1 | package eu.clarin.cmdi.vlo.importer; |
---|
2 | |
---|
3 | import java.io.File; |
---|
4 | import java.io.IOException; |
---|
5 | import java.net.MalformedURLException; |
---|
6 | import java.net.URL; |
---|
7 | import java.text.SimpleDateFormat; |
---|
8 | import java.util.ArrayList; |
---|
9 | import java.util.Arrays; |
---|
10 | import java.util.Date; |
---|
11 | import java.util.HashMap; |
---|
12 | import java.util.HashSet; |
---|
13 | import java.util.Iterator; |
---|
14 | import java.util.LinkedList; |
---|
15 | import java.util.List; |
---|
16 | import java.util.Map; |
---|
17 | import java.util.Set; |
---|
18 | |
---|
19 | import org.apache.commons.cli.CommandLine; |
---|
20 | import org.apache.commons.cli.CommandLineParser; |
---|
21 | import org.apache.commons.cli.Options; |
---|
22 | import org.apache.commons.cli.PosixParser; |
---|
23 | import org.apache.commons.io.FileUtils; |
---|
24 | import org.apache.solr.client.solrj.SolrQuery; |
---|
25 | import org.apache.solr.client.solrj.SolrServerException; |
---|
26 | import org.apache.solr.client.solrj.impl.ConcurrentUpdateSolrServer; |
---|
27 | import org.apache.solr.client.solrj.util.ClientUtils; |
---|
28 | import org.apache.solr.common.SolrDocumentList; |
---|
29 | import org.apache.solr.common.SolrInputDocument; |
---|
30 | import org.apache.solr.common.params.MapSolrParams; |
---|
31 | import org.apache.solr.common.params.SolrParams; |
---|
32 | import org.slf4j.Logger; |
---|
33 | import org.slf4j.LoggerFactory; |
---|
34 | |
---|
35 | import eu.clarin.cmdi.vlo.CommonUtils; |
---|
36 | import eu.clarin.cmdi.vlo.FacetConstants; |
---|
37 | import eu.clarin.cmdi.vlo.LanguageCodeUtils; |
---|
38 | import eu.clarin.cmdi.vlo.config.DataRoot; |
---|
39 | import eu.clarin.cmdi.vlo.config.VloConfig; |
---|
40 | import eu.clarin.cmdi.vlo.config.XmlVloConfigFactory; |
---|
41 | |
---|
42 | /** |
---|
43 | * The main metadataImporter class. Also contains the main function. |
---|
44 | * |
---|
45 | * The metadataimporter reads all the config files and then, for each |
---|
46 | * metadatafile in each defined directory structure parses and imports them as |
---|
47 | * defined in the configuration. The startImport function starts the importing |
---|
48 | * and so on. |
---|
49 | */ |
---|
50 | public class MetadataImporter { |
---|
51 | |
---|
52 | /** |
---|
53 | * Defines which files to try and parse. In this case all files ending in |
---|
54 | * "xml" or "cmdi". |
---|
55 | */ |
---|
56 | private static final String[] VALID_CMDI_EXTENSIONS = new String[]{"xml", "cmdi"}; |
---|
57 | |
---|
58 | /** |
---|
59 | * Log log log log |
---|
60 | */ |
---|
61 | protected final static Logger LOG = LoggerFactory.getLogger(MetadataImporter.class); |
---|
62 | /** |
---|
63 | * Some place to store errors. |
---|
64 | */ |
---|
65 | private static Throwable serverError; |
---|
66 | /** |
---|
67 | * the solr server. |
---|
68 | */ |
---|
69 | private ConcurrentUpdateSolrServer solrServer; |
---|
70 | /** |
---|
71 | * Defines the post-processor associations. At import, for each facet value, |
---|
72 | * this map is checked and all postprocessors associated with the facet |
---|
73 | * _type_ are applied to the value before storing the new value in the solr |
---|
74 | * document. |
---|
75 | */ |
---|
76 | final static Map<String, PostProcessor> POST_PROCESSORS = new HashMap<String, PostProcessor>(); |
---|
77 | |
---|
78 | static { |
---|
79 | POST_PROCESSORS.put(FacetConstants.FIELD_ID, new IdPostProcessor()); |
---|
80 | POST_PROCESSORS.put(FacetConstants.FIELD_CONTINENT, new ContinentNamePostProcessor()); |
---|
81 | POST_PROCESSORS.put(FacetConstants.FIELD_COUNTRY, new CountryNamePostProcessor()); |
---|
82 | POST_PROCESSORS.put(FacetConstants.FIELD_LANGUAGE_CODE, new LanguageCodePostProcessor()); |
---|
83 | POST_PROCESSORS.put(FacetConstants.FIELD_LANGUAGE_NAME, new LanguageNamePostProcessor()); |
---|
84 | POST_PROCESSORS.put(FacetConstants.FIELD_AVAILABILITY, new AvailabilityPostProcessor()); |
---|
85 | POST_PROCESSORS.put(FacetConstants.FIELD_ORGANISATION, new OrganisationPostProcessor()); |
---|
86 | POST_PROCESSORS.put(FacetConstants.FIELD_TEMPORAL_COVERAGE, new TemporalCoveragePostProcessor()); |
---|
87 | POST_PROCESSORS.put(FacetConstants.FIELD_NATIONAL_PROJECT, new NationalProjectPostProcessor()); |
---|
88 | POST_PROCESSORS.put(FacetConstants.FIELD_CLARIN_PROFILE, new CMDIComponentProfileNamePostProcessor()); |
---|
89 | POST_PROCESSORS.put(FacetConstants.FIELD_RESOURCE_CLASS, new ResourceClassPostProcessor()); |
---|
90 | } |
---|
91 | |
---|
92 | /** |
---|
93 | * Constructor |
---|
94 | */ |
---|
95 | public MetadataImporter() {} |
---|
96 | |
---|
97 | public MetadataImporter(String clDatarootsList) { |
---|
98 | this.clDatarootsList = clDatarootsList; |
---|
99 | } |
---|
100 | |
---|
101 | /** |
---|
102 | * Contains MDSelflinks (usually). Just to know what we have already done. |
---|
103 | */ |
---|
104 | protected final Set<String> processedIds = new HashSet<String>(); |
---|
105 | /** |
---|
106 | * Some caching for solr documents (we are more efficient if we ram a whole |
---|
107 | * bunch to the solr server at once. |
---|
108 | */ |
---|
109 | protected List<SolrInputDocument> docs = new ArrayList<SolrInputDocument>(); |
---|
110 | |
---|
111 | // SOME STATS |
---|
112 | protected int nrOFDocumentsSend; |
---|
113 | protected int nrOfFilesAnalyzed = 0; |
---|
114 | protected int nrOfFilesWithoutId = 0; |
---|
115 | protected int nrOfIgnoredFiles = 0; |
---|
116 | protected int nrOfFilesWithError = 0; |
---|
117 | |
---|
118 | /** |
---|
119 | * Retrieve all files with VALID_CMDI_EXTENSIONS from all DataRoot entries |
---|
120 | * and starts processing for every single file |
---|
121 | * |
---|
122 | * @throws MalformedURLException |
---|
123 | */ |
---|
124 | void startImport() throws MalformedURLException { |
---|
125 | |
---|
126 | initSolrServer(); |
---|
127 | List<DataRoot> dataRoots = checkDataRoots(); |
---|
128 | |
---|
129 | dataRoots = filterDataRootsWithCLArgs(dataRoots); |
---|
130 | |
---|
131 | long start = System.currentTimeMillis(); |
---|
132 | try { |
---|
133 | // Delete the whole Solr db |
---|
134 | if (config.getDeleteAllFirst()) { |
---|
135 | LOG.info("Deleting original data..."); |
---|
136 | solrServer.deleteByQuery("*:*"); |
---|
137 | solrServer.commit(); |
---|
138 | LOG.info("Deleting original data done."); |
---|
139 | } |
---|
140 | |
---|
141 | // Import the specified data roots |
---|
142 | for (DataRoot dataRoot : dataRoots) { |
---|
143 | LOG.info("Start of processing: " + dataRoot.getOriginName()); |
---|
144 | if (dataRoot.deleteFirst()) { |
---|
145 | LOG.info("Deleting data for data provider: " + dataRoot.getOriginName()); |
---|
146 | solrServer.deleteByQuery(FacetConstants.FIELD_DATA_PROVIDER + ":" + ClientUtils.escapeQueryChars(dataRoot.getOriginName())); |
---|
147 | LOG.info("Deleting data of provider done."); |
---|
148 | } |
---|
149 | CMDIDataProcessor processor = new CMDIParserVTDXML(POST_PROCESSORS, false); |
---|
150 | List<List<File>> centreFilesList = getFilesFromDataRoot(dataRoot.getRootFile()); |
---|
151 | // import files from every endpoint |
---|
152 | for(List<File> centreFiles : centreFilesList) { |
---|
153 | LOG.info("Processing directory: {}", centreFiles.get(0).getParent()); |
---|
154 | for (File file : centreFiles) { |
---|
155 | if (config.getMaxFileSize() > 0 |
---|
156 | && file.length() > config.getMaxFileSize()) { |
---|
157 | LOG.info("Skipping " + file.getAbsolutePath() + " because it is too large."); |
---|
158 | } else { |
---|
159 | LOG.debug("PROCESSING FILE: {}", file.getAbsolutePath()); |
---|
160 | processCmdi(file, dataRoot, processor); |
---|
161 | } |
---|
162 | } |
---|
163 | if (!docs.isEmpty()) { |
---|
164 | sendDocs(); |
---|
165 | } |
---|
166 | solrServer.commit(); |
---|
167 | if(config.isProcessHierarchies()){ |
---|
168 | updateDocumentHierarchy(); |
---|
169 | } |
---|
170 | } |
---|
171 | LOG.info("End of processing: " + dataRoot.getOriginName()); |
---|
172 | } |
---|
173 | |
---|
174 | // delete outdated entries (based on maxDaysInSolr parameter) |
---|
175 | if (config.getMaxDaysInSolr() > 0 && config.getDeleteAllFirst() == false) { |
---|
176 | LOG.info("Deleting old files that were not seen for more than " + config.getMaxDaysInSolr() + " days..."); |
---|
177 | solrServer.deleteByQuery(FacetConstants.FIELD_LAST_SEEN + ":[* TO NOW-" + config.getMaxDaysInSolr() + "DAYS]"); |
---|
178 | LOG.info("Deleting old files done."); |
---|
179 | } |
---|
180 | } catch (SolrServerException e) { |
---|
181 | LOG.error("error updating files:\n", e); |
---|
182 | LOG.error("Also see vlo_solr server logs for more information"); |
---|
183 | } catch (IOException e) { |
---|
184 | LOG.error("error updating files:\n", e); |
---|
185 | } finally { |
---|
186 | try { |
---|
187 | if (solrServer != null) { |
---|
188 | solrServer.commit(); |
---|
189 | buildSuggesterIndex(); |
---|
190 | } |
---|
191 | } catch (SolrServerException e) { |
---|
192 | LOG.error("cannot commit:\n", e); |
---|
193 | } catch (IOException e) { |
---|
194 | LOG.error("cannot commit:\n", e); |
---|
195 | } |
---|
196 | } |
---|
197 | long took = (System.currentTimeMillis() - start) / 1000; |
---|
198 | LOG.info("Found " + nrOfFilesWithoutId + " file(s) without an id. (id is generated based on fileName but that may not be unique)"); |
---|
199 | LOG.info("Found " + nrOfFilesWithError + " file(s) with errors."); |
---|
200 | LOG.info("Found " + nrOfIgnoredFiles |
---|
201 | + " file(s) that where ignored (files without resources or any link to a search service or landing page are ignored)."); |
---|
202 | LOG.info("Update of " + nrOFDocumentsSend + " took " + took + " secs. Total nr of files analyzed " + nrOfFilesAnalyzed); |
---|
203 | solrServer.shutdown(); |
---|
204 | } |
---|
205 | |
---|
206 | /** |
---|
207 | * Check a List of DataRoots for existence of RootFile (typically parent |
---|
208 | * directory of metadata files) |
---|
209 | * |
---|
210 | * @return |
---|
211 | */ |
---|
212 | protected List<DataRoot> checkDataRoots() { |
---|
213 | List<DataRoot> dataRoots = config.getDataRoots(); |
---|
214 | List<DataRoot> existingDataRoots = new LinkedList<DataRoot>(); |
---|
215 | for (DataRoot dataRoot : dataRoots) { |
---|
216 | if (!dataRoot.getRootFile().exists()) { |
---|
217 | LOG.warn("Root file " + dataRoot.getRootFile() + " does not exist. It could be configuration error! Proceeding with next ..."); |
---|
218 | } else{ |
---|
219 | existingDataRoots.add(dataRoot); |
---|
220 | } |
---|
221 | |
---|
222 | } |
---|
223 | return existingDataRoots; |
---|
224 | } |
---|
225 | |
---|
226 | /** |
---|
227 | * if user specified which data roots should be imported, |
---|
228 | * list of existing data roots will be filtered with the list from user |
---|
229 | * |
---|
230 | * @param dataRoots complete list of DataRoots |
---|
231 | * @return list of DataRoots without DataRoots excluded by the user |
---|
232 | */ |
---|
233 | protected List<DataRoot> filterDataRootsWithCLArgs(List<DataRoot> dataRoots){ |
---|
234 | if(clDatarootsList == null) |
---|
235 | return dataRoots; |
---|
236 | |
---|
237 | |
---|
238 | LOG.info("Filtering configured data root files with command line arguments: \"" + clDatarootsList + "\"" ) ; |
---|
239 | |
---|
240 | LinkedList<File> fsDataRoots = new LinkedList<File>(); |
---|
241 | |
---|
242 | List<String> paths = Arrays.asList((clDatarootsList.split("\\s+"))); |
---|
243 | |
---|
244 | //Convert String paths to File objects for comparison |
---|
245 | for(String path: paths) |
---|
246 | fsDataRoots.add(new File(path)); |
---|
247 | |
---|
248 | List<DataRoot> filteredDataRoots = new LinkedList<DataRoot>(); |
---|
249 | try{ |
---|
250 | //filter data |
---|
251 | dr: for(DataRoot dataRoot: dataRoots){ |
---|
252 | for(File fsDataRoot: fsDataRoots){ |
---|
253 | if(fsDataRoot.getCanonicalPath().equals(dataRoot.getRootFile().getCanonicalPath())){ |
---|
254 | filteredDataRoots.add(dataRoot); |
---|
255 | fsDataRoots.remove(fsDataRoot); |
---|
256 | continue dr; |
---|
257 | } |
---|
258 | } |
---|
259 | LOG.info("Root file " + dataRoot.getRootFile() + " will be omitted from processing"); |
---|
260 | } |
---|
261 | }catch (IOException e){ |
---|
262 | filteredDataRoots = dataRoots; |
---|
263 | } |
---|
264 | |
---|
265 | |
---|
266 | return filteredDataRoots; |
---|
267 | } |
---|
268 | |
---|
269 | /** |
---|
270 | * Get all files with VALID_CMDI_EXTENSIONS if rootFile is a |
---|
271 | * directory that contains center directories or rootFile if it is a file |
---|
272 | * |
---|
273 | * @param rootFile |
---|
274 | * @return List with centre Lists of all contained CMDI files if rootFile is a |
---|
275 | * directory or rootFile if it is a File |
---|
276 | */ |
---|
277 | protected List<List<File>> getFilesFromDataRoot(File rootFile) { |
---|
278 | List<List<File>> result = new ArrayList<List<File>>(); |
---|
279 | if(rootFile.isFile()) { |
---|
280 | List<File> singleFileList = new ArrayList<File>(); |
---|
281 | singleFileList.add(rootFile); |
---|
282 | result.add(singleFileList); |
---|
283 | } else { |
---|
284 | File[] centerDirs = rootFile.listFiles(); |
---|
285 | for(File centerDir : centerDirs) { |
---|
286 | List<File> centerFileList = new ArrayList<File>(); |
---|
287 | if(centerDir.isDirectory()) { |
---|
288 | centerFileList.addAll(FileUtils.listFiles(centerDir, VALID_CMDI_EXTENSIONS, true)); |
---|
289 | } |
---|
290 | |
---|
291 | if(!centerFileList.isEmpty()) |
---|
292 | result.add(centerFileList); |
---|
293 | } |
---|
294 | } |
---|
295 | return result; |
---|
296 | } |
---|
297 | |
---|
298 | /** |
---|
299 | * Create an interface to the SOLR server. |
---|
300 | * |
---|
301 | * After the interface has been created the importer can send documents to |
---|
302 | * the server. Sending documents involves a queue. The importer adds |
---|
303 | * documents to a queue, and dedicated threads will empty it, and |
---|
304 | * effectively store store the documents. |
---|
305 | * |
---|
306 | * @throws MalformedURLException |
---|
307 | */ |
---|
308 | protected void initSolrServer() throws MalformedURLException { |
---|
309 | String solrUrl = config.getSolrUrl(); |
---|
310 | LOG.info("Initializing Solr Server on " + solrUrl); |
---|
311 | |
---|
312 | /* Specify the number of documents in the queue that will trigger the |
---|
313 | * threads, two of them, emptying it. |
---|
314 | */ |
---|
315 | solrServer = new ConcurrentUpdateSolrServer(solrUrl, |
---|
316 | config.getMinDocsInSolrQueue(), 2) { |
---|
317 | /* |
---|
318 | * Let the super class method handle exceptions. Make the |
---|
319 | * exception available to the importer in the form of the |
---|
320 | * serverError variable. |
---|
321 | */ |
---|
322 | @Override |
---|
323 | public void handleError(Throwable exception) { |
---|
324 | super.handleError(exception); |
---|
325 | serverError = exception; |
---|
326 | } |
---|
327 | }; |
---|
328 | } |
---|
329 | |
---|
330 | /** |
---|
331 | * Process single CMDI file with CMDIDataProcessor |
---|
332 | * |
---|
333 | * @param file CMDI input file |
---|
334 | * @param dataOrigin |
---|
335 | * @param processor |
---|
336 | * @throws SolrServerException |
---|
337 | * @throws IOException |
---|
338 | */ |
---|
339 | protected void processCmdi(File file, DataRoot dataOrigin, CMDIDataProcessor processor) throws SolrServerException, IOException { |
---|
340 | nrOfFilesAnalyzed++; |
---|
341 | CMDIData cmdiData = null; |
---|
342 | try { |
---|
343 | cmdiData = processor.process(file); |
---|
344 | if (!idOk(cmdiData.getId())) { |
---|
345 | cmdiData.setId(dataOrigin.getOriginName() + "/" + file.getName()); //No id found in the metadata file so making one up based on the file name. Not quaranteed to be unique, but we have to set something. |
---|
346 | nrOfFilesWithoutId++; |
---|
347 | } |
---|
348 | } catch (Exception e) { |
---|
349 | LOG.error("error in file: {}", file, e); |
---|
350 | nrOfFilesWithError++; |
---|
351 | } |
---|
352 | if (cmdiData != null) { |
---|
353 | if (processedIds.add(cmdiData.getId())) { |
---|
354 | SolrInputDocument solrDocument = cmdiData.getSolrDocument(); |
---|
355 | if (solrDocument != null) { |
---|
356 | if (!cmdiData.getDataResources().isEmpty() || !cmdiData.getLandingPageResources().isEmpty() |
---|
357 | || !cmdiData.getSearchResources().isEmpty() || !cmdiData.getSearchPageResources().isEmpty() |
---|
358 | || cmdiData.getMetadataResources().isEmpty()) { |
---|
359 | // We only add metadata files that have |
---|
360 | // 1) data resources or |
---|
361 | // 2) a landing page or |
---|
362 | // 3) a search service (like SRU/CQL) or |
---|
363 | // 4) a search page or |
---|
364 | // 5) that have none of the above but also lack any metadata links (e.g. olac files describing a corpus with a link to the original archive). |
---|
365 | // Other files will have only metadata resources and are considered 'collection' metadata files they |
---|
366 | // are usually not very interesting (think imdi corpus files) and will not be included. |
---|
367 | updateDocument(solrDocument, cmdiData, file, dataOrigin); |
---|
368 | if(ResourceStructureGraph.getVertex(cmdiData.getId()) != null) |
---|
369 | ResourceStructureGraph.getVertex(cmdiData.getId()).setWasImported(true); |
---|
370 | } else { |
---|
371 | nrOfIgnoredFiles++; |
---|
372 | } |
---|
373 | } |
---|
374 | } else { |
---|
375 | LOG.warn("Skipping {}, already processed id: {}", file, cmdiData.getId()); |
---|
376 | } |
---|
377 | } |
---|
378 | } |
---|
379 | |
---|
380 | /** |
---|
381 | * Check id for validness |
---|
382 | * |
---|
383 | * @param id |
---|
384 | * @return true if id is acceptable, false otherwise |
---|
385 | */ |
---|
386 | protected boolean idOk(String id) { |
---|
387 | return id != null && !id.isEmpty(); |
---|
388 | } |
---|
389 | |
---|
390 | /** |
---|
391 | * Adds some additional information from DataRoot to solrDocument, add |
---|
392 | * solrDocument to document list, submits list to SolrServer every 1000 |
---|
393 | * files |
---|
394 | * |
---|
395 | * @param solrDocument |
---|
396 | * @param cmdiData |
---|
397 | * @param file |
---|
398 | * @param dataOrigin |
---|
399 | * @throws SolrServerException |
---|
400 | * @throws IOException |
---|
401 | */ |
---|
402 | protected void updateDocument(SolrInputDocument solrDocument, CMDIData cmdiData, File file, DataRoot dataOrigin) throws SolrServerException, |
---|
403 | IOException { |
---|
404 | if (!solrDocument.containsKey(FacetConstants.FIELD_COLLECTION)) { |
---|
405 | solrDocument.addField(FacetConstants.FIELD_COLLECTION, dataOrigin.getOriginName()); |
---|
406 | } |
---|
407 | solrDocument.addField(FacetConstants.FIELD_DATA_PROVIDER, dataOrigin.getOriginName()); |
---|
408 | solrDocument.addField(FacetConstants.FIELD_ID, cmdiData.getId()); |
---|
409 | solrDocument.addField(FacetConstants.FIELD_FILENAME, file.getAbsolutePath()); |
---|
410 | |
---|
411 | String metadataSourceUrl = dataOrigin.getPrefix(); |
---|
412 | metadataSourceUrl += file.getAbsolutePath().substring(dataOrigin.getToStrip().length()); |
---|
413 | |
---|
414 | solrDocument.addField(FacetConstants.FIELD_COMPLETE_METADATA, metadataSourceUrl); |
---|
415 | |
---|
416 | // add SearchServices (should be CQL endpoint) |
---|
417 | for (Resource resource : cmdiData.getSearchResources()) { |
---|
418 | solrDocument.addField(FacetConstants.FIELD_SEARCH_SERVICE, resource.getResourceName()); |
---|
419 | } |
---|
420 | |
---|
421 | // add landing page resource |
---|
422 | for (Resource resource : cmdiData.getLandingPageResources()) { |
---|
423 | solrDocument.addField(FacetConstants.FIELD_LANDINGPAGE, resource.getResourceName()); |
---|
424 | } |
---|
425 | |
---|
426 | // add search page resource |
---|
427 | for (Resource resource : cmdiData.getSearchPageResources()) { |
---|
428 | solrDocument.addField(FacetConstants.FIELD_SEARCHPAGE, resource.getResourceName()); |
---|
429 | } |
---|
430 | |
---|
431 | // add timestamp |
---|
432 | Date dt = new Date(); |
---|
433 | SimpleDateFormat df = new SimpleDateFormat("yyyy-MM-dd'T'HH:mm:ss'Z'"); |
---|
434 | solrDocument.addField(FacetConstants.FIELD_LAST_SEEN, df.format(dt)); |
---|
435 | |
---|
436 | // add resource proxys |
---|
437 | addResourceData(solrDocument, cmdiData); |
---|
438 | |
---|
439 | LOG.debug("Adding document for submission to SOLR: {}", file); |
---|
440 | docs.add(solrDocument); |
---|
441 | if (docs.size() == config.getMaxDocsInList()) { |
---|
442 | sendDocs(); |
---|
443 | } |
---|
444 | } |
---|
445 | |
---|
446 | /** |
---|
447 | * Adds two fields FIELD_FORMAT and FIELD_RESOURCE. The Type can be |
---|
448 | * specified in the "ResourceType" element of an imdi file or possibly |
---|
449 | * overwritten by some more specific xpath (as in the LRT cmdi files). So if |
---|
450 | * a type is overwritten and already in the solrDocument we take that type. |
---|
451 | * |
---|
452 | * @param solrDocument |
---|
453 | * @param cmdiData |
---|
454 | */ |
---|
455 | protected void addResourceData(SolrInputDocument solrDocument, CMDIData cmdiData) { |
---|
456 | List<Object> fieldValues = solrDocument.containsKey(FacetConstants.FIELD_FORMAT) ? new ArrayList<Object>(solrDocument |
---|
457 | .getFieldValues(FacetConstants.FIELD_FORMAT)) : null; |
---|
458 | solrDocument.removeField(FacetConstants.FIELD_FORMAT); //Remove old values they might be overwritten. |
---|
459 | List<Resource> resources = cmdiData.getDataResources(); |
---|
460 | for (int i = 0; i < resources.size(); i++) { |
---|
461 | Resource resource = resources.get(i); |
---|
462 | String mimeType = resource.getMimeType(); |
---|
463 | if (mimeType == null) { |
---|
464 | if (fieldValues != null && i < fieldValues.size()) { |
---|
465 | mimeType = CommonUtils.normalizeMimeType(fieldValues.get(i).toString()); |
---|
466 | } else { |
---|
467 | mimeType = CommonUtils.normalizeMimeType(""); |
---|
468 | } |
---|
469 | } |
---|
470 | |
---|
471 | FormatPostProcessor processor = new FormatPostProcessor(); |
---|
472 | mimeType = processor.process(mimeType).get(0); |
---|
473 | |
---|
474 | // TODO check should probably be moved into Solr (by using some minimum length filter) |
---|
475 | if (!mimeType.equals("")) { |
---|
476 | solrDocument.addField(FacetConstants.FIELD_FORMAT, mimeType); |
---|
477 | } |
---|
478 | solrDocument.addField(FacetConstants.FIELD_RESOURCE, mimeType + FacetConstants.FIELD_RESOURCE_SPLIT_CHAR |
---|
479 | + resource.getResourceName()); |
---|
480 | } |
---|
481 | solrDocument.addField(FacetConstants.FIELD_RESOURCE_COUNT, resources.size()); |
---|
482 | } |
---|
483 | |
---|
484 | /** |
---|
485 | * Send current list of SolrImputDocuments to SolrServer and clears list |
---|
486 | * afterwards |
---|
487 | * |
---|
488 | * @throws SolrServerException |
---|
489 | * @throws IOException |
---|
490 | */ |
---|
491 | protected void sendDocs() throws SolrServerException, IOException { |
---|
492 | LOG.info("Sending " + docs.size() + " docs to solr server. Total number of docs updated till now: " + nrOFDocumentsSend); |
---|
493 | nrOFDocumentsSend += docs.size(); |
---|
494 | solrServer.add(docs); |
---|
495 | if (serverError != null) { |
---|
496 | throw new SolrServerException(serverError); |
---|
497 | } |
---|
498 | docs = new ArrayList<SolrInputDocument>(); |
---|
499 | } |
---|
500 | |
---|
501 | /** |
---|
502 | * Builds suggester index for autocompletion |
---|
503 | * |
---|
504 | * @throws SolrServerException |
---|
505 | * @throws MalformedURLException |
---|
506 | */ |
---|
507 | private void buildSuggesterIndex() throws SolrServerException, MalformedURLException { |
---|
508 | LOG.info("Building index for autocompletion."); |
---|
509 | HashMap<String, String> paramMap = new HashMap<String, String>(); |
---|
510 | paramMap.put("qt", "/suggest"); |
---|
511 | paramMap.put("spellcheck.build", "true"); |
---|
512 | SolrParams params = new MapSolrParams(paramMap); |
---|
513 | solrServer.query(params); |
---|
514 | } |
---|
515 | |
---|
516 | /** |
---|
517 | * Updates documents in Solr with their hierarchy weight and lists of related resources (hasPart & isPartOf) |
---|
518 | * @throws SolrServerException |
---|
519 | * @throws MalformedURLException |
---|
520 | */ |
---|
521 | private void updateDocumentHierarchy() throws SolrServerException, MalformedURLException, IOException { |
---|
522 | LOG.info(ResourceStructureGraph.printStatistics(0)); |
---|
523 | Boolean updatedDocs = false; |
---|
524 | List<SolrInputDocument> updateDocs = new ArrayList<SolrInputDocument>(); |
---|
525 | Iterator<CmdiVertex> vertexIter = ResourceStructureGraph.getFoundVertices().iterator(); |
---|
526 | while(vertexIter.hasNext()) { |
---|
527 | CmdiVertex vertex = vertexIter.next(); |
---|
528 | List<String> incomingVertexNames = ResourceStructureGraph.getIncomingVertexNames(vertex); |
---|
529 | List<String> outgoingVertexNames = ResourceStructureGraph.getOutgoingVertexNames(vertex); |
---|
530 | |
---|
531 | SolrQuery query; |
---|
532 | // update vertex if changes are necessary (necessary if non-default weight or edges to other resources) |
---|
533 | if(vertex.getHierarchyWeight() != 0 || !incomingVertexNames.isEmpty() || !outgoingVertexNames.isEmpty()) { |
---|
534 | updatedDocs = true; |
---|
535 | |
---|
536 | // get document |
---|
537 | query = new SolrQuery(); |
---|
538 | query.setRequestHandler(FacetConstants.SOLR_REQUEST_HANDLER_FAST); |
---|
539 | query.set("q", FacetConstants.FIELD_ID+":"+vertex.getId()); |
---|
540 | SolrDocumentList response = solrServer.query(query).getResults(); |
---|
541 | |
---|
542 | // empty result set? may be the case if CMDI file was rejected due to missing ResourceProxys in {@link #processCmdi(File, DataRoot, CMDIDataProcessor) processCmdi} |
---|
543 | if(response.size() == 0) { |
---|
544 | LOG.debug("Doc "+vertex.getId()+" not found while updating document hierarchy information"); |
---|
545 | continue; |
---|
546 | } |
---|
547 | SolrInputDocument doc = ClientUtils.toSolrInputDocument(response.get(0)); |
---|
548 | |
---|
549 | if(vertex.getHierarchyWeight() != 0) { |
---|
550 | doc.setField(FacetConstants.FIELD_HIERARCHY_WEIGHT, Math.abs(vertex.getHierarchyWeight())); |
---|
551 | } |
---|
552 | |
---|
553 | // remove vertices that were not imported |
---|
554 | Iterator<String> incomingVertexIter = incomingVertexNames.iterator(); |
---|
555 | while(incomingVertexIter.hasNext()) { |
---|
556 | String vertexId = incomingVertexIter.next(); |
---|
557 | if(ResourceStructureGraph.getVertex(vertexId) == null || !ResourceStructureGraph.getVertex(vertexId).getWasImported()) |
---|
558 | incomingVertexIter.remove(); |
---|
559 | } |
---|
560 | Iterator<String> outgoingVertexIter = outgoingVertexNames.iterator(); |
---|
561 | while(outgoingVertexIter.hasNext()) { |
---|
562 | String vertexId = outgoingVertexIter.next(); |
---|
563 | if(ResourceStructureGraph.getVertex(vertexId) == null || !ResourceStructureGraph.getVertex(vertexId).getWasImported()) |
---|
564 | outgoingVertexIter.remove(); |
---|
565 | } |
---|
566 | |
---|
567 | if(!incomingVertexNames.isEmpty()) { |
---|
568 | doc.setField(FacetConstants.FIELD_HAS_PART, incomingVertexNames); |
---|
569 | doc.setField(FacetConstants.FIELD_HAS_PART_COUNT, incomingVertexNames.size()); |
---|
570 | } |
---|
571 | |
---|
572 | if(!outgoingVertexNames.isEmpty()) { |
---|
573 | doc.setField(FacetConstants.FIELD_IS_PART_OF, outgoingVertexNames); |
---|
574 | } |
---|
575 | updateDocs.add(doc); |
---|
576 | } |
---|
577 | |
---|
578 | if (updateDocs.size() == config.getMaxDocsInList()) { |
---|
579 | solrServer.add(updateDocs); |
---|
580 | if (serverError != null) { |
---|
581 | throw new SolrServerException(serverError); |
---|
582 | } |
---|
583 | updateDocs = new ArrayList<SolrInputDocument>(); |
---|
584 | } |
---|
585 | } |
---|
586 | if(!updateDocs.isEmpty()) { |
---|
587 | solrServer.add(updateDocs); |
---|
588 | if (serverError != null) { |
---|
589 | throw new SolrServerException(serverError); |
---|
590 | } |
---|
591 | } |
---|
592 | |
---|
593 | if(updatedDocs) { |
---|
594 | solrServer.commit(); |
---|
595 | } |
---|
596 | |
---|
597 | ResourceStructureGraph.clearResourceGraph(); |
---|
598 | } |
---|
599 | |
---|
600 | public static VloConfig config; |
---|
601 | |
---|
602 | public static LanguageCodeUtils languageCodeUtils; |
---|
603 | |
---|
604 | //data roots passed from command line |
---|
605 | private String clDatarootsList = null; |
---|
606 | |
---|
607 | /** |
---|
608 | * @param args |
---|
609 | * @throws MalformedURLException |
---|
610 | * @throws IOException |
---|
611 | */ |
---|
612 | public static void main(String[] args) throws MalformedURLException, IOException { |
---|
613 | |
---|
614 | // path to the configuration file |
---|
615 | String configFile = null; |
---|
616 | |
---|
617 | // use the Apache cli framework for getting command line parameters |
---|
618 | Options options = new Options(); |
---|
619 | |
---|
620 | // Data root list passed from command line with -l option |
---|
621 | String cldrList = null; |
---|
622 | |
---|
623 | /** |
---|
624 | * Add a "c" option, the option indicating the specification of an XML |
---|
625 | * configuration file |
---|
626 | * |
---|
627 | * "l" option - to specify which data roots (from config file) to import |
---|
628 | * imports all by default |
---|
629 | */ |
---|
630 | options.addOption("c", true, "-c <file> : use parameters specified in <file>"); |
---|
631 | options.addOption("l", true, "-l <dataroot> [ ' ' <dataroot> ]* : space separated list of dataroots to be processed.\n" |
---|
632 | + "If dataroot is not specified in config file it will be ignored."); |
---|
633 | options.getOption("l").setOptionalArg(true); |
---|
634 | |
---|
635 | CommandLineParser parser = new PosixParser(); |
---|
636 | |
---|
637 | try { |
---|
638 | // parse the command line arguments |
---|
639 | CommandLine cmd = parser.parse(options, args); |
---|
640 | if (cmd.hasOption("c")) { |
---|
641 | |
---|
642 | // the "c" option was specified, now get its value |
---|
643 | configFile = cmd.getOptionValue("c"); |
---|
644 | } |
---|
645 | |
---|
646 | if(cmd.hasOption("l")){ |
---|
647 | cldrList = cmd.getOptionValue("l"); |
---|
648 | } |
---|
649 | |
---|
650 | } catch (org.apache.commons.cli.ParseException ex) { |
---|
651 | |
---|
652 | /** |
---|
653 | * Caught an exception caused by command line parsing. Try to get |
---|
654 | * the name of the configuration file by querying the system |
---|
655 | * property. |
---|
656 | */ |
---|
657 | String message = "Command line parsing failed. " + ex.getMessage(); |
---|
658 | LOG.error(message); |
---|
659 | System.err.println(message); |
---|
660 | } |
---|
661 | |
---|
662 | if (configFile == null) { |
---|
663 | |
---|
664 | String message; |
---|
665 | |
---|
666 | message = "Could not get config file name via the command line, trying the system properties."; |
---|
667 | LOG.info(message); |
---|
668 | |
---|
669 | String key; |
---|
670 | |
---|
671 | key = "configFile"; |
---|
672 | configFile = System.getProperty(key); |
---|
673 | } |
---|
674 | |
---|
675 | if (configFile == null) { |
---|
676 | |
---|
677 | String message; |
---|
678 | |
---|
679 | message = "Could not get filename as system property either - stopping."; |
---|
680 | LOG.error(message); |
---|
681 | } else { |
---|
682 | // read the configuration from the externally supplied file |
---|
683 | final URL configUrl; |
---|
684 | if (configFile.startsWith("file:")) { |
---|
685 | configUrl = new URL(configFile); |
---|
686 | } else { |
---|
687 | configUrl = new File(configFile).toURI().toURL(); |
---|
688 | } |
---|
689 | System.out.println("Reading configuration from " + configUrl.toString()); |
---|
690 | LOG.info("Reading configuration from " + configUrl.toString()); |
---|
691 | final XmlVloConfigFactory configFactory = new XmlVloConfigFactory(configUrl); |
---|
692 | MetadataImporter.config = configFactory.newConfig(); |
---|
693 | MetadataImporter.languageCodeUtils = new LanguageCodeUtils(MetadataImporter.config); |
---|
694 | |
---|
695 | // optionally, modify the configuration here |
---|
696 | // create and start the importer |
---|
697 | MetadataImporter importer = new MetadataImporter(cldrList); |
---|
698 | importer.startImport(); |
---|
699 | |
---|
700 | // finished importing |
---|
701 | if (MetadataImporter.config.printMapping()) { |
---|
702 | File file = new File("xsdMapping.txt"); |
---|
703 | FacetMappingFactory.printMapping(file); |
---|
704 | LOG.info("Printed facetMapping in " + file); |
---|
705 | } |
---|
706 | } |
---|
707 | } |
---|
708 | } |
---|