public IList <Document> ReadDocuments(IList <int> documentIds)
{
var addresses = new List <BlockInfo>(documentIds.Count);
foreach (var id in documentIds)
{
addresses.Add(new BlockInfo(id * _blockSize, _blockSize));
}
var docAddresses = _addressReader.Read(addresses);
var index = 0;
var documents = new List <Document>(documentIds.Count);
_stream.Seek(Version.KeyIndexOffset, SeekOrigin.Begin);
var keyIndex = TableSerializer.ReadKeyIndex(_stream, Version.KeyIndexSize);
using (var documentReader = new DocumentReader(
_stream, Version.Compression, keyIndex, leaveOpen: true))
foreach (var document in documentReader.Read(docAddresses))
{
document.Id = documentIds[index++];
documents.Add(document);
}
return(documents);
}
public DocHash Read(int docId)
{
var pos = docId * TableSerializer.SizeOfDocHash() + _offset;
_stream.Seek(pos, SeekOrigin.Begin);
var hash = TableSerializer.DeserializeDocHash(_stream);
return(hash);
}
public DtblStream(Stream stream, BatchInfo ix, int skip = 0, int take = int.MaxValue)
: base(ix.PrimaryKeyFieldName)
{
_dataFile = stream;
_ix = ix;
_dataFile.Seek(_ix.KeyIndexOffset, SeekOrigin.Begin);
var keyIndex = TableSerializer.ReadKeyIndex(_dataFile, _ix.KeyIndexSize);
_hashReader = new DocHashReader(_dataFile, _ix.DocHashOffset, leaveOpen: false);
_addressReader = new DocumentAddressReader(_dataFile, _ix.DocAddressesOffset);
_documentReader = new DocumentReader(_dataFile, _ix.Compression, keyIndex, leaveOpen: false);
_skip = skip;
_take = take;
}
public static BatchInfo Load(string fileName)
{
var time = new Stopwatch();
time.Start();
BatchInfo ix;
using (var fs = new FileStream(fileName, FileMode.Open, FileAccess.Read, FileShare.Read))
{
ix = TableSerializer.DeserializeBatchInfo(fs);
}
Log.DebugFormat("loaded ix in {0}", time.Elapsed);
return(ix);
}
protected override Document Deserialize(long offset, int size, Stream stream)
{
stream.Seek(_offset + offset, SeekOrigin.Begin);
return(TableSerializer.DeserializeDocument(stream, size, _compression, _keyIndex));
}
protected override IList <DocumentPosting> Deserialize(long offset, int size, Stream stream)
{
stream.Seek(offset, SeekOrigin.Begin);
return(TableSerializer.DeserializePostings(stream, size));
}
