Examples of OAIPMHLoader

Examples of net.yacy.document.importer.OAIPMHLoader

60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
                String oaipmhurl = post.get("urlstartone");
                if (oaipmhurl.indexOf("?") < 0) oaipmhurl = oaipmhurl + "?verb=ListRecords&metadataPrefix=oai_dc";
                DigestURI url = null;
                try {
                    url = new DigestURI(oaipmhurl);
                    OAIPMHLoader r = new OAIPMHLoader(sb.loader, url, sb.surrogatesInPath, "oaipmh-one");
                    ResumptionToken rt = r.getResumptionToken();
                    prop.put("import-one", 1);
                    prop.put("import-one_count", (rt == null) ? "not available" : Integer.toString(rt.getRecordCounter()));
                    prop.put("import-one_source", r.source());
                    prop.put("import-one_rt", r.getResumptionToken().toString());
                   
                    // set next default url
                    try {
                        DigestURI nexturl = (rt == null) ? null : rt.resumptionURL();
                        if (rt != null) prop.put("defaulturl", (nexturl == null) ? "" : nexturl.toNormalform(true, false));
                    } catch (MalformedURLException e) {
                        prop.put("defaulturl", e.getMessage());
                    } catch (IOException e) {
                        // reached end of resumption
                        prop.put("defaulturl", e.getMessage());
                    }
                } catch (MalformedURLException e) {
                    Log.logException(e);
                    prop.put("import-one", 2);
                    prop.put("import-one_error", e.getMessage());
                } catch (IOException e) {
                    Log.logException(e);
                    prop.put("import-one", 2);
                    prop.put("import-one_error", e.getMessage());
                }
            }
           
            if (post.get("urlstart", "").length() > 0) {
                String oaipmhurl = post.get("urlstart", "");
                sb.tables.recordAPICall(post, "IndexImportOAIPMH_p.html", WorkTables.TABLE_API_TYPE_CRAWLER, "OAI-PMH import for " + oaipmhurl);
                DigestURI url = null;
                try {
                    url = new DigestURI(oaipmhurl);
                    OAIPMHImporter job = new OAIPMHImporter(sb.loader, url);
                    job.start();
                    prop.put("status", 1);
                    prop.put("optiongetlist", 1);
                    prop.put("iframetype", 1);
                } catch (MalformedURLException e) {
                    Log.logException(e);
                    prop.put("status", 2);
                    prop.put("status_message", e.getMessage());
                }
            }
           
           
            if (post.get("loadrows", "").length() > 0) {
                // create a time-ordered list of events to execute
                Set<String> sources = new TreeSet<String>();
                for (final Map.Entry<String, String> entry: post.entrySet()) {
                    if (entry.getValue().startsWith("mark_")) {
                        sources.add(entry.getValue().substring(5));
                    }
                }
                prop.put("status", 1);
                prop.put("optiongetlist", 1);
                prop.put("iframetype", 1);
               
                // prepare the set for random read from it (to protect the servers at the beginning of the list)
                List<String> sourceList = new ArrayList<String>(sources.size());
                for (String oaipmhurl: sources) sourceList.add(oaipmhurl);
                Random r = new Random(System.currentTimeMillis());
               
                // start jobs for the sources
                DigestURI url = null;
                while (sourceList.size() > 0) {
                    String oaipmhurl = sourceList.remove(r.nextInt(sourceList.size()));
                    try {
                        url = new DigestURI(oaipmhurl);
                        OAIPMHImporter job = new OAIPMHImporter(sb.loader, url);
                        job.start();
                    } catch (MalformedURLException e) {
View Full Code Here

Examples of net.yacy.document.importer.OAIPMHLoader

59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
                String oaipmhurl = post.get("urlstartone");
                if (oaipmhurl.indexOf('?',0) < 0) oaipmhurl = oaipmhurl + "?verb=ListRecords&metadataPrefix=oai_dc";
                DigestURI url = null;
                try {
                    url = new DigestURI(oaipmhurl);
                    final OAIPMHLoader r = new OAIPMHLoader(sb.loader, url, sb.surrogatesInPath, "oaipmh-one");
                    final ResumptionToken rt = r.getResumptionToken();
                    prop.put("import-one", 1);
                    prop.put("import-one_count", (rt == null) ? "not available" : Integer.toString(rt.getRecordCounter()));
                    prop.put("import-one_source", r.source());
                    prop.put("import-one_rt", r.getResumptionToken().toString());

                    // set next default url
                    try {
                        final DigestURI nexturl = (rt == null) ? null : rt.resumptionURL();
                        if (rt != null) prop.put("defaulturl", (nexturl == null) ? "" : nexturl.toNormalform(true, false));
                    } catch (final MalformedURLException e) {
                        prop.put("defaulturl", e.getMessage());
                    } catch (final IOException e) {
                        // reached end of resumption
                        prop.put("defaulturl", e.getMessage());
                    }
                } catch (final MalformedURLException e) {
                    Log.logException(e);
                    prop.put("import-one", 2);
                    prop.put("import-one_error", e.getMessage());
                } catch (final IOException e) {
                    Log.logException(e);
                    prop.put("import-one", 2);
                    prop.put("import-one_error", e.getMessage());
                }
            }

            if (post.get("urlstart", "").length() > 0) {
                final String oaipmhurl = post.get("urlstart", "");
                sb.tables.recordAPICall(post, "IndexImportOAIPMH_p.html", WorkTables.TABLE_API_TYPE_CRAWLER, "OAI-PMH import for " + oaipmhurl);
                DigestURI url = null;
                try {
                    url = new DigestURI(oaipmhurl);
                    final OAIPMHImporter job = new OAIPMHImporter(sb.loader, url);
                    job.start();
                    prop.put("status", 1);
                    prop.put("optiongetlist", 1);
                    prop.put("iframetype", 1);
                } catch (final MalformedURLException e) {
                    Log.logException(e);
                    prop.put("status", 2);
                    prop.put("status_message", e.getMessage());
                }
            }


            if (post.get("loadrows", "").length() > 0) {
                // create a time-ordered list of events to execute
                final Set<String> sources = new TreeSet<String>();
                for (final Map.Entry<String, String> entry: post.entrySet()) {
                    if (entry.getValue().startsWith("mark_")) {
                        sources.add(entry.getValue().substring(5));
                    }
                }
                prop.put("status", 1);
                prop.put("optiongetlist", 1);
                prop.put("iframetype", 1);

                // prepare the set for random read from it (to protect the servers at the beginning of the list)
                final List<String> sourceList = new ArrayList<String>(sources.size());
                for (final String oaipmhurl: sources) sourceList.add(oaipmhurl);
                final Random r = new Random(System.currentTimeMillis());

                // start jobs for the sources
                DigestURI url = null;
                while (sourceList.size() > 0) {
                    final String oaipmhurl = sourceList.remove(r.nextInt(sourceList.size()));
                    try {
                        url = new DigestURI(oaipmhurl);
                        final OAIPMHImporter job = new OAIPMHImporter(sb.loader, url);
                        job.start();
                    } catch (final MalformedURLException e) {
View Full Code Here
TOP
Copyright © 2018 www.massapi.com. All rights reserved.
All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.