Changeset 6043 for SRUAggregator/trunk/src/main/java/eu/clarin/sru/fcs/aggregator/rest/RestService.java

Timestamp:

02/24/15 16:58:55 (9 years ago)

Author:

emanuel.dima@uni-tuebingen.de

Message:

22. alpha 21: downloading as text, csv, excel and tcf works

File:

• SRUAggregator/trunk/src/main/java/eu/clarin/sru/fcs/aggregator/rest/RestService.java (modified) (9 diffs)

SRUAggregator/trunk/src/main/java/eu/clarin/sru/fcs/aggregator/rest/RestService.java

@@ -13 +13 @@
 import eu.clarin.sru.fcs.aggregator.search.Search;
 import eu.clarin.sru.fcs.aggregator.lang.LanguagesISO693_3;
+import eu.clarin.sru.fcs.aggregator.search.ExportException;
 import eu.clarin.sru.fcs.aggregator.search.Exports;
 import java.io.IOException;
@@ -21 +22 @@
 import java.util.Map;
 import java.util.Set;
+import java.util.logging.Level;
+import java.util.logging.Logger;
 import javax.servlet.ServletContext;
 import javax.servlet.http.HttpServletRequest;
@@ -33 +36 @@
 import javax.ws.rs.core.MediaType;
 import javax.ws.rs.core.Response;
+import opennlp.tools.tokenize.TokenizerModel;
 import org.slf4j.LoggerFactory;
 
@@ -42 +46 @@
 @Path("/")
 public class RestService {
+
         private static final String EXPORT_FILENAME_PREFIX = "ClarinDFederatedContentSearch-";
         private static final String TCF_MEDIA_TYPE = "text/tcf+xml";
         private static final String EXCEL_MEDIA_TYPE = "application/vnd.openxmlformats-officedocument.spreadsheetml.sheet";
+        private static final String SEARCH_RESULTS_ENCODING = "UTF-8";
 
         private static final org.slf4j.Logger log = LoggerFactory.getLogger(RestService.class);
@@ -82 +88 @@
                                                 put("isScan", true);
                                                 put("institutions", scan.getInstitutions());
+                                                put("date", scan.getDate());
                                         }
                                 });
@@ -91 +98 @@
                                                 put("isScan", false);
                                                 put("institutions", search.getInstitutions());
+                                                put("date", scan.getDate());
                                         }
                                 });
@@ -105 +113 @@
                 log.info("get language codes", codes);
                 for (String code : codes) {
-                        String name = LanguagesISO693_3.getInstance().nameForCode(code);
+                        String name = LanguagesISO693_3.getInstance().nameForCode_3(code);
                         languages.put(code, name != null ? name : code);
                 }
@@ -184 +192 @@
                 }
 
-                if (format == null || format.trim().isEmpty() || format.trim().equals("text")) {
+        if (format == null || format.trim().isEmpty() || format.trim().equals("text")) {
                         String text = Exports.getExportText(search.getResults());
                         return download(text, MediaType.TEXT_PLAIN, search.getQuery() + ".txt");
@@ -215 +223 @@
                                 .build();
         }
+
+        @GET
+        @Path("search/{id}/toWeblicht")
+        public Response sendSearchResultsToWeblicht(@PathParam("id") Long searchId,
+                        @QueryParam("format") String format) throws Exception {
+                Search search = Aggregator.getInstance().getSearchById(searchId);
+                if (search == null) {
+                        return Response.status(Response.Status.NOT_FOUND).entity("Search job not found").build();
+                }
+
+                String url = null;
+                if (format == null || format.isEmpty() || format.trim().equals("text")) {
+                        String text = Exports.getExportText(search.getResults());
+                        if (text != null) {
+                                byte[] bytes = text.getBytes(SEARCH_RESULTS_ENCODING);
+                                url = DataTransfer.uploadToDropOff(bytes, "text/plan", ".txt");
+                        }
+                } else if (format.equals("tokens")) {
+                        byte[] bytes = Exports.getExportTokenizedTCF(
+                                        search.getResults(), search.getSearchLanguage(),
+                                        Aggregator.getInstance().getTokenizerModel());
+                        if (bytes != null) {
+                                url = DataTransfer.uploadToDropOff(bytes, "text/tcf+xml", ".tcf");
+                        }
+                } else {
+                        return Response.status(400).entity("incorrect format parameter").build();
+                }
+
+                return url == null ? Response.status(503).entity("error while exporting to weblicht").build()
+                                : Response.ok().entity(url).build();
+        }
 }