nssalian commented on code in PR #17342:
URL: https://github.com/apache/iceberg/pull/17342#discussion_r3678641296
##########
parquet/src/test/java/org/apache/iceberg/parquet/TestVariantMetrics.java:
##########
@@ -266,13 +269,174 @@ public void testShreddedBinaryUpperBoundOverflow()
throws IOException {
assertThat(metrics.lowerBounds().get(2))
.extracting(b -> Variant.from(b).value().asObject().get(ROOT_FIELD))
-
.isEqualTo(Variants.of(BinaryUtil.truncateBinaryMin(ByteBuffer.wrap(bytes),
16)));
+ .isEqualTo(
+
Variants.of(BinaryUtil.truncateBinaryMin(ByteBuffer.wrap(BINARY_20_BYTES_ALL_FF),
16)));
assertThat(metrics.upperBounds().get(2))
.extracting(b -> Variant.from(b).value().asObject().get(ROOT_FIELD))
.isNull();
}
+ @Test
+ public void testShreddedBinaryBoundsTruncateLength() throws IOException {
+ // a per-column truncate(8) overrides the default 16-byte truncation
+ VariantValue value = Variants.of(ByteBuffer.wrap(BINARY_20_BYTES));
+
+ MetricsConfig metricsConfig =
+ MetricsConfig.from(
+ ImmutableMap.of(TableProperties.METRICS_MODE_COLUMN_CONF_PREFIX +
"var", "truncate(8)"),
+ SCHEMA,
+ null);
+
+ Metrics metrics =
+ writeParquetWithMetricsConfig(
+ (id, name) -> ParquetVariantUtil.toParquetSchema(value),
+ metricsConfig,
+ Variant.of(EMPTY, value),
+ Variant.of(EMPTY, Variants.ofNull()),
+ null);
+
+ assertThat(metrics.lowerBounds().get(2))
+ .extracting(b -> Variant.from(b).value().asObject().get(ROOT_FIELD))
+
.isEqualTo(Variants.of(BinaryUtil.truncateBinaryMin(ByteBuffer.wrap(BINARY_20_BYTES),
8)));
+
+ assertThat(metrics.upperBounds().get(2))
+ .extracting(b -> Variant.from(b).value().asObject().get(ROOT_FIELD))
+
.isEqualTo(Variants.of(BinaryUtil.truncateBinaryMax(ByteBuffer.wrap(BINARY_20_BYTES),
8)));
+ }
+
+ @Test
+ public void testShreddedBinaryBoundsFull() throws IOException {
+ // full mode leaves the bounds untruncated
+ VariantValue value = Variants.of(ByteBuffer.wrap(BINARY_20_BYTES));
+
+ MetricsConfig metricsConfig =
+ MetricsConfig.from(
+ ImmutableMap.of(TableProperties.METRICS_MODE_COLUMN_CONF_PREFIX +
"var", "full"),
+ SCHEMA,
+ null);
+
+ Metrics metrics =
+ writeParquetWithMetricsConfig(
+ (id, name) -> ParquetVariantUtil.toParquetSchema(value),
+ metricsConfig,
+ Variant.of(EMPTY, value),
+ Variant.of(EMPTY, Variants.ofNull()),
+ null);
+
+ assertThat(metrics.lowerBounds().get(2))
+ .extracting(b -> Variant.from(b).value().asObject().get(ROOT_FIELD))
+ .isEqualTo(Variants.of(ByteBuffer.wrap(BINARY_20_BYTES)));
+
+ assertThat(metrics.upperBounds().get(2))
+ .extracting(b -> Variant.from(b).value().asObject().get(ROOT_FIELD))
+ .isEqualTo(Variants.of(ByteBuffer.wrap(BINARY_20_BYTES)));
+ }
+
+ @Test
+ public void testShreddedBinaryBoundsCounts() throws IOException {
+ // counts mode drops shredded bounds
+ VariantValue value = Variants.of(ByteBuffer.wrap(BINARY_20_BYTES));
+
+ MetricsConfig metricsConfig =
+ MetricsConfig.from(
+ ImmutableMap.of(TableProperties.METRICS_MODE_COLUMN_CONF_PREFIX +
"var", "counts"),
+ SCHEMA,
+ null);
+
+ Metrics metrics =
+ writeParquetWithMetricsConfig(
+ (id, name) -> ParquetVariantUtil.toParquetSchema(value),
+ metricsConfig,
+ Variant.of(EMPTY, value),
+ Variant.of(EMPTY, Variants.ofNull()),
+ null);
+
+ assertThat(metrics.valueCounts()).containsKey(2);
+ assertThat(metrics.lowerBounds()).doesNotContainKey(2);
+ assertThat(metrics.upperBounds()).doesNotContainKey(2);
+ }
+
+ @Test
+ public void testShreddedStringBoundsTruncateLength() throws IOException {
+ // a per-column truncate(8) overrides the default 16-char truncation
+ VariantValue value = Variants.of("iceberg_variant");
+
+ MetricsConfig metricsConfig =
+ MetricsConfig.from(
+ ImmutableMap.of(TableProperties.METRICS_MODE_COLUMN_CONF_PREFIX +
"var", "truncate(8)"),
+ SCHEMA,
+ null);
+
+ Metrics metrics =
+ writeParquetWithMetricsConfig(
+ (id, name) -> ParquetVariantUtil.toParquetSchema(value),
+ metricsConfig,
+ Variant.of(EMPTY, value),
+ Variant.of(EMPTY, Variants.ofNull()),
+ null);
+
+ assertThat(metrics.lowerBounds().get(2))
+ .extracting(b -> Variant.from(b).value().asObject().get(ROOT_FIELD))
+
.isEqualTo(Variants.of(UnicodeUtil.truncateStringMin("iceberg_variant", 8)));
+
+ assertThat(metrics.upperBounds().get(2))
+ .extracting(b -> Variant.from(b).value().asObject().get(ROOT_FIELD))
+
.isEqualTo(Variants.of(UnicodeUtil.truncateStringMax("iceberg_variant", 8)));
+ }
+
+ @Test
+ public void testShreddedStringBoundsFull() throws IOException {
+ // full mode leaves the string bound untruncated
+ VariantValue value = Variants.of("iceberg_variant");
Review Comment:
Good catch @uros-b. Thanks. Bumped to `"iceberg_variant_full"` (20 chars) so
the default 16-char truncation and full mode produce different bounds
--
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.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
