From ab259568ce223ec74ee67b625ff6e67f5a9e5a09 Mon Sep 17 00:00:00 2001 From: "zhongheng.gy" Date: Mon, 25 May 2026 14:13:47 +0800 Subject: [PATCH] [format] Pass through parquet statistics and page-size-check config in RowDataParquetBuilder --- .../parquet/writer/RowDataParquetBuilder.java | 18 +++++++++++++++++- 1 file changed, 17 insertions(+), 1 deletion(-) diff --git a/paimon-format/src/main/java/org/apache/paimon/format/parquet/writer/RowDataParquetBuilder.java b/paimon-format/src/main/java/org/apache/paimon/format/parquet/writer/RowDataParquetBuilder.java index 1a66a55130e1..7d72c8a1a91a 100644 --- a/paimon-format/src/main/java/org/apache/paimon/format/parquet/writer/RowDataParquetBuilder.java +++ b/paimon-format/src/main/java/org/apache/paimon/format/parquet/writer/RowDataParquetBuilder.java @@ -96,7 +96,23 @@ public ParquetWriter createWriter(OutputFile out, String compressio .withBloomFilterEnabled( conf.getBoolean( ParquetOutputFormat.BLOOM_FILTER_ENABLED, - ParquetProperties.DEFAULT_BLOOM_FILTER_ENABLED)); + ParquetProperties.DEFAULT_BLOOM_FILTER_ENABLED)) + .withMinRowCountForPageSizeCheck( + conf.getInt( + ParquetOutputFormat.MIN_ROW_COUNT_FOR_PAGE_SIZE_CHECK, + ParquetProperties.DEFAULT_MINIMUM_RECORD_COUNT_FOR_CHECK)) + .withMaxRowCountForPageSizeCheck( + conf.getInt( + ParquetOutputFormat.MAX_ROW_COUNT_FOR_PAGE_SIZE_CHECK, + ParquetProperties.DEFAULT_MAXIMUM_RECORD_COUNT_FOR_CHECK)) + .withStatisticsTruncateLength( + conf.getInt( + ParquetOutputFormat.STATISTICS_TRUNCATE_LENGTH, + ParquetProperties.DEFAULT_STATISTICS_TRUNCATE_LENGTH)) + .withColumnIndexTruncateLength( + conf.getInt( + ParquetOutputFormat.COLUMN_INDEX_TRUNCATE_LENGTH, + ParquetProperties.DEFAULT_COLUMN_INDEX_TRUNCATE_LENGTH)); new ColumnConfigParser() .withColumnConfig( ParquetOutputFormat.ENABLE_DICTIONARY,