/** Initialize a scanner over the given input split using this task attempt configuration. */
public void initialize(InputSplit inSplit, JobConf job) throws IOException {
baseSplit = (org.apache.accumulo.core.client.mapreduce.RangeInputSplit) inSplit;
log.debug("Initializing input split: " + baseSplit.toString());
Instance instance = baseSplit.getInstance(getClientConfiguration(job));
if (null == instance) {
instance = getInstance(job);
}
String principal = baseSplit.getPrincipal();
if (null == principal) {
principal = getPrincipal(job);
}
AuthenticationToken token = baseSplit.getToken();
if (null == token) {
token = getAuthenticationToken(job);
}
Authorizations authorizations = baseSplit.getAuths();
if (null == authorizations) {
authorizations = getScanAuthorizations(job);
}
String classLoaderContext = getClassLoaderContext(job);
String table = baseSplit.getTableName();
// in case the table name changed, we can still use the previous name for terms of
// configuration,
// but the scanner will use the table id resolved at job setup time
InputTableConfig tableConfig = getInputTableConfig(job, baseSplit.getTableName());
log.debug("Creating connector with user: "******"Creating scanner for table: " + table);
log.debug("Authorizations are: " + authorizations);
if (baseSplit instanceof BatchInputSplit) {
BatchScanner scanner;
BatchInputSplit multiRangeSplit = (BatchInputSplit) baseSplit;
try {
// Note: BatchScanner will use at most one thread per tablet, currently BatchInputSplit
// will not span tablets
int scanThreads = 1;
scanner =
instance
.getConnector(principal, token)
.createBatchScanner(baseSplit.getTableName(), authorizations, scanThreads);
setupIterators(job, scanner, baseSplit.getTableName(), baseSplit);
if (null != classLoaderContext) {
scanner.setClassLoaderContext(classLoaderContext);
}
} catch (Exception e) {
throw new IOException(e);
}
scanner.setRanges(multiRangeSplit.getRanges());
scannerBase = scanner;
} else if (baseSplit instanceof RangeInputSplit) {
split = (RangeInputSplit) baseSplit;
Boolean isOffline = baseSplit.isOffline();
if (null == isOffline) {
isOffline = tableConfig.isOfflineScan();
}
Boolean isIsolated = baseSplit.isIsolatedScan();
if (null == isIsolated) {
isIsolated = tableConfig.shouldUseIsolatedScanners();
}
Boolean usesLocalIterators = baseSplit.usesLocalIterators();
if (null == usesLocalIterators) {
usesLocalIterators = tableConfig.shouldUseLocalIterators();
}
Scanner scanner;
try {
if (isOffline) {
scanner =
new OfflineScanner(
instance,
new Credentials(principal, token),
baseSplit.getTableId(),
authorizations);
} else if (DeprecationUtil.isMockInstance(instance)) {
scanner =
instance
.getConnector(principal, token)
.createScanner(baseSplit.getTableName(), authorizations);
} else {
ClientConfiguration clientConf = getClientConfiguration(job);
ClientContext context =
new ClientContext(instance, new Credentials(principal, token), clientConf);
scanner = new ScannerImpl(context, baseSplit.getTableId(), authorizations);
}
if (isIsolated) {
log.info("Creating isolated scanner");
scanner = new IsolatedScanner(scanner);
}
if (usesLocalIterators) {
log.info("Using local iterators");
scanner = new ClientSideIteratorScanner(scanner);
}
setupIterators(job, scanner, baseSplit.getTableName(), baseSplit);
} catch (Exception e) {
throw new IOException(e);
}
scanner.setRange(baseSplit.getRange());
scannerBase = scanner;
} else {
throw new IllegalArgumentException(
"Can not initialize from " + baseSplit.getClass().toString());
}
Collection<Pair<Text, Text>> columns = baseSplit.getFetchedColumns();
if (null == columns) {
columns = tableConfig.getFetchedColumns();
}
// setup a scanner within the bounds of this split
for (Pair<Text, Text> c : columns) {
if (c.getSecond() != null) {
log.debug("Fetching column " + c.getFirst() + ":" + c.getSecond());
scannerBase.fetchColumn(c.getFirst(), c.getSecond());
} else {
log.debug("Fetching column family " + c.getFirst());
scannerBase.fetchColumnFamily(c.getFirst());
}
}
SamplerConfiguration samplerConfig = baseSplit.getSamplerConfiguration();
if (null == samplerConfig) {
samplerConfig = tableConfig.getSamplerConfiguration();
}
if (samplerConfig != null) {
scannerBase.setSamplerConfiguration(samplerConfig);
}
scannerIterator = scannerBase.iterator();
numKeysRead = 0;
}