1 /*
2 * This file is part of *** M y C o R e ***
3 * See http://www.mycore.de/ for details.
4 *
5 * MyCoRe is free software: you can redistribute it and/or modify
6 * it under the terms of the GNU General Public License as published by
7 * the Free Software Foundation, either version 3 of the License, or
8 * (at your option) any later version.
9 *
10 * MyCoRe is distributed in the hope that it will be useful,
11 * but WITHOUT ANY WARRANTY; without even the implied warranty of
12 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13 * GNU General Public License for more details.
14 *
15 * You should have received a copy of the GNU General Public License
16 * along with MyCoRe. If not, see <http://www.gnu.org/licenses/>.
17 */
18
19 package org.mycore.mods.enrichment;
20
21 import java.io.IOException;
22 import java.net.URLEncoder;
23 import java.nio.charset.StandardCharsets;
24 import java.text.MessageFormat;
25 import java.util.Locale;
26
27 import org.apache.logging.log4j.LogManager;
28 import org.apache.logging.log4j.Logger;
29 import org.jdom2.Document;
30 import org.jdom2.Element;
31 import org.mycore.common.xml.MCRURIResolver;
32 import org.mycore.common.xml.MCRXMLHelper;
33 import org.mycore.mods.MCRMODSCommands;
34 import org.xml.sax.SAXException;
35
36 /**
37 * Returns publication data in MODS format for a given identifier.
38 * Each resolver belongs to a certain data source, e.g. the data source
39 * "PubMed" may have two resolves to get publication data by DOI or PubMed artice ID.
40 *
41 * The resolver will use an URI to get the publication data.
42 * MCR.MODS.EnrichmentResolver.DataSource.[SourceID].[TypeID].URI=[URI]
43 *
44 * This is typically a HTTP URL followed by a XSL stylesheet to transform the
45 * source format to MODS, e.g.
46 * MCR.MODS.EnrichmentResolver.DataSource.DataCite.doi.URI=xslStyle:datacite2mods:https://data.datacite.org/application/vnd.datacite.datacite+xml/{0}
47 *
48 * Within the URI, the pattern {0} will be replaced by the given identifier value,
49 * optionally the pattern {1} will be replaced by the value uri-encoded as http request parameter
50 *
51 * @author Frank L\u00FCtzenkirchen
52 */
53 class MCRIdentifierResolver {
54
55 private static final Logger LOGGER = LogManager.getLogger(MCRIdentifierResolver.class);
56
57 private MCRDataSource ds;
58
59 private MCRIdentifierType idType;
60
61 private String uriPattern;
62
63 MCRIdentifierResolver(MCRDataSource ds, MCRIdentifierType idType, String uriPattern) {
64 this.ds = ds;
65 this.idType = idType;
66 this.uriPattern = uriPattern;
67 }
68
69 MCRIdentifierType getType() {
70 return idType;
71 }
72
73 /**
74 * Tries to resolve publication data for the given identifier.
75 *
76 * @param identifier the identifier's value, e.g. a DOI or ISBN
77 * @return the publication data in MODS format, or null if the data source did not return data for this identifier
78 */
79 Element resolve(String identifier) {
80 Object[] params = new Object[] { identifier, URLEncoder.encode(identifier, StandardCharsets.UTF_8) };
81 String uri = new MessageFormat(uriPattern, Locale.ROOT).format(params);
82
83 Element resolved = null;
84 try {
85 resolved = MCRURIResolver.instance().resolve(uri);
86 } catch (Exception ex) {
87 LOGGER.warn("Exception resolving " + uri, ex);
88 return null;
89 }
90
91 // Normalize various error/not found cases
92 if (resolved == null || !"mods".equals(resolved.getName()) || resolved.getChildren().isEmpty()) {
93 LOGGER.warn(ds + " returned none or empty MODS for " + idType + " " + identifier);
94 return null;
95 }
96
97 try {
98 ensureIsValidMODS(resolved);
99 return resolved;
100 } catch (Exception ex) {
101 LOGGER.warn(ds + " returned invalid MODS for " + identifier + ": " + ex.getMessage(), ex);
102 return null;
103 }
104 }
105
106 void ensureIsValidMODS(Element mods) throws SAXException, IOException {
107 MCRXMLHelper.validate(new Document().addContent(mods.detach()), MCRMODSCommands.MODS_V3_XSD_URI);
108 }
109 }