Use length of VarcharType and CharType to calculate EXPECTED_ENTRY_SIZE

trinodb · Jul 30, 2020 · 547eb1a · 547eb1a
1 parent deb9c13
commit 547eb1a
Show file tree

Hide file tree

Showing 4 changed files with 36 additions and 1 deletion.
diff --git a/presto-main/src/main/java/io/prestosql/type/TypeUtils.java b/presto-main/src/main/java/io/prestosql/type/TypeUtils.java
@@ -23,13 +23,15 @@
 import io.prestosql.spi.block.Block;
 import io.prestosql.spi.block.BlockBuilder;
 import io.prestosql.spi.type.ArrayType;
+import io.prestosql.spi.type.CharType;
 import io.prestosql.spi.type.FixedWidthType;
 import io.prestosql.spi.type.MapType;
 import io.prestosql.spi.type.RowType;
 import io.prestosql.spi.type.StandardTypes;
 import io.prestosql.spi.type.TimestampType;
 import io.prestosql.spi.type.TimestampWithTimeZoneType;
 import io.prestosql.spi.type.Type;
+import io.prestosql.spi.type.VarcharType;
 
 import java.lang.invoke.MethodHandle;
 import java.util.List;
@@ -57,6 +59,16 @@ public static int expectedValueSize(Type type, int defaultSize)
         if (type instanceof FixedWidthType) {
             return ((FixedWidthType) type).getFixedSize();
         }
+        // If bound on length of varchar or char is smaller than defaultSize, use that as expected size
+        // The data can take up to 4 bytes per character due to UTF-8 encoding, but we assume it is ASCII and only needs one byte.
+        if (type instanceof VarcharType) {
+            return ((VarcharType) type).getLength()
+                    .map(length -> Math.min(length, defaultSize))
+                    .orElse(defaultSize);
+        }
+        if (type instanceof CharType) {
+            return Math.min(((CharType) type).getLength(), defaultSize);
+        }
         return defaultSize;
     }
 

diff --git a/presto-spi/src/main/java/io/prestosql/spi/type/AbstractVariableWidthType.java b/presto-spi/src/main/java/io/prestosql/spi/type/AbstractVariableWidthType.java
@@ -24,7 +24,7 @@ public abstract class AbstractVariableWidthType
         extends AbstractType
         implements VariableWidthType
 {
-    private static final int EXPECTED_BYTES_PER_ENTRY = 32;
+    protected static final int EXPECTED_BYTES_PER_ENTRY = 32;
 
     protected AbstractVariableWidthType(TypeSignature signature, Class<?> javaType)
     {

diff --git a/presto-spi/src/main/java/io/prestosql/spi/type/CharType.java b/presto-spi/src/main/java/io/prestosql/spi/type/CharType.java
@@ -18,6 +18,7 @@
 import io.prestosql.spi.PrestoException;
 import io.prestosql.spi.block.Block;
 import io.prestosql.spi.block.BlockBuilder;
+import io.prestosql.spi.block.BlockBuilderStatus;
 import io.prestosql.spi.connector.ConnectorSession;
 
 import java.util.Objects;
@@ -108,6 +109,14 @@ public int compareTo(Block leftBlock, int leftPosition, Block rightBlock, int ri
         return compareChars(leftSlice, rightSlice);
     }
 
+    @Override
+    public BlockBuilder createBlockBuilder(BlockBuilderStatus blockBuilderStatus, int expectedEntries)
+    {
+        // If bound on length of char is smaller than EXPECTED_BYTES_PER_ENTRY, use that as expectedBytesPerEntry
+        // The data can take up to 4 bytes per character due to UTF-8 encoding, but we assume it is ASCII and only needs one byte.
+        return createBlockBuilder(blockBuilderStatus, expectedEntries, Math.min(length, EXPECTED_BYTES_PER_ENTRY));
+    }
+
     @Override
     public void appendTo(Block block, int position, BlockBuilder blockBuilder)
     {

diff --git a/presto-spi/src/main/java/io/prestosql/spi/type/VarcharType.java b/presto-spi/src/main/java/io/prestosql/spi/type/VarcharType.java
@@ -18,6 +18,7 @@
 import io.airlift.slice.Slices;
 import io.prestosql.spi.block.Block;
 import io.prestosql.spi.block.BlockBuilder;
+import io.prestosql.spi.block.BlockBuilderStatus;
 import io.prestosql.spi.connector.ConnectorSession;
 
 import java.util.Objects;
@@ -136,6 +137,19 @@ public int compareTo(Block leftBlock, int leftPosition, Block rightBlock, int ri
         return leftBlock.compareTo(leftPosition, 0, leftLength, rightBlock, rightPosition, 0, rightLength);
     }
 
+    @Override
+    public BlockBuilder createBlockBuilder(BlockBuilderStatus blockBuilderStatus, int expectedEntries)
+    {
+        return createBlockBuilder(
+                blockBuilderStatus,
+                expectedEntries,
+                getLength()
+                        // If bound on length is smaller than EXPECTED_BYTES_PER_ENTRY, use that as expectedBytesPerEntry
+                        // The data can take up to 4 bytes per character due to UTF-8 encoding, but we assume it is ASCII and only needs one byte.
+                        .map(length -> Math.min(length, EXPECTED_BYTES_PER_ENTRY))
+                        .orElse(EXPECTED_BYTES_PER_ENTRY));
+    }
+
     @Override
     public Optional<Range> getRange()
     {