source: other-projects/hathitrust/wcsa/extracted-features-solr/trunk/solr-ingest/src/main/java/org/hathitrust/extractedfeatures/PerVolumeJSON.java@ 32101

Last change on this file since 32101 was 31784, checked in by davidb, 7 years ago

Output to highlight skipping per-page indexing
  • Property svn:executable set to *
File size: 9.4 KB
Line 
1package org.hathitrust.extractedfeatures;
2
3import java.io.IOException;
4import java.util.ArrayList;
5import java.util.HashMap;
6import java.util.Iterator;
7
8import org.apache.hadoop.io.Text;
9import org.apache.spark.api.java.function.FlatMapFunction;
10import org.apache.spark.api.java.function.Function;
11import org.apache.spark.api.java.function.VoidFunction;
12import org.apache.spark.util.DoubleAccumulator;
13import org.json.JSONArray;
14import org.json.JSONObject;
15
16/*
17class PagedJSON implements Function<String, Boolean> {
18
19 private static final long serialVersionUID = 1L;
20
21 public Boolean call(String s) { return s.contains("a"); }
22}
23 */
24
25
26//public class PerVolumeJSON implements VoidFunction<String>
27public class PerVolumeJSON implements Function<Text,Integer>
28{
29 private static final long serialVersionUID = 1L;
30 protected String _input_dir;
31 protected String _whitelist_filename;
32 protected String _langmap_directory;
33
34 protected final ArrayList<String> _solr_endpoints;
35 protected final int _solr_endpoints_len;
36
37 //protected String _solr_url;
38 protected String _output_dir;
39
40 protected int _verbosity;
41
42 protected WhitelistBloomFilter _whitelist_bloomfilter;
43 protected UniversalPOSLangMap _universal_langmap;
44
45 boolean _icu_tokenize;
46 boolean _strict_file_io;
47
48 public PerVolumeJSON(String input_dir, String whitelist_filename, String langmap_directory,
49 ArrayList<String> solr_endpoints, String output_dir, int verbosity,
50 boolean icu_tokenize, boolean strict_file_io)
51 {
52 System.out.println("*** PerVolumeJSON Constructor, langmap_directory = " + langmap_directory);
53
54 _input_dir = input_dir;
55 _whitelist_filename = whitelist_filename;
56 _langmap_directory = langmap_directory;
57
58 _solr_endpoints = solr_endpoints;
59 _solr_endpoints_len = solr_endpoints.size();
60
61 //_solr_url = solr_url;
62 _output_dir = output_dir;
63 _verbosity = verbosity;
64
65 _icu_tokenize = icu_tokenize;
66 _strict_file_io = strict_file_io;
67
68 _whitelist_bloomfilter = null;
69 _universal_langmap = null;
70 }
71
72
73 public Integer call(Text json_text) throws IOException
74
75 {
76 if (_whitelist_filename != null) {
77
78 synchronized (_whitelist_filename) {
79 if (_whitelist_bloomfilter == null) {
80
81 _whitelist_bloomfilter = new WhitelistBloomFilter(_whitelist_filename,true);
82 }
83 }
84 }
85
86 if (_langmap_directory != null) {
87
88 synchronized (_langmap_directory) {
89 if (_universal_langmap == null) {
90 _universal_langmap = new UniversalPOSLangMap(_langmap_directory);
91 }
92 }
93 }
94
95 int ef_num_pages = 0;
96
97 String solr_url = null;
98 if (_solr_endpoints_len > 0) {
99 int random_choice = (int)(_solr_endpoints_len * Math.random());
100 solr_url = _solr_endpoints.get(random_choice);
101 }
102
103 try {
104
105
106 JSONObject extracted_feature_record = new JSONObject(json_text.toString());
107
108 if (extracted_feature_record != null) {
109 String volume_id = extracted_feature_record.getString("id");
110
111 JSONObject ef_metadata = extracted_feature_record.getJSONObject("metadata");
112 //String title= ef_metadata.getString("title");
113
114 //
115 // Top-level metadata Solr doc
116 //
117 JSONObject solr_add_metadata_doc_json = SolrDocJSON.generateToplevelMetadataSolrDocJSON(volume_id,ef_metadata);
118 if (solr_add_metadata_doc_json != null) {
119
120 if ((_verbosity >=2)) {
121 System.out.println("==================");
122 System.out.println("Metadata JSON: " + solr_add_metadata_doc_json.toString());
123 System.out.println("==================");
124 }
125
126 if (solr_url != null) {
127
128 if ((_verbosity >=2) ) {
129 System.out.println("==================");
130 System.out.println("Posting to: " + solr_url);
131 System.out.println("==================");
132 }
133 SolrDocJSON.postSolrDoc(solr_url, solr_add_metadata_doc_json, volume_id, "top-level-metadata");
134 }
135 }
136
137 //
138 // Now move on to POS extracted features per-page
139 //
140 boolean index_pages = true;
141 if (index_pages) {
142
143 JSONObject ef_features = extracted_feature_record.getJSONObject("features");
144
145 int ef_page_count = ef_features.getInt("pageCount");
146
147 if (_verbosity >= 1) {
148 System.out.println("Processing: " + volume_id);
149 System.out.println(" pageCount = " + ef_page_count);
150 }
151
152 JSONArray ef_pages = ef_features.getJSONArray("pages");
153 ef_num_pages = ef_pages.length();
154
155
156 for (int i = 0; i < ef_page_count; i++) {
157 String formatted_i = String.format("page-%06d", i);
158 String page_id = volume_id + "." + formatted_i;
159
160 if (_verbosity >= 2) {
161 System.out.println(" Page: " + page_id);
162 }
163
164
165 JSONObject ef_page = ef_pages.getJSONObject(i);
166
167 if (ef_page != null) {
168 // Convert to Solr add form
169 JSONObject solr_add_doc_json
170 = SolrDocJSON.generateSolrDocJSON(volume_id, page_id,
171 ef_metadata, ef_page,
172 _whitelist_bloomfilter, _universal_langmap, _icu_tokenize);
173
174
175 if ((_verbosity >=2) && (i==20)) {
176 System.out.println("==================");
177 System.out.println("Sample output Solr add JSON [page 20]: " + solr_add_doc_json.toString());
178 System.out.println("==================");
179 }
180
181
182 if (solr_url != null) {
183 SolrDocJSON.postSolrDoc(solr_url, solr_add_doc_json,
184 volume_id, page_id);
185 }
186 }
187 else {
188 System.err.println("Skipping: " + page_id);
189 }
190
191 }
192 }
193 else {
194 System.err.println("Skipping per-page POS text indexing");
195 }
196
197 }
198 }
199 catch (Exception e) {
200 if (_strict_file_io) {
201 throw e;
202 }
203 else {
204 e.printStackTrace();
205 }
206 }
207
208 return ef_num_pages;
209
210 }
211
212 /*
213 //public void call(String json_file_in) throws IOException
214 public Integer call(String json_file_in) throws IOException
215
216 {
217 if ((_whitelist_filename != null) && (_whitelist_bloomfilter == null)) {
218 _whitelist_bloomfilter = new WhitelistBloomFilter(_whitelist_filename,true);
219 }
220
221 int ef_num_pages = 0;
222
223 ArrayList<String> ids = new ArrayList<String>(); // want it to be non-null so can return valid iterator
224
225 String full_json_file_in = _input_dir + "/" + json_file_in;
226 JSONObject extracted_feature_record = JSONClusterFileIO.readJSONFile(full_json_file_in);
227
228 if (extracted_feature_record != null) {
229 String volume_id = extracted_feature_record.getString("id");
230
231 //JSONObject ef_metadata = extracted_feature_record.getJSONObject("metadata");
232 //String title= ef_metadata.getString("title");
233
234 JSONObject ef_features = extracted_feature_record.getJSONObject("features");
235
236 int ef_page_count = ef_features.getInt("pageCount");
237
238 if (_verbosity >= 1) {
239 System.out.println("Processing: " + json_file_in);
240 System.out.println(" pageCount = " + ef_page_count);
241 }
242
243 JSONArray ef_pages = ef_features.getJSONArray("pages");
244 ef_num_pages = ef_pages.length();
245
246 // Make directory for page-level JSON output
247 String json_dir = ClusterFileIO.removeSuffix(json_file_in,".json.bz2");
248 String page_json_dir = json_dir + "/pages";
249
250 if (_output_dir != null) {
251 ClusterFileIO.createDirectoryAll(_output_dir + "/" + page_json_dir);
252 }
253
254 ids = new ArrayList<String>(ef_num_pages);
255 for (int i = 0; i < ef_page_count; i++) {
256 String formatted_i = String.format("page-%06d", i);
257 String page_id = volume_id + "." + formatted_i;
258
259 if (_verbosity >= 2) {
260 System.out.println(" Page: " + page_id);
261 }
262
263 String output_json_bz2 = page_json_dir +"/" + formatted_i + ".json.bz2";
264 ids.add(page_id);
265
266 if (_verbosity >=2) {
267 if (i==0) {
268 System.out.println("Sample output JSON page file [i=0]: " + output_json_bz2);
269 }
270 }
271 JSONObject ef_page = ef_pages.getJSONObject(i);
272
273 if (ef_page != null) {
274 // Convert to Solr add form
275 JSONObject solr_add_doc_json
276 = SolrDocJSON.generateSolrDocJSON(volume_id, page_id, ef_page, _whitelist_bloomfilter, _icu_tokenize);
277
278
279 if ((_verbosity >=2) && (i==20)) {
280 System.out.println("==================");
281 System.out.println("Sample output Solr add JSON [page 20]: " + solr_add_doc_json.toString());
282 System.out.println("==================");
283 }
284
285
286 if (_solr_url != null) {
287 if ((_verbosity >=2) && (i==20)) {
288 System.out.println("==================");
289 System.out.println("Posting to: " + _solr_url);
290 System.out.println("==================");
291 }
292 SolrDocJSON.postSolrDoc(_solr_url, solr_add_doc_json);
293 }
294
295 if (_output_dir != null) {
296 if ((_verbosity >=2) && (i==20)) {
297 System.out.println("==================");
298 System.out.println("Saving to: " + _output_dir);
299 System.out.println("==================");
300 }
301 SolrDocJSON.saveSolrDoc(solr_add_doc_json, _output_dir + "/" + output_json_bz2);
302 }
303 }
304 else {
305 System.err.println("Skipping: " + page_id);
306 }
307
308 }
309 }
310 else {
311 // File did not exist, or could not be parsed
312 String mess = "Failed to read in bzipped JSON file '" + full_json_file_in + "'";
313 if (_strict_file_io) {
314 throw new IOException(mess);
315 }
316 else {
317 System.err.println("Warning: " + mess);
318 System.out.println("Warning: " + mess);
319 }
320 }
321
322 return ef_num_pages;
323
324 }
325 */
326}
327
Note: See TracBrowser for help on using the repository browser.