** Provides a metadata parser implementation that knows how to locate, prepare for, then import metadata from a previous Greenstone collection. Is aware of such factors as the presence of Metadata Set files and hierarchy files. Updates the profiler where possible to allow for faster subsequent imports from a certain collection. Caches all the information about encountered collections in CollectCFG objects which are softly cached (ie are cached, but are reclaimed before an OutOfMemory exception would be thrown). 
  * @author John Thompson, Greenstone Digital Library, University of Waikato
@@ -137 +129 @@
     }
 
-    // Continue only if we are sure this is a greenstone collection
-    if(collection_dir != null && collect_cfg != null) {
-        // 2. Attempt to merge in any mdses and make note of those that are successfully imported (by removing reference from collect.cfg).
-        ///ystem.err.print("2 ");
+    // 2. If a collection configuration file was found, attempt to merge in any mdses and make note of those that are successfully imported (by removing reference from collect.cfg).
+    ///ystem.err.print("2 ");
+    if(collect_cfg != null) {
         ArrayList mdses = collect_cfg.getMetadataSets();
         for(int i = 0; i < mdses.size(); i++) {
@@ -148 +139 @@
         mdses.clear();
         mdses = null;
-
-        // 3. Locate all of the metadata.xml files that may have an affect on the origin file. Make sure the metadata.xml closest to the origin files directory is last (to ensure property inheritance regarding accumulate/overwrite).
-        ///ystem.err.print("3 ");
-        ArrayList search_files = new ArrayList();
-        File file = origin.getFile();
-        String filename = null;
-        boolean file_level;
-        if(file.isFile()) {
-        file_level = false;
+    }
+
+    // 3. Locate all of the metadata.xml files that may have an affect on the origin file. Make sure the metadata.xml closest to the origin files directory is last (to ensure property inheritance regarding accumulate/overwrite).
+    ///ystem.err.print("3 ");
+    ArrayList search_files = new ArrayList();
+    File file = origin.getFile();
+    String filename = null;
+    boolean file_level;
+    if(file.isFile()) {
+        file_level = false;
+        filename = file.getName();
+        file = file.getParentFile();
+    }
+    else {
+        file_level = true;
+    }
+    while(file != null && (collection_dir == null || !file.equals(collection_dir))) {
+        File test_file = new File(file, Utility.METADATA_XML);
+        if(test_file.exists()) {
+        search_files.add(0, new MetadataXMLFileSearch(test_file, filename));
+        }
+        if(filename != null) {
+        filename = file.getName() + SEPARATOR + filename;
+        }
+        else {
         filename = file.getName();
-        file = file.getParentFile();
-        }
-        else {
-        file_level = true;
-        }
-        while(!file.equals(collection_dir)) {
-        File test_file = new File(file, Utility.METADATA_XML);
-        if(test_file.exists()) {
-            search_files.add(0, new MetadataXMLFileSearch(test_file, filename));
-        }
-        if(filename != null) {
-            filename = file.getName() + SEPARATOR + filename;
+        }
+        file = file.getParentFile();
+    }
+    filename = null;
+    file = null;
+    // Start with an initially empty ArrayList of metadata
+    ArrayList metadata = new ArrayList();
+    // Now search each of these metadata xml for metadata, remembering to accumulate or overwrite as we go along.
+    for(int i = 0; i < search_files.size(); i++) {
+        MetadataXMLFileSearch a_search = (MetadataXMLFileSearch) search_files.get(i);
+        ///ystem.err.println("Search " + a_search.file.getAbsolutePath() + " for " + (a_search.filename != null ? a_search.filename : ".*"));
+        // Retrieve the document
+        BasicGDMDocument document = getDocument(a_search.file);
+        if(document != null) {
+        // If this is a dummy run, our original source file is actually the metadata.xml file and we retrieve all metadata for this collection, as if accumulated!
+        if(dummy_run) {
+            metadata = document.getAllMetadata();
         }
         else {
-            filename = file.getName();
-        }
-        file = file.getParentFile();
-        }
-        filename = null;
-        file = null;
-        // Start with an initially empty ArrayList of metadata
-        ArrayList metadatum = new ArrayList();
-        // Now search each of these metadata xml for metadata, remembering to accumulate or overwrite as we go along.
-        for(int i = 0; i < search_files.size(); i++) {
-        MetadataXMLFileSearch a_search = (MetadataXMLFileSearch) search_files.get(i);
-        ///ystem.err.println("Search " + a_search.file.getAbsolutePath() + " for " + (a_search.filename != null ? a_search.filename : ".*"));
-        // Retrieve the document
-        BasicGDMDocument document = getDocument(a_search.file);
-        if(document != null) {
-            // If this is a dummy run, our original source file is actually the metadata.xml file and we retrieve all metadata for this collection, as if accumulated!
-            if(dummy_run) {
-            metadatum = document.getAllMetadata();
-            }
-            else {
-            metadatum = document.getMetadata(a_search.filename, metadatum, folder_level);
-            }
-            document = null;
-        }
-        a_search = null;
-        }
-        search_files = null;
-        // Finally assign the metadata
-        ///ystem.err.println("Found " + metadatum.size() + " pieces of metadata for " + destination);
-        if(metadatum.size() > 0) {
-        addMetadata(destination, metadatum, collection_dir, collect_cfg, dummy_run);
-        }
-    }
-    else {
-        ///ystem.err.println("Not a greenstone collection (no collect.cfg found).");
+            metadata = document.getMetadata(a_search.filename, metadata, folder_level);
+        }
+        document = null;
+        }
+        a_search = null;
+    }
+    search_files = null;
+    // Finally assign the metadata
+    ///ystem.err.println("Found " + metadata.size() + " pieces of metadata for " + destination);
+    if(metadata.size() > 0) {
+        addMetadata(origin, destination, metadata, collection_dir, collect_cfg, dummy_run);
     }
     return dialog_cancelled;
@@ -215 +203 @@
     }
 
-    private void addMetadata(FileNode destination, ArrayList metadatum, File collection_dir, CollectCFG collect_cfg, boolean dummy_run) {
+    private void addMetadata(FileNode origin, FileNode destination, ArrayList metadata, File collection_dir, CollectCFG collect_cfg, boolean dummy_run) {
     // before we try to addMetadata, we need to check that there are some metadata sets for the collection - otherwise we cant add or import
     Vector meta_sets = Gatherer.c_man.getCollection().msm.getSets(false);
@@ -230 +218 @@
     ///ystem.err.print("6 ");
     // Used in a complicated test later on.
-    for(int i = 0; !dialog_cancelled && i < metadatum.size(); i++) {
-        BasicMetadata basic_metadata = (BasicMetadata) metadatum.get(i);
-        BasicMetadata metadata = (BasicMetadata) metadatum.get(i);
-        metadata.collection = collection_dir;
+    for(int i = 0; !dialog_cancelled && i < metadata.size(); i++) {
+        BasicMetadata basic_metadata = ((BasicMetadata) metadata.get(i)).copy();
+        BasicMetadata metadatum = (BasicMetadata) metadata.get(i);
+        metadatum.collection = collection_dir; // May be null. Doesn't matter.
         Metadata final_metadata = null;
         // If this BasicMetadata already exists in the transform cache then we can save ourselves a lot of work.
@@ -242 +230 @@
         if(final_metadata == null) {
         ///ystem.err.println("No existing Metadata object for BasicMetadata: " + basic_metadata);
-        // 6a. Check if an hfile is associated with this metadata, and if so load it, cache it in the collection.cfg object, then resolve metadata value index.
-        HFile h_file = collect_cfg.getHFile(metadata.element);
-        if(h_file != null && !dummy_run) {
-            ///ystem.err.print(metadata.value + " maps to ");
-            metadata.value = h_file.getValue(metadata.value);
-            ///ystem.err.println(metadata.value);
-        }
-        h_file = null;
-        // 6b. Check if there is a profile regarding the current metadata.
-        ///ystem.err.println("Retrieve existing action: " + collection_dir.getAbsolutePath() + ", " + metadata.element);
-        if(Gatherer.c_man.getCollection().msm.profiler.containsAction(collection_dir.getAbsolutePath(), metadata.element)) {
-            String new_element_name = Gatherer.c_man.getCollection().msm.profiler.getAction(collection_dir.getAbsolutePath(), metadata.element);
-            ///ystem.err.println("Profile result = " + new_element_name);
-            if(new_element_name == null) {
-            metadata = null;
-            }
-            else {
-            metadata.element = new_element_name;
-            }
-            new_element_name = null;
-        }
-        ///atherer.println("Assigning metadata.");
-        if(metadata != null) {
+        // 6a. Check if an hfile is associated with this metadata, and if so load it, cache it in the collection.cfg object, then resolve metadata value index. Of course we can only do this if a collection configuration file was found in the first place.
+        if(collect_cfg != null) {
+            HFile h_file = collect_cfg.getHFile(metadatum.element);
+            if(h_file != null && !dummy_run) {
+            ///ystem.err.print(metadata.value + " maps to ");
+            metadatum.value = h_file.getValue(metadatum.value);
+            ///ystem.err.println(metadatum.value);
+            }
+            h_file = null;
+        }
+        // 6b. Check if there is a profile regarding the current metadata. The profile may be stored for the collection directory, or if no such directory is available, then try the ancestor folders of the origin file.
+        ///ystem.err.println("Retrieve existing action: " + collection_dir.getAbsolutePath() + ", " + metadatum.element);
+        if(collection_dir != null) {
+            // Note that the first test is whether a profile action exist, while the 'getAction' can return null as the profile action. 
+            if(Gatherer.c_man.getCollection().msm.profiler.containsAction(collection_dir.getAbsolutePath(), metadatum.element)) {
+            String new_element_name = Gatherer.c_man.getCollection().msm.profiler.getAction(collection_dir.getAbsolutePath(), metadatum.element);
+            ///ystem.err.println("Profile result = " + new_element_name);
+            if(new_element_name == null) {
+                metadatum = null;
+            }
+            else {
+                metadatum.element = new_element_name;
+            }
+            new_element_name = null;
+            }
+        }
+        else {
+            boolean found = false;
+            File current_folder = origin.getFile().getParentFile();
+            while(!found && metadatum != null && current_folder != null) {
+            if(Gatherer.c_man.getCollection().msm.profiler.containsAction(current_folder.getAbsolutePath(), metadatum.element)) {
+                found = true;
+                String new_element_name = Gatherer.c_man.getCollection().msm.profiler.getAction(current_folder.getAbsolutePath(), metadatum.element);
+                ///ystem.err.println("Profile result = " + new_element_name);
+                if(new_element_name == null) {
+                metadatum = null;
+                }
+                else {
+                metadatum.element = new_element_name;
+                }
+                new_element_name = null;
+            }
+            else {
+                current_folder = current_folder.getParentFile();
+            }
+            }
+            current_folder = null;
+        }
+        ///atherer.println("Assigning metadatum.");
+        if(metadatum != null) {
             // 6c. Try to add metadata. If there is no matching metadata element:
-            ElementWrapper element = Gatherer.c_man.getCollection().msm.getElement(metadata.element);
+            ElementWrapper element = Gatherer.c_man.getCollection().msm.getElement(metadatum.element);
             // Arg. The element returned may come from the Greenstone dls, which of course should never be involved during importing. To solve check the namespace isn't "" and if it is nullify the element. Nullify. NULLIFY, Bwuhahahaha...
             if(element != null && element.getNamespace().equals("")) {
@@ -273 +288 @@
             // 6ci. If no match exists, prompt the user to add/merge with specific metadata element. The user can also choose to ignore this metadata.
             if(element == null) {
-            element = selectElement(metadata.element);
+            element = selectElement(metadatum.element);
             if(!dialog_cancelled) {
                 // 6ciii. If either of the above work, remember to add to profile.
                 if(element == null) {
-                ///ystem.err.println("Adding profile action: " + collection_dir.getAbsolutePath() + ", " + metadata.element + ", null");
-                Gatherer.c_man.getCollection().msm.profiler.addAction(collection_dir.getAbsolutePath(), metadata.element, null);
+                ///ystem.err.println("Adding profile action: " + collection_dir.getAbsolutePath() + ", " + metadatum.element + ", null");
+                if(collection_dir != null) {
+                    Gatherer.c_man.getCollection().msm.profiler.addAction(collection_dir.getAbsolutePath(), metadatum.element, null);
+                }
+                else {
+                    Gatherer.c_man.getCollection().msm.profiler.addAction(origin.getFile().getParentFile().getAbsolutePath(), metadatum.element, null);
+                }
                 }
                 else {
-                ///ystem.err.println("Adding profile action: " + collection_dir.getAbsolutePath() + ", " + metadata.element + ", " + element.getName());
-                Gatherer.c_man.getCollection().msm.profiler.addAction(collection_dir.getAbsolutePath(), metadata.element, element.getName());
+                ///ystem.err.println("Adding profile action: " + collection_dir.getAbsolutePath() + ", " + metadatum.element + ", " + element.getName());
+                if(collection_dir != null) {
+                    Gatherer.c_man.getCollection().msm.profiler.addAction(collection_dir.getAbsolutePath(), metadatum.element, null);
+}
+                else {
+                    Gatherer.c_man.getCollection().msm.profiler.addAction(origin.getFile().getParentFile().getAbsolutePath(), metadatum.element, null);
+                }
                 }
             }
@@ -292 +317 @@
             if(model != null) {
                 // One little 'fix' for importing from the demo or dls files. The Title metadata found in the metadata.xml isn't used in preference for the automatically extracted Titles. However we want to use them, so we should remove '.*(<filename>)$' for a certain file <filename>.
-                String raw_value = metadata.value.trim();
+                String raw_value = metadatum.value.trim();
                 String filename_munged = destination.getFile().getName();
                 int index = -1;
@@ -299 +324 @@
                 }
                 filename_munged = "(" + filename_munged + ")";
-                ///atherer.println("Hack: filename = " + destination.getFile().getName() + ", munged = " + filename_munged + ", raw_value = " + raw_value);
                 if(raw_value.endsWith(filename_munged)) {
                 raw_value = (raw_value.substring(0, raw_value.length() - filename_munged.length())).trim();
@@ -305 +329 @@
                 GValueNode node = model.addValue(raw_value);
                 final_metadata = new Metadata(element, node);
-                ///ystem.err.println("Adding final metadata: " + metadata.toString());
+                ///ystem.err.println("Adding final metadata: " + metadatum.toString());
                 node = null;
             }
@@ -322 +346 @@
         }
         if(!dummy_run && final_metadata != null && !dialog_cancelled) {
-        final_metadata.setAccumulate(metadata.accumulates);
+        final_metadata.setAccumulate(metadatum.accumulates);
         // Now we can finally add the metadata.
         ///ystem.err.println("Adding Metadata: " + final_metadata);
@@ -329 +353 @@
                 // Otherwise there is no way to add this metadata. No value model no metadata value.
         final_metadata = null;
-        metadata = null;
+        metadatum = null;
     }
     }
@@ -623 +647 @@
     }
 
+    public BasicMetadata copy() {
+        return new BasicMetadata(element, value, accumulates);
+    }
+
     public int compareTo(Object other) {
         return toString().compareTo(other.toString());
@@ -632 +660 @@
     public boolean equals(Object object) {
         BasicMetadata other = (BasicMetadata) object;
-        if(collection != null) {
+        if(collection != null && other.collection != null) {
         return (collection.equals(other.collection) && element.equals(other.element) && value.equals(other.value));
         }