@Test
public void stringSerialization() {
// Characters
DataOutput out = serialize.getDataOutput(((int) Character.MAX_VALUE) * 2 + 8, true);
for (char c = Character.MIN_VALUE; c < Character.MAX_VALUE; c++) {
out.writeObjectNotNull(Character.valueOf(c));
}
ReadBuffer b = out.getStaticBuffer().asReadBuffer();
for (char c = Character.MIN_VALUE; c < Character.MAX_VALUE; c++) {
assertEquals(c, serialize.readObjectNotNull(b, Character.class).charValue());
}
// String
for (int t = 0; t < 10000; t++) {
DataOutput out1 = serialize.getDataOutput(32 + 5, true);
DataOutput out2 = serialize.getDataOutput(32 + 5, true);
String s1 = RandomGenerator.randomString(1, 32);
String s2 = RandomGenerator.randomString(1, 32);
out1.writeObjectNotNull(s1);
out2.writeObjectNotNull(s2);
StaticBuffer b1 = out1.getStaticBuffer();
StaticBuffer b2 = out2.getStaticBuffer();
assertEquals(s1, serialize.readObjectNotNull(b1.asReadBuffer(), String.class));
assertEquals(s2, serialize.readObjectNotNull(b2.asReadBuffer(), String.class));
assertEquals(
s1 + " vs " + s2, Integer.signum(s1.compareTo(s2)), Integer.signum(b1.compareTo(b2)));
}
}
/**
* This class relies heavily on the behavior of segmented scans with respect to which hash keys
* are scanned by each segment. Here's a rough ASCII example to help illustrate:
* ___________________________ |hk:A |hk:B | ---------------------------- ^segment 1 ^segment 2
*
* <p>Because we are scanning in segments across the entire hash key space, it is possible for the
* same hash key to appear in two different segments. We are also running all of the scan segments
* in parallel, so we have no control over which segment returns first.
*
* <p>In the example, if segment 2 was the first segment to post a result, we would store hk:B as
* a "boundary" key. That way when segment 1 eventually reaches hk:B in its scan, we know that
* another segment has already returned this hash key and we can safely skip returning it.
*
* <p>By doing this, we avoid returning a RecordIterator for the same hash key twice and we only
* need to store at most 2 hash keys per segment.
*/
@Override
public List<SingleKeyRecordIterator> buildRecordIterators(ScanContext scanContext) {
final ScanResult dynamoDbResult = scanContext.getScanResult();
final int segment = scanContext.getScanRequest().getSegment();
final List<Map<String, AttributeValue>> items = dynamoDbResult.getItems();
// If the scan returned no results, we need to shortcut and just throw back an empty result set
if (items.isEmpty()) {
return Collections.emptyList();
}
List<SingleKeyRecordIterator> recordIterators = Lists.newLinkedList();
final Iterator<Map<String, AttributeValue>> itemIterator = items.iterator();
final Map<String, AttributeValue> firstItem = itemIterator.next();
final StaticBuffer firstKey = new KeyBuilder(firstItem).build(Constants.TITAN_HASH_KEY);
// Computes the full set of boundary keys up to this point. This includes the previous end key
// for this segment.
final ImmutableSet<StaticBuffer> boundaryKeys = aggregateBoundaryKeys();
// The first key in this scan segment might already have been returned by a previous scan
// segment
if (!boundaryKeys.contains(firstKey)) {
recordIterators.add(buildRecordIteratorForHashKey(firstKey));
}
StaticBuffer hashKey = firstKey;
while (itemIterator.hasNext()) {
final Optional<StaticBuffer> nextKey = findNextHashKey(itemIterator, hashKey);
if (nextKey.isPresent()) {
// Found a new hash key. Make a record iterator and look for the next unique hash key
hashKey = nextKey.get();
recordIterators.add(buildRecordIteratorForHashKey(hashKey));
}
}
// If we've already seen the final hashKey in a previous scan segment result, we want to avoid
// returning it again.
if (!hashKey.equals(firstKey) && boundaryKeys.contains(hashKey)) {
recordIterators.remove(recordIterators.size() - 1);
}
// Update the boundary keys for this segment
if (scanContext.isFirstResult()) {
setInitialBoundaryKeys(segment, firstKey, hashKey);
} else {
updateLastKey(segment, hashKey);
}
return recordIterators;
}
private Optional<StaticBuffer> findNextHashKey(
Iterator<Map<String, AttributeValue>> itemIterator, StaticBuffer previousKey) {
Optional<StaticBuffer> result = Optional.absent();
while (itemIterator.hasNext() && !result.isPresent()) {
final StaticBuffer nextKey =
new KeyBuilder(itemIterator.next()).build(Constants.TITAN_HASH_KEY);
if (!nextKey.equals(previousKey)) {
result = Optional.of(nextKey);
}
}
return result;
}
@Override
public boolean equals(Object other) {
if (this == other) return true;
else if (other == null) return false;
else if (!getClass().isInstance(other)) return false;
KeySliceQuery oth = (KeySliceQuery) other;
return key.equals(oth.key) && super.equals(oth);
}
@Override
public int hashCode() {
if (hashcode == 0) {
hashcode = key.hashCode() * 102329 + super.hashCode();
if (hashcode == 0) hashcode = 1;
}
return hashcode;
}
public boolean subsumes(KeySliceQuery oth) {
return key.equals(oth.key) && super.subsumes(oth);
}
