source: trunk/gli/src/org/greenstone/gatherer/metadata/DocXMLFile.java@ 8131

Last change on this file since 8131 was 8131, checked in by mdewsnip, 20 years ago

More improvements to the new metadata code, including language-specific metadata element display and a 5x speed up in the skimming of the doc.xml files.

  • Property svn:keywords set to Author Date Id Revision
File size: 8.0 KB
Line 
1package org.greenstone.gatherer.metadata;
2
3
4import java.io.*;
5import java.util.*;
6import org.greenstone.gatherer.util.Utility;
7import org.greenstone.gatherer.util.XMLTools;
8import org.w3c.dom.*;
9
10
11/** This class represents one doc.xml file */
12public class DocXMLFile
13 extends File
14{
15 static final private String ARCHIVE_ELEMENT = "Archive";
16 static final private String DESCRIPTION_ELEMENT = "Description";
17 static final private String METADATA_ELEMENT = "Metadata";
18 static final private String SECTION_ELEMENT = "Section";
19
20 private ArrayList files_in_doc_xml_file = new ArrayList();
21
22
23 public DocXMLFile(String doc_xml_file_path)
24 {
25 super(doc_xml_file_path);
26
27 MetadataSet extracted_metadata_set = MetadataSetManager.getMetadataSet(MetadataSetManager.EXTRACTED_METADATA_NAMESPACE);
28
29 // Skim the doc.xml file as quickly as possible (don't parse as XML), looking at the Metadata elements
30 System.err.println("Skimming doc.xml file " + this + "...");
31 try {
32 BufferedReader buffered_reader = new BufferedReader(new FileReader(this));
33 String line = null;
34 while ((line = buffered_reader.readLine()) != null) {
35 // This line doesn't contain a metadata element
36 if (line.indexOf("<Metadata ") == -1) {
37 continue;
38 }
39
40 // Extract the metadata element name
41 int name_index = line.indexOf(" name=\"") + " name=\"".length();
42 String metadata_element_name_full = line.substring(name_index, line.indexOf("\"", name_index));
43
44 // If the metadata has a namespace it isn't extracted metadata, so we're not interested
45 String metadata_set_namespace = MetadataTools.getMetadataSetNamespace(metadata_element_name_full);
46 if (!metadata_set_namespace.equals("")) {
47 continue;
48 }
49
50 // Extracted metadata!
51 String metadata_element_name = metadata_element_name_full;
52
53 // Note which file this doc.xml is for
54 if (metadata_element_name.equals("gsdlsourcefilename")) {
55 // Extract the gsdlsourcefilename element value
56 int value_index = line.indexOf(">", name_index) + ">".length();
57 String gsdlsourcefilename_value = line.substring(value_index, line.indexOf("<", value_index));
58
59 // We're only interested in the path relative to the import folder
60 int import_index = gsdlsourcefilename_value.indexOf("import");
61 if (import_index != -1) {
62 gsdlsourcefilename_value = gsdlsourcefilename_value.substring(import_index + "import".length());
63
64 boolean is_unix_path = gsdlsourcefilename_value.startsWith("/");
65 gsdlsourcefilename_value = gsdlsourcefilename_value.substring(1);
66
67 // Make sure the path matches the OS that is running
68 if (is_unix_path && Utility.isWindows()) {
69 // Convert path from Unix to Windows
70 gsdlsourcefilename_value = gsdlsourcefilename_value.replaceAll("/", File.separator);
71 }
72 if (!is_unix_path && !Utility.isWindows()) {
73 // Convert path from Windows to Unix
74 gsdlsourcefilename_value = gsdlsourcefilename_value.replaceAll("\\\\", File.separator);
75 }
76
77 // Remember this for quick access later
78 if (gsdlsourcefilename_value != null) {
79 files_in_doc_xml_file.add(gsdlsourcefilename_value);
80 }
81 }
82 else {
83 // We don't really know what is going on...
84 System.err.println("Warning: Could not understand gsdlsourcefilename " + gsdlsourcefilename_value);
85 }
86 }
87
88 // Ignore lower-case metadata elements (gsdlsourcefilename, gsdlassocfile etc.)
89 // and those starting with '/' (/srclink)
90 char first_character = metadata_element_name.charAt(0);
91 if (Character.isLowerCase(first_character) || first_character == '/') {
92 continue;
93 }
94
95 MetadataElement metadata_element = extracted_metadata_set.getMetadataElement(metadata_element_name);
96 if (metadata_element == null) {
97 // This element isn't defined in ex.mds, so create it for this session
98 System.err.println("Extracted metadata element not defined: " + metadata_element_name);
99 extracted_metadata_set.addMetadataElementForThisSession(metadata_element_name);
100 }
101 }
102 }
103 catch (Exception ex) {
104 System.err.println("Exception: " + ex);
105 ex.printStackTrace();
106 }
107 }
108
109
110 public ArrayList getMetadataExtractedFromFile(File file)
111 {
112 // Build up a list of metadata extracted from this file
113 ArrayList metadata_values = new ArrayList();
114
115 String file_relative_path = file.getAbsolutePath();
116 int import_index = file_relative_path.indexOf("import");
117 if (import_index != -1) {
118 file_relative_path = file_relative_path.substring(import_index + "import".length() + 1);
119 }
120
121 // Check whether this doc.xml file contains extracted metadata for the specified file
122 boolean contains_extracted_metadata_for_file = false;
123 for (int i = 0; i < files_in_doc_xml_file.size(); i++) {
124 if (file_relative_path.equals(files_in_doc_xml_file.get(i))) {
125 contains_extracted_metadata_for_file = true;
126 // System.err.println("Found extracted metadata in file " + getAbsolutePath());
127 break;
128 }
129 }
130
131 // ...it doesn't
132 if (!contains_extracted_metadata_for_file) {
133 return metadata_values;
134 }
135
136 // Parse the doc.xml file
137 Document document = XMLTools.parseXMLFile(this);
138 if (document == null) {
139 System.err.println("Error: Could not parse doc.xml file " + getAbsolutePath());
140 return metadata_values;
141 }
142
143 MetadataSet extracted_metadata_set = MetadataSetManager.getMetadataSet(MetadataSetManager.EXTRACTED_METADATA_NAMESPACE);
144
145 // Read all the Archive elements in the file
146 NodeList archive_elements_nodelist = document.getElementsByTagName(ARCHIVE_ELEMENT);
147 for (int i = 0; i < archive_elements_nodelist.getLength(); i++) {
148 Element current_archive_element = (Element) archive_elements_nodelist.item(i);
149
150 // Read the child Section elements of the archive (but not all descendants)
151 ArrayList child_section_elements = XMLTools.getChildElementsByTagName(current_archive_element, SECTION_ELEMENT);
152 for (int j = 0; j < child_section_elements.size(); j++) {
153 Element current_section_element = (Element) child_section_elements.get(j);
154
155 // Read the Description elements of this section only (not child sections as well)
156 ArrayList child_description_elements = XMLTools.getChildElementsByTagName(current_section_element, DESCRIPTION_ELEMENT);
157 for (int k = 0; k < child_description_elements.size(); k++) {
158 Element current_description_element = (Element) child_description_elements.get(k);
159
160 // Read all the Metadata elements in this description element
161 NodeList metadata_elements_nodelist = current_description_element.getElementsByTagName(METADATA_ELEMENT);
162 for (int l = 0; l < metadata_elements_nodelist.getLength(); l++) {
163 Element current_metadata_element = (Element) metadata_elements_nodelist.item(l);
164 String metadata_element_name_full = current_metadata_element.getAttribute("name");
165
166 // If the metadata has a namespace it isn't extracted metadata, so we're not interested
167 String metadata_set_namespace = MetadataTools.getMetadataSetNamespace(metadata_element_name_full);
168 if (!metadata_set_namespace.equals("")) {
169 continue;
170 }
171
172 // Extracted metadata!
173 String metadata_element_name = metadata_element_name_full;
174
175 // Ignore lower-case metadata elements (gsdlsourcefilename, gsdlassocfile etc.)
176 // and those starting with '/' (/srclink)
177 char first_character = metadata_element_name.charAt(0);
178 if (Character.isLowerCase(first_character) || first_character == '/') {
179 continue;
180 }
181
182 MetadataElement metadata_element = extracted_metadata_set.getMetadataElement(metadata_element_name);
183
184 // Value trees are not stored for extracted metadata, so create a new value tree node now
185 String current_metadata_element_value = XMLTools.getElementTextValue(current_metadata_element);
186 metadata_element.addMetadataValue(current_metadata_element_value);
187 MetadataValueTreeNode metadata_value_tree_node = metadata_element.getMetadataValueTreeNode(current_metadata_element_value);
188
189 // Add the new metadata value to the list
190 MetadataValue metadata_value = new MetadataValue(metadata_element, metadata_value_tree_node);
191 metadata_values.add(metadata_value);
192 }
193 }
194 }
195 }
196
197 return metadata_values;
198 }
199}
Note: See TracBrowser for help on using the repository browser.