Class VarByteChunkForwardIndexWriterV4

  • All Implemented Interfaces:
    Closeable, AutoCloseable, VarByteChunkWriter

    @NotThreadSafe
    public class VarByteChunkForwardIndexWriterV4
    extends Object
    implements VarByteChunkWriter
    Chunk-based raw (non-dictionary-encoded) forward index writer where each chunk contains variable number of docs, and the entries are variable length.

    The layout of the file is as follows:

    • Header Section
      • File format version (int)
      • Target decompressed chunk size (int)
      • Compression type enum value (int)
      • Start offset of chunk data (int)
      • Data header (for each chunk)
        • First docId in the chunk (int), where MSB is used to mark huge chunk
        • Start offset of the chunk (unsigned int)
    • Individual Chunks
      • Regular chunk
        • Header Section: start offsets (stored as int) of the entry within the data section
        • Data Section
      • Huge chunk: contains one single value