1 | package org.hathitrust.extractedfeatures;
|
---|
2 |
|
---|
3 | import java.io.IOException;
|
---|
4 | import java.io.Reader;
|
---|
5 | import java.io.StringReader;
|
---|
6 | import java.util.ArrayList;
|
---|
7 | import java.util.Iterator;
|
---|
8 | import java.util.List;
|
---|
9 | import java.util.Set;
|
---|
10 |
|
---|
11 | import org.apache.lucene.analysis.TokenStream;
|
---|
12 | import org.apache.lucene.analysis.core.LowerCaseFilter;
|
---|
13 | import org.apache.lucene.analysis.icu.segmentation.ICUTokenizer;
|
---|
14 | import org.apache.lucene.analysis.tokenattributes.CharTermAttribute;
|
---|
15 | import org.apache.spark.api.java.function.Function;
|
---|
16 | import org.apache.spark.util.DoubleAccumulator;
|
---|
17 | import org.bson.Document;
|
---|
18 | import org.json.JSONArray;
|
---|
19 | import org.json.JSONObject;
|
---|
20 |
|
---|
21 | import com.mongodb.MongoClient;
|
---|
22 | import com.mongodb.MongoClientURI;
|
---|
23 | import com.mongodb.MongoException;
|
---|
24 | import com.mongodb.client.MongoCollection;
|
---|
25 | import com.mongodb.client.MongoDatabase;
|
---|
26 |
|
---|
27 |
|
---|
28 | class PerVolumeMongoDBDocumentsMap implements Function<String, Integer>
|
---|
29 | {
|
---|
30 | private static final long serialVersionUID = 1L;
|
---|
31 |
|
---|
32 | protected String _input_dir;
|
---|
33 | protected int _verbosity;
|
---|
34 |
|
---|
35 | protected DoubleAccumulator _progress_accum;
|
---|
36 | protected double _progress_step;
|
---|
37 |
|
---|
38 | boolean _strict_file_io;
|
---|
39 |
|
---|
40 | public PerVolumeMongoDBDocumentsMap(String input_dir, int verbosity,
|
---|
41 | DoubleAccumulator progress_accum, double progress_step,
|
---|
42 | boolean strict_file_io)
|
---|
43 | {
|
---|
44 | _input_dir = input_dir;
|
---|
45 | _verbosity = verbosity;
|
---|
46 |
|
---|
47 | _progress_accum = progress_accum;
|
---|
48 | _progress_step = progress_step;
|
---|
49 |
|
---|
50 | _strict_file_io = strict_file_io;
|
---|
51 | }
|
---|
52 |
|
---|
53 | protected void fixup_section(Document ef_count)
|
---|
54 | {
|
---|
55 |
|
---|
56 | Set<String> key_set = ef_count.keySet();
|
---|
57 | String[] key_array = key_set.toArray(new String[key_set.size()]);
|
---|
58 |
|
---|
59 |
|
---|
60 | //Set<String> key_set = ef_count.keySet();
|
---|
61 | //for (String key : key_set) {
|
---|
62 |
|
---|
63 | //Iterator<String> key_iterator = ef_count.keySet().iterator();
|
---|
64 | //while (key_iterator.hasNext()) {
|
---|
65 | for (int i=0; i<key_array.length; i++) {
|
---|
66 |
|
---|
67 | String key = key_array[i];
|
---|
68 | //String key = key_iterator.next();
|
---|
69 | if (key.matches("\\.")) {
|
---|
70 | String new_key = key.replaceAll("\\.", "<PERIOD>");
|
---|
71 | ef_count.put(new_key, ef_count.get(key));
|
---|
72 | ef_count.remove(key);
|
---|
73 | key = new_key;
|
---|
74 | }
|
---|
75 |
|
---|
76 | if (key.matches("\\$")) {
|
---|
77 | String new_key = key.replaceAll("\\$", "<DOLLAR>");
|
---|
78 | ef_count.put(new_key, ef_count.get(key));
|
---|
79 | ef_count.remove(key);
|
---|
80 | }
|
---|
81 |
|
---|
82 | }
|
---|
83 | }
|
---|
84 |
|
---|
85 | protected void fixup_page(String volume_id, String page_id, Document ef_page)
|
---|
86 | {
|
---|
87 | if (ef_page != null) {
|
---|
88 | String[] zone_keys = { "header", "body", "footer" };
|
---|
89 |
|
---|
90 | for (String zone_key: zone_keys) {
|
---|
91 | Document ef_zone = (Document)ef_page.get(zone_key);
|
---|
92 | if (ef_zone != null) {
|
---|
93 | String[] count_keys = { "beginCharCounts", "endCharCount", "tokenPosCount" };
|
---|
94 |
|
---|
95 | for (String count_key: count_keys) {
|
---|
96 | Document ef_sub_section = (Document)ef_zone.get(count_key);
|
---|
97 | if (ef_sub_section != null) {
|
---|
98 | fixup_section(ef_sub_section);
|
---|
99 |
|
---|
100 | if (count_key.equals("tokenPostCount")) {
|
---|
101 | Set<String> key_set = ef_sub_section.keySet();
|
---|
102 | for (String key : key_set) {
|
---|
103 | Document token_section = (Document)ef_sub_section.get(key);
|
---|
104 | fixup_section(token_section);
|
---|
105 | }
|
---|
106 | }
|
---|
107 |
|
---|
108 |
|
---|
109 | }
|
---|
110 | }
|
---|
111 | }
|
---|
112 | }
|
---|
113 | }
|
---|
114 | else {
|
---|
115 | System.err.println("Warning: null page for '" + page_id + "'");
|
---|
116 | }
|
---|
117 |
|
---|
118 | }
|
---|
119 | protected void fixup_volume(String json_file_in, Document extracted_feature_record)
|
---|
120 | {
|
---|
121 | String full_json_file_in = _input_dir + "/" + json_file_in;
|
---|
122 |
|
---|
123 | if (extracted_feature_record != null) {
|
---|
124 | String volume_id = extracted_feature_record.getString("id");
|
---|
125 | extracted_feature_record.put("_id",volume_id);
|
---|
126 | extracted_feature_record.remove("id");
|
---|
127 |
|
---|
128 | Document ef_features = (Document)extracted_feature_record.get("features");
|
---|
129 |
|
---|
130 | int ef_page_count = ef_features.getInteger("pageCount");
|
---|
131 |
|
---|
132 | if (_verbosity >= 1) {
|
---|
133 | System.out.println("Processing: " + json_file_in);
|
---|
134 | System.out.println(" pageCount = " + ef_page_count);
|
---|
135 | }
|
---|
136 |
|
---|
137 | List<Document> ef_pages = (List<Document>)ef_features.get("pages");
|
---|
138 | int ef_num_pages = ef_pages.size();
|
---|
139 | if (ef_num_pages != ef_page_count) {
|
---|
140 | System.err.println("Warning: number of page elements in JSON (" + ef_num_pages + ")"
|
---|
141 | +" does not match 'pageCount' metadata (" + ef_page_count + ")");
|
---|
142 | }
|
---|
143 |
|
---|
144 | if (_verbosity >= 2) {
|
---|
145 | System.out.print(" Pages: ");
|
---|
146 | }
|
---|
147 |
|
---|
148 | for (int i = 0; i < ef_page_count; i++) {
|
---|
149 | String formatted_i = String.format("page-%06d", i);
|
---|
150 | String page_id = volume_id + "." + formatted_i;
|
---|
151 |
|
---|
152 | if (_verbosity >= 2) {
|
---|
153 | if (i>0) {
|
---|
154 | System.out.print(", ");
|
---|
155 | }
|
---|
156 | System.out.print(page_id);
|
---|
157 | }
|
---|
158 |
|
---|
159 | if (i==(ef_page_count-1)) {
|
---|
160 | if (_verbosity >= 2) {
|
---|
161 | System.out.println();
|
---|
162 | }
|
---|
163 | }
|
---|
164 |
|
---|
165 | Document ef_page = (Document)ef_pages.get(i);
|
---|
166 |
|
---|
167 | if (ef_page != null) {
|
---|
168 |
|
---|
169 | fixup_page(volume_id, page_id, ef_page);
|
---|
170 | }
|
---|
171 | else {
|
---|
172 | System.err.println("Skipping: " + page_id);
|
---|
173 | }
|
---|
174 | }
|
---|
175 | }
|
---|
176 | else {
|
---|
177 | // File did not exist, or could not be parsed
|
---|
178 | String mess = "Failed to read in bzipped JSON file '" + full_json_file_in + "'";
|
---|
179 |
|
---|
180 | System.err.println("Warning: " + mess);
|
---|
181 | System.out.println("Warning: " + mess);
|
---|
182 |
|
---|
183 | }
|
---|
184 | }
|
---|
185 |
|
---|
186 | public Integer call(String json_file_in) throws IOException
|
---|
187 | {
|
---|
188 | try {
|
---|
189 | MongoClientURI mongo_url = new MongoClientURI("mongodb://gc3:27017,gc4:27017,gc5:27017");
|
---|
190 | MongoClient mongoClient = new MongoClient(mongo_url);
|
---|
191 |
|
---|
192 | MongoDatabase database = mongoClient.getDatabase("htrc_ef");
|
---|
193 | MongoCollection<Document> collection = database.getCollection("volumes");
|
---|
194 |
|
---|
195 | String full_json_file_in = _input_dir + "/" + json_file_in;
|
---|
196 | System.out.println("Processing: " + full_json_file_in);
|
---|
197 | String extracted_feature_json_doc = ClusterFileIO.readTextFile(full_json_file_in);
|
---|
198 |
|
---|
199 | Document doc = Document.parse(extracted_feature_json_doc);
|
---|
200 |
|
---|
201 | fixup_volume(json_file_in,doc);
|
---|
202 |
|
---|
203 | collection.insertOne(doc);
|
---|
204 |
|
---|
205 | /*
|
---|
206 | //Mongo mongo = new Mongo("localhost", 27017);
|
---|
207 | MongoClient mongo = new MongoClient( "localhost" , 27017 );
|
---|
208 |
|
---|
209 | DB db = mongo.getDB("yourdb");
|
---|
210 | DBCollection coll = db.getCollection("dummyColl");
|
---|
211 |
|
---|
212 | // convert JSON to DBObject directly
|
---|
213 | DBObject dbObject = (DBObject) JSON
|
---|
214 | .parse("{'name':'mkyong', 'age':30}");
|
---|
215 | coll.insert(dbObject);
|
---|
216 |
|
---|
217 |
|
---|
218 | DBCursor cursorDoc = coll.find();
|
---|
219 | while (cursorDoc.hasNext()) {
|
---|
220 | System.out.println(cursorDoc.next());
|
---|
221 | }
|
---|
222 |
|
---|
223 | System.out.println("Done");
|
---|
224 | */
|
---|
225 | mongoClient.close();
|
---|
226 |
|
---|
227 | } catch (MongoException e) {
|
---|
228 | e.printStackTrace();
|
---|
229 | }
|
---|
230 |
|
---|
231 | return 1;
|
---|
232 | }
|
---|
233 | public Integer callPageCount(String json_file_in) throws IOException
|
---|
234 | {
|
---|
235 | Integer page_count = 0;
|
---|
236 |
|
---|
237 | String full_json_file_in = _input_dir + "/" + json_file_in;
|
---|
238 | JSONObject extracted_feature_record = JSONClusterFileIO.readJSONFile(full_json_file_in);
|
---|
239 |
|
---|
240 | if (extracted_feature_record != null) {
|
---|
241 | String volume_id = extracted_feature_record.getString("id");
|
---|
242 |
|
---|
243 | JSONObject ef_features = extracted_feature_record.getJSONObject("features");
|
---|
244 |
|
---|
245 | if (_verbosity >= 1) {
|
---|
246 | System.out.println("Processing: " + json_file_in);
|
---|
247 | }
|
---|
248 |
|
---|
249 | if (ef_features != null) {
|
---|
250 | String page_count_str = ef_features.getString("pageCount");
|
---|
251 | if (!page_count_str.equals("")) {
|
---|
252 | page_count = Integer.parseInt(page_count_str);
|
---|
253 | }
|
---|
254 | else {
|
---|
255 | System.err.println("No 'pageCount' in 'features' in volume id '" + volume_id + "' => defaulting to 0");
|
---|
256 | }
|
---|
257 | }
|
---|
258 | else {
|
---|
259 | System.err.println("No 'features' section in JSON file => Skipping id: " + volume_id);
|
---|
260 | }
|
---|
261 |
|
---|
262 | }
|
---|
263 | else {
|
---|
264 | // File did not exist, or could not be parsed
|
---|
265 | String mess = "Failed to read in bzipped JSON file '" + full_json_file_in + "'";
|
---|
266 | if (_strict_file_io) {
|
---|
267 | throw new IOException(mess);
|
---|
268 | }
|
---|
269 | else {
|
---|
270 | System.err.println("Warning: " + mess);
|
---|
271 | System.out.println("Warning: " + mess);
|
---|
272 | }
|
---|
273 | }
|
---|
274 |
|
---|
275 | _progress_accum.add(_progress_step);
|
---|
276 |
|
---|
277 | return page_count;
|
---|
278 | }
|
---|
279 |
|
---|
280 |
|
---|
281 | }
|
---|
282 |
|
---|