Package fr.eolya.utils.nosql.mongodb

Examples of fr.eolya.utils.nosql.mongodb.MongoDBCollection

43
44
45
46
47
48
49
50
51
    String query = "{\"name\": \"version\"}"; 
    return coll.getValue(MongoDBHelper.JSON2BasicDBObject (query), "value");   
  }

  public String getSourceClass(String sourceType) {
    MongoDBCollection coll = new MongoDBCollection(db,"plugins");
    String query = String.format("{\"id\": %1$s}", sourceType); 
    return coll.getValue(MongoDBHelper.JSON2BasicDBObject (query), "class_java");   
  }
View Full Code Here
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
    String query = String.format("{\"id\": %1$s}", sourceType); 
    return coll.getValue(MongoDBHelper.JSON2BasicDBObject (query), "class_java");   
  }

  public HashMap<String,String> getTarget(String targetId) {
    MongoDBCollection coll = new MongoDBCollection(db,"targets");
    String query = String.format("{\"id\": %1$s}", targetId); 
    try {
      // TODO : replace by getItemMap
      return JSONHelper.getJSONMapString(coll.getJson(MongoDBHelper.JSON2BasicDBObject (query)));
    } catch (Exception e) {
      //e.printStackTrace();
      return null;
    }
  }
View Full Code Here
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
    this.con.close();
  }

  public void fixStartupSourcesStatus () {
    // TODO : should work on a specific account or all account of an engine if either engine_id or account_id are specified ???
    MongoDBCollection coll = new MongoDBCollection(db,"sources");
    String query = String.format("{\"$or\": [{\"crawl_process_status\": \"1\"}, {\"crawl_process_status\": \"5\"}, {\"_poped\": true}]}"); 

    BasicDBObject docsearch = MongoDBHelper.JSON2BasicDBObject(query);
    synchronized (sourcesCollMonitor) {
      DBCursor cur = coll.getColl().find(docsearch);
      if (cur.count()==0) return;
      while (cur.hasNext()) {
        BasicDBObject doc = (BasicDBObject) cur.next();
        BasicDBObject doc2 = (BasicDBObject) doc.copy();
        doc2.put("crawl_nexttime", new Date());
        doc2.put("crawl_process_status", ISource.CRAWL_PROCESS_MODE_NONE);
        doc2.put("crawl_priority", "2");
        doc2.put("running_crawl_lastupdate", null);
        doc2.put("_poped", false);
        coll.update(doc, doc2);         
      }
    }
  }
View Full Code Here
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
  public boolean updateSourceStatusStartup(int id, long queueSize, long doneQueueSize) {
   
    logger.log("        Source " + String.valueOf(id) + " starting : update status");
   
    MongoDBCollection coll = new MongoDBCollection(db,"sources");
    String query = String.format("{\"id\": %1$s}", id); 

    BasicDBObject docsearch = MongoDBHelper.JSON2BasicDBObject(query);

    synchronized (sourcesCollMonitor) {
      DBCursor cur = coll.getColl().find(docsearch);

      if (cur.count()!=1) return false;

      BasicDBObject doc = (BasicDBObject) cur.next();
      BasicDBObject doc2 = (BasicDBObject) doc.copy();

      if (doneQueueSize>0 || queueSize>0) {
        doc2.put("running_crawl_item_processed", doneQueueSize);
        doc2.put("running_crawl_item_to_process", queueSize);

      } else {
        doc2.put("running_crawl_item_processed", 0);
        doc2.put("running_crawl_item_to_process", 0);
      }
      doc2.put("crawl_lasttime_start", new Date());         
      doc2.put("running_crawl_lastupdate", new Date());  
      doc2.put("crawl_process_status", ISource.CRAWL_PROCESS_STATUS_CRAWLING);
      doc2.put("crawl_mode", ISource.CRAWL_PROCESS_MODE_NONE);
      doc2.put("_poped", false);
      coll.update(doc, doc2);       
    }
    return true;
  }
View Full Code Here
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
    }
    return true;
  }

  public boolean updateSourceProcessingInfo(int id, long queueSize, long doneQueueSize, String processingInfo) {
    MongoDBCollection coll = new MongoDBCollection(db,"sources");
    String query = String.format("{\"id\": %1$s}", id); 

    BasicDBObject docsearch = MongoDBHelper.JSON2BasicDBObject(query);
    synchronized (sourcesCollMonitor) {
      DBCursor cur = coll.getColl().find(docsearch);
      if (cur.count()!=1) return false;
      BasicDBObject doc = (BasicDBObject) cur.next();
      BasicDBObject doc2 = (BasicDBObject) doc.copy();
      doc2.put("running_crawl_item_processed", doneQueueSize);
      doc2.put("running_crawl_item_to_process", queueSize);
      doc2.put("running_crawl_lastupdate", new Date());  
      doc2.put("processing_info", processingInfo);  
      coll.update(doc, doc2);       
    }
    return true;
  }
View Full Code Here
142
143
144
145
146
147
148
149
    }
    return true;
  }

  private long getSourcePageCount(int id) {
    MongoDBCollection coll = new MongoDBCollection(dbQueues,"pages_" + String.valueOf(id));
    return coll.count("");
  }
View Full Code Here
147
148
149
150
151
152
153
154
155
    MongoDBCollection coll = new MongoDBCollection(dbQueues,"pages_" + String.valueOf(id));
    return coll.count("");
  }

  private long getSourcePageCountSuccess(int id) {
    MongoDBCollection coll = new MongoDBCollection(dbQueues,"pages_" + String.valueOf(id));
    String query = "{\"crawl_status\": 200}"; 
    return coll.count(query);
  }
View Full Code Here
153
154
155
156
157
158
159
160
161
    String query = "{\"crawl_status\": 200}"; 
    return coll.count(query);
  }

  private long getSourcePageCountPushed(int id) {
    MongoDBCollection coll = new MongoDBCollection(dbQueues,"pages_" + String.valueOf(id));
    String query = "{\"crawl_status\": 200, \"crawl_mode\": \"a\"}"; 
    return coll.count(query);
  }
View Full Code Here
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
  public boolean updateSourceStatusStop(ISource src, long queueSize, long doneQueueSize, boolean crawlerStopRequested, boolean pause, boolean pauseBySchedule, XMLConfig config) {

    logger.log("        Source " + String.valueOf(src.getId()) + " ending : update status and counters");

    MongoDBCollection coll = new MongoDBCollection(db,"sources");
    String query = String.format("{\"id\": %1$s}", src.getId()); 

    BasicDBObject docsearch = MongoDBHelper.JSON2BasicDBObject(query);

    synchronized (sourcesCollMonitor) {
      DBCursor cur = coll.getColl().find(docsearch);

      if (cur.count()!=1) return false;

      BasicDBObject doc = (BasicDBObject) cur.next();
      BasicDBObject doc2 = (BasicDBObject) doc.copy();


      doc2.put("crawl_lastpagecount", doneQueueSize);  
      doc2.put("crawl_pagecount", getSourcePageCount(src.getId()));  
      doc2.put("crawl_pagecount_success", getSourcePageCountSuccess(src.getId()));  
      doc2.put("crawl_pagecount_pushed", getSourcePageCountPushed(src.getId()));  

      // TODO: v4
      /*
            if (item.getDepth()==0) {
                 rootUrlStatusCode = statusCode;
                 rootUrlStatusMessage = statusMessage.trim();
             }
       */
      //doc2.put("crawl_status", rootUrlStatusCode);  
      //doc2.put("crawl_status", rootUrlStatusMessage);  

      /*
            s += ",crawl_lastpagecount = " + String.valueOf(done.size());
            s += ",crawl_pagecount = '" + pageCount + "'";
            s += ",crawl_pagecount_success = '" + pageCountSuccess + "'";
            s += ",crawl_pagecount_pushed = '" + pageCountPushed + "'";
            s += ",crawl_status = '" + rootUrlStatusCode + "'";
            s += ",crawl_status_message = '" + rootUrlStatusMessage.replace("'", "\\'") + "'";
       */

      if (queueSize==0)
        doc2.put("crawl_firstcompleted", "1");  

      if (pause) {
        doc2.put("running_crawl_item_processed", doneQueueSize);
        doc2.put("running_crawl_item_to_process", queueSize);
        if (pauseBySchedule) {
          doc2.put("crawl_process_status", ISource.CRAWL_PROCESS_STATUS_PAUSE_BY_SCHEDULE);
        }
        doc2.put("crawl_priority", "4");
      } else {
        doc2.put("running_crawl_item_processed", 0);
        doc2.put("running_crawl_item_to_process", 0);
        doc2.put("crawl_process_status", ISource.CRAWL_PROCESS_MODE_NONE);             
        //if (crawlerStopRequested) {
        //  doc2.put("crawl_priority", 2);
        //}
        //else {
        doc2.put("crawl_priority", "0");

        int crawlPeriod;
        if (src.isOptimized())           
          crawlPeriod = Integer.parseInt(config.getProperty("/crawler/param[@name='period_optimized_source']", "24"));
        else
          crawlPeriod = Integer.parseInt(config.getProperty("/crawler/param[@name='period']", "168"));

        Date read_nexttime = getNextCrawlStartDate(src.getCurrentCrawlNextTime(), crawlPeriod, Integer.parseInt(src.getCrawlPeriod()));
        doc2.put("crawl_nexttime", read_nexttime);
        //}
      }

      doc2.put("running_crawl_lastupdate", new Date());  
      doc2.put("crawl_lasttime_end", new Date());  
      doc2.put("crawl_mode", ISource.CRAWL_PROCESS_MODE_NONE);

      if (src.getExtra()!=null && src.getExtra().asXml()!=null)
        doc2.put("extra", src.getExtra().asXml().replaceAll("'", "''"));
      else
        doc2.put("extra", "");

      coll.update(doc, doc2);   
     
      updateSourceLog(src, 0);
    }
    return true;
  }
View Full Code Here
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
    }
    return true;
  }
 
  public boolean updateSourceLog(ISource src, int retention) {
    MongoDBCollection coll = new MongoDBCollection(db,"sources_log");
    coll.createIndex("id_source", false);
    coll.createIndex("createtime", false);

    BasicDBObject doc = new BasicDBObject();
    doc.put("id_source", src.getId());
    doc.put("createtime", new Date().getTime());
    doc.put("log",src.memLogGet());

    coll.add(doc);
   
    return true;
  }
View Full Code Here
TOP

Related Classes of fr.eolya.utils.nosql.mongodb.MongoDBCollection

  • fr.eolya.crawler.cache.mongodb.MongoDBDocumentCache
  • fr.eolya.crawler.database.mongodb.MongoDBCrawlerDB
  • org.bson.types.ObjectId
    • Copyright © 2018 www.massapicom. 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.