public static void BeforeClass()
{
Random random = Random();
INTS = new int[COUNT];
LONGS = new long[COUNT];
RANDOM_TEST_BYTES = new byte[COUNT * (5 + 4 + 9 + 8)];
ByteArrayDataOutput bdo = new ByteArrayDataOutput(RANDOM_TEST_BYTES);
for (int i = 0; i < COUNT; i++)
{
int i1 = INTS[i] = random.Next();
bdo.WriteVInt(i1);
bdo.WriteInt(i1);
long l1;
if (Rarely())
{
// a long with lots of zeroes at the end
l1 = LONGS[i] = TestUtil.NextLong(random, 0, int.MaxValue) << 32;
}
else
{
l1 = LONGS[i] = TestUtil.NextLong(random, 0, long.MaxValue);
}
bdo.WriteVLong(l1);
bdo.WriteLong(l1);
}
}
/// <summary>
/// Builds an <seealso cref="SynonymMap"/> and returns it.
/// </summary>
public virtual SynonymMap Build()
{
ByteSequenceOutputs outputs = ByteSequenceOutputs.Singleton;
// TODO: are we using the best sharing options?
var builder = new Builder <BytesRef>(FST.INPUT_TYPE.BYTE4, outputs);
BytesRef scratch = new BytesRef(64);
ByteArrayDataOutput scratchOutput = new ByteArrayDataOutput();
HashSet <int?> dedupSet;
if (dedup)
{
dedupSet = new HashSet <int?>();
}
else
{
dedupSet = null;
}
var spare = new byte[5];
IEnumerable <CharsRef> keys = workingSet.Keys;
CharsRef[] sortedKeys = keys.ToArray();
#pragma warning disable 612, 618
System.Array.Sort(sortedKeys, CharsRef.UTF16SortedAsUTF8Comparer);
#pragma warning restore 612, 618
IntsRef scratchIntsRef = new IntsRef();
//System.out.println("fmap.build");
for (int keyIdx = 0; keyIdx < sortedKeys.Length; keyIdx++)
{
CharsRef input = sortedKeys[keyIdx];
MapEntry output = workingSet[input];
int numEntries = output.ords.Count;
// output size, assume the worst case
int estimatedSize = 5 + numEntries * 5; // numEntries + one ord for each entry
scratch.Grow(estimatedSize);
scratchOutput.Reset(scratch.Bytes, scratch.Offset, scratch.Bytes.Length);
Debug.Assert(scratch.Offset == 0);
// now write our output data:
int count = 0;
for (int i = 0; i < numEntries; i++)
{
if (dedupSet != null)
{
// box once
int?ent = output.ords[i];
if (dedupSet.Contains(ent))
{
continue;
}
dedupSet.Add(ent);
}
scratchOutput.WriteVInt(output.ords[i]);
count++;
}
int pos = scratchOutput.Position;
scratchOutput.WriteVInt(count << 1 | (output.includeOrig ? 0 : 1));
int pos2 = scratchOutput.Position;
int vIntLen = pos2 - pos;
// Move the count + includeOrig to the front of the byte[]:
Array.Copy(scratch.Bytes, pos, spare, 0, vIntLen);
Array.Copy(scratch.Bytes, 0, scratch.Bytes, vIntLen, pos);
Array.Copy(spare, 0, scratch.Bytes, 0, vIntLen);
if (dedupSet != null)
{
dedupSet.Clear();
}
scratch.Length = scratchOutput.Position - scratch.Offset;
//System.out.println(" add input=" + input + " output=" + scratch + " offset=" + scratch.offset + " length=" + scratch.length + " count=" + count);
builder.Add(Lucene.Net.Util.Fst.Util.ToUTF32(input.ToString(), scratchIntsRef), BytesRef.DeepCopyOf(scratch));
}
FST <BytesRef> fst = builder.Finish();
return(new SynonymMap(fst, words, maxHorizontalContext));
}
public virtual void TestVariableBinary()
{
BaseDirectoryWrapper dir = NewFSDirectory(CreateTempDir("2BVariableBinary"));
if (dir is MockDirectoryWrapper)
{
((MockDirectoryWrapper)dir).Throttling = MockDirectoryWrapper.Throttling_e.NEVER;
}
IndexWriter w = new IndexWriter(dir, new IndexWriterConfig(TEST_VERSION_CURRENT, new MockAnalyzer(Random()))
.SetMaxBufferedDocs(IndexWriterConfig.DISABLE_AUTO_FLUSH).SetRAMBufferSizeMB(256.0).SetMergeScheduler(new ConcurrentMergeScheduler()).SetMergePolicy(NewLogMergePolicy(false, 10)).SetOpenMode(IndexWriterConfig.OpenMode_e.CREATE));
Document doc = new Document();
sbyte[] bytes = new sbyte[4];
ByteArrayDataOutput encoder = new ByteArrayDataOutput((byte[])(Array)bytes);
BytesRef data = new BytesRef(bytes);
BinaryDocValuesField dvField = new BinaryDocValuesField("dv", data);
doc.Add(dvField);
for (int i = 0; i < int.MaxValue; i++)
{
encoder.Reset((byte[])(Array)bytes);
encoder.WriteVInt(i % 65535); // 1, 2, or 3 bytes
data.Length = encoder.Position;
w.AddDocument(doc);
if (i % 100000 == 0)
{
Console.WriteLine("indexed: " + i);
Console.Out.Flush();
}
}
w.ForceMerge(1);
w.Dispose();
Console.WriteLine("verifying...");
Console.Out.Flush();
DirectoryReader r = DirectoryReader.Open(dir);
int expectedValue = 0;
ByteArrayDataInput input = new ByteArrayDataInput();
foreach (AtomicReaderContext context in r.Leaves)
{
AtomicReader reader = context.AtomicReader;
BytesRef scratch = new BytesRef(bytes);
BinaryDocValues dv = reader.GetBinaryDocValues("dv");
for (int i = 0; i < reader.MaxDoc; i++)
{
dv.Get(i, scratch);
input.Reset((byte[])(Array)scratch.Bytes, scratch.Offset, scratch.Length);
Assert.AreEqual(expectedValue % 65535, input.ReadVInt());
Assert.IsTrue(input.Eof());
expectedValue++;
}
}
r.Dispose();
dir.Dispose();
}
public virtual void TestVariableBinary([ValueSource(typeof(ConcurrentMergeSchedulers), "Values")]IConcurrentMergeScheduler scheduler)
{
BaseDirectoryWrapper dir = NewFSDirectory(CreateTempDir("2BVariableBinary"));
if (dir is MockDirectoryWrapper)
{
((MockDirectoryWrapper)dir).Throttling = MockDirectoryWrapper.Throttling_e.NEVER;
}
var config = new IndexWriterConfig(TEST_VERSION_CURRENT, new MockAnalyzer(Random()))
.SetMaxBufferedDocs(IndexWriterConfig.DISABLE_AUTO_FLUSH)
.SetRAMBufferSizeMB(256.0)
.SetMergeScheduler(scheduler)
.SetMergePolicy(NewLogMergePolicy(false, 10))
.SetOpenMode(IndexWriterConfig.OpenMode_e.CREATE);
IndexWriter w = new IndexWriter(dir, config);
Document doc = new Document();
var bytes = new byte[4];
ByteArrayDataOutput encoder = new ByteArrayDataOutput(bytes);
BytesRef data = new BytesRef(bytes);
BinaryDocValuesField dvField = new BinaryDocValuesField("dv", data);
doc.Add(dvField);
for (int i = 0; i < int.MaxValue; i++)
{
encoder.Reset(bytes);
encoder.WriteVInt(i % 65535); // 1, 2, or 3 bytes
data.Length = encoder.Position;
w.AddDocument(doc);
if (i % 100000 == 0)
{
Console.WriteLine("indexed: " + i);
Console.Out.Flush();
}
}
w.ForceMerge(1);
w.Dispose();
Console.WriteLine("verifying...");
Console.Out.Flush();
DirectoryReader r = DirectoryReader.Open(dir);
int expectedValue = 0;
ByteArrayDataInput input = new ByteArrayDataInput();
foreach (AtomicReaderContext context in r.Leaves)
{
AtomicReader reader = context.AtomicReader;
BytesRef scratch = new BytesRef(bytes);
BinaryDocValues dv = reader.GetBinaryDocValues("dv");
for (int i = 0; i < reader.MaxDoc; i++)
{
dv.Get(i, scratch);
input.Reset((byte[])(Array)scratch.Bytes, scratch.Offset, scratch.Length);
Assert.AreEqual(expectedValue % 65535, input.ReadVInt());
Assert.IsTrue(input.Eof());
expectedValue++;
}
}
r.Dispose();
dir.Dispose();
}
