Changeset 31204 for other-projects/hathitrust/wcsa/extracted-features-solr
- Timestamp:
- 2016-12-12T14:28:20+13:00 (7 years ago)
- Location:
- other-projects/hathitrust/wcsa/extracted-features-solr/trunk/solr-ingest/src/main/java
- Files:
-
- 5 added
- 1 edited
Legend:
- Unmodified
- Added
- Removed
-
other-projects/hathitrust/wcsa/extracted-features-solr/trunk/solr-ingest/src/main/java/org/hathitrust/extractedfeatures/WhitelistBloomFilter.java
r31203 r31204 3 3 import java.io.BufferedInputStream; 4 4 import java.io.BufferedOutputStream; 5 import java.io.DataInputStream; 6 import java.io.DataOutputStream; 5 7 import java.io.File; 6 8 import java.io.FileInputStream; … … 9 11 import java.io.IOException; 10 12 import java.io.InputStream; 13 import java.io.OutputStream; 11 14 import java.io.Serializable; 12 15 import java.nio.charset.Charset; … … 18 21 import javax.annotation.Nullable; 19 22 23 import static com.google.common.base.Preconditions.checkNotNull; 24 20 25 import com.google.common.base.Preconditions; 21 import com.google.common.hash.BloomFilter; 26 //import com.google.common.hash.BloomFilter; 27 import com.google.common.hash.BloomFilterAdvanced; 28 //import com.google.common.hash.BloomFilterStrategies.BitArray; 22 29 import com.google.common.hash.Funnel; 23 30 import com.google.common.hash.Funnels; 24 31 import com.google.common.hash.PrimitiveSink; 32 import com.google.common.primitives.SignedBytes; 33 import com.google.common.primitives.UnsignedBytes; 25 34 26 35 public class WhitelistBloomFilter { … … 28 37 29 38 30 protected BloomFilter <CharSequence> _bloomFilter;39 protected BloomFilterAdvanced<CharSequence> _bloomFilter; 31 40 protected static final String SERIALIZED_SUFFIX = "-serialized"; 32 41 protected static final double FALSE_POSITIVE_PERCENTAGE = 0.01; … … 78 87 //Funnel<CharSequence> string_funnel = Funnels.stringFunnel(StandardCharsets.UTF_8); 79 88 Funnel<CharSequence> string_funnel = stringFunnel(StandardCharsets.UTF_8); 80 _bloomFilter = BloomFilter .create(string_funnel, num_lines,FALSE_POSITIVE_PERCENTAGE);89 _bloomFilter = BloomFilterAdvanced.create(string_funnel, num_lines,FALSE_POSITIVE_PERCENTAGE); 81 90 } 82 91 catch (IOException e) { … … 139 148 } 140 149 141 protected static BloomFilter <CharSequence> serializeIn(File ser_file)142 { 143 BloomFilter <CharSequence> bloomFilter = null;150 protected static BloomFilterAdvanced<CharSequence> serializeIn(File ser_file) 151 { 152 BloomFilterAdvanced<CharSequence> bloomFilter = null; 144 153 145 154 try { … … 149 158 //Funnel<CharSequence> string_funnel = Funnels.stringFunnel(StandardCharsets.UTF_8); 150 159 Funnel<CharSequence> string_funnel = stringFunnel(StandardCharsets.UTF_8); 151 bloomFilter = BloomFilter .readFrom(bfis,string_funnel);160 bloomFilter = BloomFilterAdvanced.readFrom(bfis,string_funnel); 152 161 153 162 bfis.close(); … … 225 234 226 235 } 227 228 236 } 229 237
Note:
See TracChangeset
for help on using the changeset viewer.