[GitHub] [iotdb] JackieTien97 commented on a change in pull request #2184: [IOTDB-1081] New TsFile Format

[GitBox <gi...@apache.org>]

JackieTien97 commented on a change in pull request #2184:
URL: https://github.com/apache/iotdb/pull/2184#discussion_r559931528



##########
File path: tsfile/src/main/java/org/apache/iotdb/tsfile/file/header/ChunkHeader.java
##########
@@ -33,101 +34,119 @@
 
 public class ChunkHeader {
 
+
+  /**
+   * 1 means this chunk has more than one page, so each page has its own page statistic 4 means this
+   * chunk has only one page, and this page has no page statistic
+   */
+  private byte chunkType;
   private String measurementID;
   private int dataSize;
   private TSDataType dataType;
   private CompressionType compressionType;
   private TSEncoding encodingType;
-  private int numOfPages;
 
-  // this field does not need to be serialized.
+  // the following fields do not need to be serialized.
+  private int numOfPages;
   private int serializedSize;
 
   public ChunkHeader(String measurementID, int dataSize, TSDataType dataType,
-      CompressionType compressionType,
-      TSEncoding encoding, int numOfPages) {
-    this(measurementID, dataSize, getSerializedSize(measurementID), dataType, compressionType,
-        encoding, numOfPages);
+      CompressionType compressionType, TSEncoding encoding, int numOfPages) {
+    this(numOfPages <= 1 ? MetaMarker.ONLY_ONE_PAGE_CHUNK_HEADER : MetaMarker.CHUNK_HEADER,
+        measurementID, dataSize, getSerializedSize(measurementID, dataSize), dataType,
+        compressionType,
+        encoding);
+    this.numOfPages = numOfPages;
   }
 
-  private ChunkHeader(String measurementID, int dataSize, int headerSize, TSDataType dataType,
-      CompressionType compressionType, TSEncoding encoding, int numOfPages) {
+  public ChunkHeader(byte chunkType, String measurementID, int dataSize, TSDataType dataType,
+      CompressionType compressionType, TSEncoding encoding) {
+    this(chunkType, measurementID, dataSize, getSerializedSize(measurementID, dataSize), dataType,
+        compressionType, encoding);
+  }
+
+  public ChunkHeader(byte chunkType, String measurementID, int dataSize, int headerSize,
+      TSDataType dataType, CompressionType compressionType, TSEncoding encoding) {
+    this.chunkType = chunkType;
     this.measurementID = measurementID;
     this.dataSize = dataSize;
     this.dataType = dataType;
     this.compressionType = compressionType;
-    this.numOfPages = numOfPages;
     this.encodingType = encoding;
     this.serializedSize = headerSize;
   }
 
+  /**
+   * the exact serialized size of chunk header
+   */
+  public static int getSerializedSize(String measurementID, int dataSize) {
+    int measurementIdLength = measurementID.getBytes(TSFileConfig.STRING_CHARSET).length;
+    return Byte.BYTES // chunkType
+        + ReadWriteForEncodingUtils.varIntSize(measurementIdLength) // measurementID length
+        + measurementIdLength // measurementID
+        + ReadWriteForEncodingUtils.uVarIntSize(dataSize) // dataSize
+        + TSDataType.getSerializedSize() // dataType
+        + CompressionType.getSerializedSize() // compressionType
+        + TSEncoding.getSerializedSize(); // encodingType
+  }
+
+  /**
+   * The estimated serialized size of chunk header. Only used when we don't know the actual dataSize
+   * attribute
+   */
   public static int getSerializedSize(String measurementID) {
-    return Byte.BYTES // marker
-        + Integer.BYTES // measurementID length
-        + measurementID.getBytes(TSFileConfig.STRING_CHARSET).length // measurementID
-        + Integer.BYTES // dataSize
+
+    int measurementIdLength = measurementID.getBytes(TSFileConfig.STRING_CHARSET).length;
+    return  Byte.BYTES // chunkType
+        + ReadWriteForEncodingUtils.varIntSize(measurementIdLength) // measurementID length
+        + measurementIdLength // measurementID
+        + Integer.BYTES + 1 // uVarInt dataSize
         + TSDataType.getSerializedSize() // dataType
         + CompressionType.getSerializedSize() // compressionType
-        + TSEncoding.getSerializedSize() // encodingType
-        + Integer.BYTES; // numOfPages
+        + TSEncoding.getSerializedSize(); // encodingType
   }
 
   /**
    * deserialize from inputStream.
-   *
-   * @param markerRead Whether the marker of the CHUNK_HEADER has been read
    */
-  public static ChunkHeader deserializeFrom(InputStream inputStream, boolean markerRead)
+  public static ChunkHeader deserializeFrom(InputStream inputStream, byte chunkType)

Review comment:
       Yes, but we can get that from the method signature, because this method pass the marker as a parameter. And, i will add some comments for these two methods




----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org