This is an automated email from the ASF dual-hosted git repository.
lzljs3620320 pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/paimon.git
The following commit(s) were added to refs/heads/master by this push:
new e18f6ed6fd [parquet] Fix minor format codes in parquet readers
e18f6ed6fd is described below
commit e18f6ed6fde0f3b10170540aebefa0d437bb19aa
Author: Jingsong <[email protected]>
AuthorDate: Sun Dec 8 22:46:44 2024 +0800
[parquet] Fix minor format codes in parquet readers
---
.../paimon/format/parquet/position/CollectionPosition.java | 6 +++---
.../apache/paimon/format/parquet/position/LevelDelegation.java | 1 +
.../paimon/format/parquet/reader/AbstractColumnReader.java | 5 +----
.../apache/paimon/format/parquet/reader/BooleanColumnReader.java | 5 -----
.../format/parquet/reader/NestedPrimitiveColumnReader.java | 4 +---
.../apache/paimon/format/parquet/reader/ParquetReadState.java | 9 +++------
.../paimon/format/parquet/reader/TimestampColumnReader.java | 5 +++--
.../java/org/apache/paimon/format/parquet/type/ParquetField.java | 1 +
8 files changed, 13 insertions(+), 23 deletions(-)
diff --git
a/paimon-format/src/main/java/org/apache/paimon/format/parquet/position/CollectionPosition.java
b/paimon-format/src/main/java/org/apache/paimon/format/parquet/position/CollectionPosition.java
index e72a4280f4..beb5de7a92 100644
---
a/paimon-format/src/main/java/org/apache/paimon/format/parquet/position/CollectionPosition.java
+++
b/paimon-format/src/main/java/org/apache/paimon/format/parquet/position/CollectionPosition.java
@@ -22,14 +22,14 @@ import javax.annotation.Nullable;
/** To represent collection's position in repeated type. */
public class CollectionPosition {
+
@Nullable private final boolean[] isNull;
private final long[] offsets;
-
private final long[] length;
-
private final int valueCount;
- public CollectionPosition(boolean[] isNull, long[] offsets, long[] length,
int valueCount) {
+ public CollectionPosition(
+ @Nullable boolean[] isNull, long[] offsets, long[] length, int
valueCount) {
this.isNull = isNull;
this.offsets = offsets;
this.length = length;
diff --git
a/paimon-format/src/main/java/org/apache/paimon/format/parquet/position/LevelDelegation.java
b/paimon-format/src/main/java/org/apache/paimon/format/parquet/position/LevelDelegation.java
index 25bbedc861..8e30d90ba2 100644
---
a/paimon-format/src/main/java/org/apache/paimon/format/parquet/position/LevelDelegation.java
+++
b/paimon-format/src/main/java/org/apache/paimon/format/parquet/position/LevelDelegation.java
@@ -20,6 +20,7 @@ package org.apache.paimon.format.parquet.position;
/** To delegate repetition level and definition level. */
public class LevelDelegation {
+
private final int[] repetitionLevel;
private final int[] definitionLevel;
diff --git
a/paimon-format/src/main/java/org/apache/paimon/format/parquet/reader/AbstractColumnReader.java
b/paimon-format/src/main/java/org/apache/paimon/format/parquet/reader/AbstractColumnReader.java
index 5e3f4a7e6a..d4a0ab039b 100644
---
a/paimon-format/src/main/java/org/apache/paimon/format/parquet/reader/AbstractColumnReader.java
+++
b/paimon-format/src/main/java/org/apache/paimon/format/parquet/reader/AbstractColumnReader.java
@@ -69,9 +69,6 @@ public abstract class AbstractColumnReader<VECTOR extends
WritableColumnVector>
/** If true, the current page is dictionary encoded. */
private boolean isCurrentPageDictionaryEncoded;
- /** Total values in the current page. */
- // private int pageValueCount;
-
/**
* Helper struct to track intermediate states while reading Parquet pages
in the column chunk.
*/
@@ -90,7 +87,7 @@ public abstract class AbstractColumnReader<VECTOR extends
WritableColumnVector>
*/
/** Run length decoder for data and dictionary. */
- protected RunLengthDecoder runLenDecoder;
+ RunLengthDecoder runLenDecoder;
/** Data input stream. */
ByteBufferInputStream dataInputStream;
diff --git
a/paimon-format/src/main/java/org/apache/paimon/format/parquet/reader/BooleanColumnReader.java
b/paimon-format/src/main/java/org/apache/paimon/format/parquet/reader/BooleanColumnReader.java
index 83d3c5a07d..4355392bf5 100644
---
a/paimon-format/src/main/java/org/apache/paimon/format/parquet/reader/BooleanColumnReader.java
+++
b/paimon-format/src/main/java/org/apache/paimon/format/parquet/reader/BooleanColumnReader.java
@@ -42,11 +42,6 @@ public class BooleanColumnReader extends
AbstractColumnReader<WritableBooleanVec
checkTypeName(PrimitiveType.PrimitiveTypeName.BOOLEAN);
}
- @Override
- protected boolean supportLazyDecode() {
- return true;
- }
-
@Override
protected void afterReadPage() {
bitOffset = 0;
diff --git
a/paimon-format/src/main/java/org/apache/paimon/format/parquet/reader/NestedPrimitiveColumnReader.java
b/paimon-format/src/main/java/org/apache/paimon/format/parquet/reader/NestedPrimitiveColumnReader.java
index 7db7aedbf6..f0a82a6d71 100644
---
a/paimon-format/src/main/java/org/apache/paimon/format/parquet/reader/NestedPrimitiveColumnReader.java
+++
b/paimon-format/src/main/java/org/apache/paimon/format/parquet/reader/NestedPrimitiveColumnReader.java
@@ -65,6 +65,7 @@ import static org.apache.parquet.column.ValuesType.VALUES;
/** Reader to read nested primitive column. */
public class NestedPrimitiveColumnReader implements
ColumnReader<WritableColumnVector> {
+
private static final Logger LOG =
LoggerFactory.getLogger(NestedPrimitiveColumnReader.class);
private final IntArrayList repetitionLevelList = new IntArrayList(0);
@@ -95,9 +96,6 @@ public class NestedPrimitiveColumnReader implements
ColumnReader<WritableColumnV
private IntIterator definitionLevelColumn;
private ParquetDataColumnReader dataColumn;
- /** Total values in the current page. */
- // private int pageValueCount;
-
/**
* Helper struct to track intermediate states while reading Parquet pages
in the column chunk.
*/
diff --git
a/paimon-format/src/main/java/org/apache/paimon/format/parquet/reader/ParquetReadState.java
b/paimon-format/src/main/java/org/apache/paimon/format/parquet/reader/ParquetReadState.java
index a600367682..aa89ea9821 100644
---
a/paimon-format/src/main/java/org/apache/paimon/format/parquet/reader/ParquetReadState.java
+++
b/paimon-format/src/main/java/org/apache/paimon/format/parquet/reader/ParquetReadState.java
@@ -25,6 +25,7 @@ import java.util.PrimitiveIterator;
/** Parquet reader state for column index. */
public class ParquetReadState {
+
/** A special row range used when there is no row indexes (hence all rows
must be included). */
private static final RowRange MAX_ROW_RANGE = new RowRange(Long.MIN_VALUE,
Long.MAX_VALUE);
@@ -105,15 +106,11 @@ public class ParquetReadState {
}
public boolean isFinished() {
- return this.currentRange.equals(this.END_ROW_RANGE);
+ return this.currentRange.equals(END_ROW_RANGE);
}
public boolean isMaxRange() {
- return this.currentRange.equals(this.MAX_ROW_RANGE);
- }
-
- public RowRange getCurrentRange() {
- return currentRange;
+ return this.currentRange.equals(MAX_ROW_RANGE);
}
/** Advance to the next range. */
diff --git
a/paimon-format/src/main/java/org/apache/paimon/format/parquet/reader/TimestampColumnReader.java
b/paimon-format/src/main/java/org/apache/paimon/format/parquet/reader/TimestampColumnReader.java
index 8767173315..d6ac96ea44 100644
---
a/paimon-format/src/main/java/org/apache/paimon/format/parquet/reader/TimestampColumnReader.java
+++
b/paimon-format/src/main/java/org/apache/paimon/format/parquet/reader/TimestampColumnReader.java
@@ -36,8 +36,9 @@ import static
org.apache.paimon.utils.Preconditions.checkArgument;
/**
* Timestamp {@link ColumnReader}. We only support INT96 bytes now,
julianDay(4) + nanosOfDay(8).
- * See
https://github.com/apache/parquet-format/blob/master/DataTypes.md#timestamp
TIMESTAMP_MILLIS
- * and TIMESTAMP_MICROS are the deprecated ConvertedType.
+ * See <a
+ *
href="https://github.com/apache/parquet-format/blob/master/LogicalTypes.md#timestamp";>Parquet
+ * Timestamp</a> TIMESTAMP_MILLIS and TIMESTAMP_MICROS are the deprecated
ConvertedType.
*/
public class TimestampColumnReader extends
AbstractColumnReader<WritableTimestampVector> {
diff --git
a/paimon-format/src/main/java/org/apache/paimon/format/parquet/type/ParquetField.java
b/paimon-format/src/main/java/org/apache/paimon/format/parquet/type/ParquetField.java
index 94fe6b91d9..291e9ebbce 100644
---
a/paimon-format/src/main/java/org/apache/paimon/format/parquet/type/ParquetField.java
+++
b/paimon-format/src/main/java/org/apache/paimon/format/parquet/type/ParquetField.java
@@ -22,6 +22,7 @@ import org.apache.paimon.types.DataType;
/** Field that represent parquet's field type. */
public abstract class ParquetField {
+
private final DataType type;
private final int repetitionLevel;
private final int definitionLevel;
