Changeset 31201 for other-projects/hathitrust

Timestamp:

2016-12-12T10:35:42+13:00 (7 years ago)

Author:

davidb

Message:

Trigger serialization of whitelist in main program

Location:

other-projects/hathitrust/wcsa/extracted-features-solr/trunk/solr-ingest/src/main/java/org/hathitrust/extractedfeatures

Files:

• ProcessForSolrIngest.java (modified) (1 diff)
• WhitelistBloomFilter.java (added)

other-projects/hathitrust/wcsa/extracted-features-solr/trunk/solr-ingest/src/main/java/org/hathitrust/extractedfeatures/ProcessForSolrIngest.java

@@ -310 +310 @@
             = new ProcessForSolrIngest(input_dir,json_list_filename,solr_url,output_dir,verbosity);
         
+        
+        boolean use_whitelist = Boolean.getBoolean("wcsa-ef-ingest.use-whitelist");
+                
+        if (use_whitelist) {
+            String whitelist_filename = System.getProperty("wcsa-ef-ingest.whitelist-filename");
+
+            WhitelistBloomFilter whitelist_bloomfilter = new WhitelistBloomFilter(whitelist_filename,true);
+            whitelist_bloomfilter.contains("foo");
+        }
+        
+        
         String process_ef_json_mode = System.getProperty("wcsa-ef-ingest.process-ef-json-mode","per-page");
         if (process_ef_json_mode.equals("per-volume")) {