@Override
public Set<Element> get(Object key) {
Get get = new Get(ByteArraySerializer.fromObject(key));
Result result;
try {
result = backingTable.get(get);
} catch (IOException e) {
LOG.severe("Cannot get from backing table");
e.printStackTrace();
return null;
}
NavigableMap<byte[], byte[]> map = result.getFamilyMap(Bytes.toBytes(VALUES));
if (null == map) return null;
HashSet<Element> elementHashSet = new HashSet<Element>();
for (byte[] byteArray : map.keySet()) {
if (indexClass.equals(HVertex.class) || indexClass.equals(Vertex.class)) {
HVertex hVertex = new HVertex(hGraph);
hVertex.setId(byteArray);
elementHashSet.add(hVertex);
} else {
final HEdge hEdge = new HEdge(hGraph);
hEdge.setId(byteArray);
elementHashSet.add(hEdge);
}
}
return elementHashSet;
}
public int run(String[] args) throws Exception {
Configuration argConf = getConf();
// JobConf conf = new JobConf(diffdb.class);
Configuration config = HBaseConfiguration.create();
HBaseAdmin hbAdmin = new HBaseAdmin(config);
dbutil db_util = new dbutil(config);
HTable runTable = new HTable(config, "gestore_runs");
Get runGet = new Get(argConf.get("id").getBytes());
Result pipeline = runTable.get(runGet);
NavigableMap<byte[], byte[]> pipeMap = pipeline.getFamilyMap("d".getBytes());
Map.Entry<byte[], byte[]> results = pipeMap.pollFirstEntry();
HashMap<String, HashMap<String, String>> resultMap =
new HashMap<String, HashMap<String, String>>();
while (results != null) {
String resultKey = new String(results.getKey());
String resultValue = new String(results.getValue());
String field = "type";
HashMap<String, String> tempMap = new HashMap<String, String>();
String entry = resultKey;
if (resultKey.endsWith("_db_timestamp")) {
field = "db_timestamp";
entry = resultKey.substring(0, resultKey.lastIndexOf("_db_timestamp"));
} else if (resultKey.endsWith("_filename")) {
field = "filename";
entry = resultKey.substring(0, resultKey.lastIndexOf("_filename"));
} else if (resultKey.endsWith("_regex")) {
field = "regex";
entry = resultKey.substring(0, resultKey.lastIndexOf("_regex"));
}
if (resultMap.containsKey(entry)) {
tempMap = resultMap.get(entry);
}
tempMap.put(field, resultValue);
resultMap.put(entry, tempMap);
// System.out.println("Key: " + resultKey + " Value: " + resultValue);
results = pipeMap.pollFirstEntry();
}
for (String key : resultMap.keySet()) {
System.out.println("File ID: " + key);
for (String subKey : resultMap.get(key).keySet()) {
// System.out.println("\t " + subKey + "\t\t" + resultMap.get(key).get(subKey));
System.out.format(" %1$-20s %2$s\n", subKey, resultMap.get(key).get(subKey));
}
}
return 0;
}
@Test
public void testDisableAndEnableTable() throws IOException {
final byte[] row = Bytes.toBytes("row");
final byte[] qualifier = Bytes.toBytes("qualifier");
final byte[] value = Bytes.toBytes("value");
final byte[] table = Bytes.toBytes("testDisableAndEnableTable");
HTable ht = TEST_UTIL.createTable(table, HConstants.CATALOG_FAMILY);
Put put = new Put(row);
put.add(HConstants.CATALOG_FAMILY, qualifier, value);
ht.put(put);
Get get = new Get(row);
get.addColumn(HConstants.CATALOG_FAMILY, qualifier);
ht.get(get);
this.admin.disableTable(table);
// Test that table is disabled
get = new Get(row);
get.addColumn(HConstants.CATALOG_FAMILY, qualifier);
boolean ok = false;
try {
ht.get(get);
} catch (NotServingRegionException e) {
ok = true;
} catch (RetriesExhaustedException e) {
ok = true;
}
assertTrue(ok);
this.admin.enableTable(table);
// Test that table is enabled
try {
ht.get(get);
} catch (RetriesExhaustedException e) {
ok = false;
}
assertTrue(ok);
}
public List<Variant> getRecordSimpleStats(
String study, int missing_gt, float maf, String maf_allele) {
BasicDBObject compare =
new BasicDBObject("studies.stats.allele_maf", maf_allele)
.append("studies.stats.MAF", maf)
.append("studies.stats.missing", missing_gt);
List<Get> hbaseQuery = new ArrayList<>();
DBCollection collection = db.getCollection("variants");
Iterator<DBObject> result = collection.find(compare);
String chromosome = new String();
while (result.hasNext()) {
DBObject variant = result.next();
String position = variant.get("_id").toString();
// hbase query construction
Get get = new Get(position.getBytes());
hbaseQuery.add(get);
}
// Complete results, from HBase
tableName = study;
effectTableName = tableName + "effect";
Map<String, Variant> resultsMap = new HashMap<>();
try {
HTable table = new HTable(admin.getConfiguration(), tableName);
HTable effectTable = new HTable(admin.getConfiguration(), effectTableName);
Result[] hbaseResultEffect = effectTable.get(hbaseQuery);
Result[] hbaseResultStats = table.get(hbaseQuery);
// List<Variant> results = new LinkedList<>();
for (Result r : hbaseResultStats) {
String position = new String(r.getRow(), CHARSET_UTF_8);
String[] aux = position.split("_");
String inner_position = aux[1];
String chr = aux[0];
// position parsing
if (chr.startsWith("0")) {
chr = chr.substring(1);
}
while (inner_position.startsWith("0")) {
inner_position = inner_position.substring(1);
}
List<VariantFieldsProtos.VariantSample> samples = new LinkedList<>();
NavigableMap<byte[], byte[]> infoMap = r.getFamilyMap("i".getBytes());
byte[] byteStats = infoMap.get((study + "_stats").getBytes());
VariantFieldsProtos.VariantStats stats =
VariantFieldsProtos.VariantStats.parseFrom(byteStats);
byte[] byteInfo = infoMap.get((study + "_data").getBytes());
VariantFieldsProtos.VariantInfo info = VariantFieldsProtos.VariantInfo.parseFrom(byteInfo);
String alternate = StringUtils.join(info.getAlternateList(), ", ");
String reference = info.getReference();
Variant partialResult =
new Variant(
chr,
Integer.parseInt(inner_position),
Integer.parseInt(inner_position),
reference,
alternate);
String format = StringUtils.join(info.getFormatList(), ":");
NavigableMap<byte[], byte[]> sampleMap = r.getFamilyMap("d".getBytes());
Map<String, Map<String, String>> resultSampleMap = new HashMap<>();
// StringBuilder sampleRaw = new StringBuilder();
for (byte[] s : sampleMap.keySet()) {
String qual = (new String(s, CHARSET_UTF_8)).replaceAll(study + "_", "");
VariantFieldsProtos.VariantSample sample =
VariantFieldsProtos.VariantSample.parseFrom(sampleMap.get(s));
String sample1 = sample.getSample();
String[] values = sample1.split(":");
String[] fields = format.split(":");
Map<String, String> singleSampleMap = new HashMap<>();
for (int i = 0; i < fields.length; i++) {
singleSampleMap.put(fields[i], values[i]);
}
resultSampleMap.put(qual, singleSampleMap);
}
VariantStats variantStats =
new VariantStats(
chromosome,
Integer.parseInt(inner_position),
reference,
alternate,
stats.getMaf(),
stats.getMgf(),
stats.getMafAllele(),
stats.getMgfGenotype(),
stats.getMissingAlleles(),
stats.getMissingGenotypes(),
stats.getMendelianErrors(),
stats.getIsIndel(),
stats.getCasesPercentDominant(),
stats.getControlsPercentDominant(),
stats.getCasesPercentRecessive(),
stats.getControlsPercentRecessive());
partialResult.setStats(variantStats);
resultsMap.put(new String(r.getRow(), CHARSET_UTF_8), partialResult);
}
for (Result r : hbaseResultEffect) {
if (!r.isEmpty()) {
NavigableMap<byte[], byte[]> effectMap = r.getFamilyMap("e".getBytes());
Variant partialResult = resultsMap.get(new String(r.getRow(), CHARSET_UTF_8));
System.out.println("Recuperado " + partialResult.toString());
String s = partialResult.getReference() + "_" + partialResult.getAlternate();
VariantEffectProtos.EffectInfo effectInfo =
VariantEffectProtos.EffectInfo.parseFrom(effectMap.get(s.getBytes()));
VariantEffect variantEffect =
new VariantEffect(
partialResult.getChromosome(),
(int) partialResult.getStart(),
partialResult.getReference(),
partialResult.getAlternate(),
effectInfo.getFeatureId(),
effectInfo.getFeatureName(),
effectInfo.getFeatureType(),
effectInfo.getFeatureBiotype(),
effectInfo.getFeatureChromosome(),
effectInfo.getFeatureStart(),
effectInfo.getFeatureEnd(),
effectInfo.getFeatureStrand(),
effectInfo.getSnpId(),
effectInfo.getAncestral(),
effectInfo.getAlternative(),
effectInfo.getGeneId(),
effectInfo.getTranscriptId(),
effectInfo.getGeneName(),
effectInfo.getConsequenceType(),
effectInfo.getConsequenceTypeObo(),
effectInfo.getConsequenceTypeDesc(),
effectInfo.getConsequenceTypeType(),
effectInfo.getAaPosition(),
effectInfo.getAminoacidChange(),
effectInfo.getCodonChange());
resultsMap.put(new String(r.getRow(), CHARSET_UTF_8), partialResult);
}
}
} catch (InvalidProtocolBufferException e) {
System.err.println(e.getClass().getName() + ": " + e.getMessage());
} catch (IOException e) {
System.err.println(e.getClass().getName() + ": " + e.getMessage());
}
List<Variant> results = new ArrayList<>(resultsMap.values());
return results;
}
@Test
public void testHostRank() throws Exception {
if (System.getProperty("prop.mapred.job.tracker") != null) {
if (LOG.isInfoEnabled())
LOG.info("testHBaseInputOutput: Ignore this test if not local mode.");
return;
}
File jarTest = new File(System.getProperty("prop.jarLocation"));
if (!jarTest.exists()) {
fail(
"Could not find Giraph jar at "
+ "location specified by 'prop.jarLocation'. "
+ "Make sure you built the main Giraph artifact?.");
}
MiniHBaseCluster cluster = null;
MiniZooKeeperCluster zkCluster = null;
FileSystem fs = null;
try {
// using the restart method allows us to avoid having the hbase
// root directory overwritten by /home/$username
zkCluster = testUtil.startMiniZKCluster();
testUtil.restartHBaseCluster(2);
cluster = testUtil.getMiniHBaseCluster();
final byte[] OL_BYTES = Bytes.toBytes("ol");
final byte[] S_BYTES = Bytes.toBytes("s");
final byte[] METADATA_BYTES = Bytes.toBytes("mtdt");
final byte[] HR_BYTES = Bytes.toBytes("_hr_");
final byte[] TAB = Bytes.toBytes(TABLE_NAME);
Configuration conf = cluster.getConfiguration();
HTableDescriptor desc = new HTableDescriptor(TAB);
desc.addFamily(new HColumnDescriptor(OL_BYTES));
desc.addFamily(new HColumnDescriptor(S_BYTES));
desc.addFamily(new HColumnDescriptor(METADATA_BYTES));
HBaseAdmin hbaseAdmin = new HBaseAdmin(conf);
if (hbaseAdmin.isTableAvailable(TABLE_NAME)) {
hbaseAdmin.disableTable(TABLE_NAME);
hbaseAdmin.deleteTable(TABLE_NAME);
}
hbaseAdmin.createTable(desc);
/**
* Enter the initial data (a,b), (b,c), (a,c) a = 1.0 - google b = 1.0 - yahoo c = 1.0 - bing
*/
HTable table = new HTable(conf, TABLE_NAME);
Put p1 = new Put(Bytes.toBytes("com.google.www"));
p1.add(OL_BYTES, Bytes.toBytes("www.yahoo.com"), Bytes.toBytes("ab"));
Put p2 = new Put(Bytes.toBytes("com.google.www"));
p2.add(OL_BYTES, Bytes.toBytes("www.bing.com"), Bytes.toBytes("ac"));
p2.add(OL_BYTES, Bytes.toBytes("www.bing.com"), Bytes.toBytes("invalid1"));
p2.add(OL_BYTES, Bytes.toBytes("www.google.com"), Bytes.toBytes("invalid2"));
Put p3 = new Put(Bytes.toBytes("com.yahoo.www"));
p3.add(OL_BYTES, Bytes.toBytes("www.bing.com"), Bytes.toBytes("bc"));
// p3.add(OL_BYTES, Bytes.toBytes(""), Bytes.toBytes("invalid4"));
Put p4 = new Put(Bytes.toBytes("com.bing.www"));
// TODO: Handle below case. use apache isValid method.
p4.add(OL_BYTES, Bytes.toBytes("http://invalidurl"), Bytes.toBytes("invalid5"));
p4.add(S_BYTES, S_BYTES, Bytes.toBytes(10.0d));
Put p5 = new Put(Bytes.toBytes("dummy"));
p5.add(S_BYTES, S_BYTES, Bytes.toBytes(10.0d));
table.put(p1);
table.put(p2);
table.put(p3);
table.put(p4);
table.put(p5);
// Set Giraph configuration
// now operate over HBase using Vertex I/O formats
conf.set(TableInputFormat.INPUT_TABLE, TABLE_NAME);
conf.set(TableOutputFormat.OUTPUT_TABLE, TABLE_NAME);
// Start the giraph job
GiraphJob giraphJob = new GiraphJob(conf, BspCase.getCallingMethodName());
GiraphConfiguration giraphConf = giraphJob.getConfiguration();
giraphConf.setZooKeeperConfiguration(cluster.getMaster().getZooKeeper().getQuorum());
setupConfiguration(giraphJob);
giraphConf.setComputationClass(LinkRankComputation.class);
giraphConf.setMasterComputeClass(LinkRankVertexMasterCompute.class);
giraphConf.setOutEdgesClass(ByteArrayEdges.class);
giraphConf.setVertexInputFormatClass(Nutch2HostInputFormat.class);
giraphConf.setVertexOutputFormatClass(Nutch2HostOutputFormat.class);
giraphConf.setInt("giraph.linkRank.superstepCount", 10);
giraphConf.setInt("giraph.linkRank.scale", 10);
giraphConf.set("giraph.linkRank.family", "mtdt");
giraphConf.set("giraph.linkRank.qualifier", "_hr_");
giraphConf.setVertexInputFilterClass(HostRankVertexFilter.class);
assertTrue(giraphJob.run(true));
if (LOG.isInfoEnabled()) LOG.info("Giraph job successful. Checking output qualifier.");
/** Check the results * */
Result result;
String key;
byte[] calculatedScoreByte;
HashMap expectedValues = new HashMap<String, Double>();
expectedValues.put("com.google.www", 1.3515060339386287d);
expectedValues.put("com.yahoo.www", 4.144902009567587d);
expectedValues.put("com.bing.www", 9.063893290511482d);
for (Object keyObject : expectedValues.keySet()) {
key = keyObject.toString();
result = table.get(new Get(key.getBytes()));
calculatedScoreByte = result.getValue(METADATA_BYTES, HR_BYTES);
assertNotNull(calculatedScoreByte);
assertTrue(calculatedScoreByte.length > 0);
Assert.assertEquals(
"Scores are not the same",
(Double) expectedValues.get(key),
Bytes.toDouble(calculatedScoreByte),
DELTA);
}
} finally {
if (cluster != null) {
cluster.shutdown();
}
if (zkCluster != null) {
zkCluster.shutdown();
}
// clean test files
if (fs != null) {
fs.delete(hbaseRootdir);
}
}
}
