fix SplitRecordsForBlobSize (#19208)

zverevgeny · zverevgeny · commit 76b249aa3d2c · 2025-06-03T12:33:19.000Z
diff --git a/ydb/library/formats/arrow/splitter/stats.cpp b/ydb/library/formats/arrow/splitter/stats.cpp
@@ -45,7 +45,7 @@ std::vector<i64> TSimpleSerializationStat::SplitRecords(
 }
 
 std::vector<i64> TBatchSerializationStat::SplitRecordsForBlobSize(const i64 recordsCount, const ui64 blobSize) const {
-    if (!SerializedBytes) {
+    if (!SerializedBytes || blobSize < GetSerializedBytesPerRecord()) {
         return { recordsCount };
     }
     const ui32 recordsCountPerBlob = blobSize / GetSerializedBytesPerRecord();

Original file line number	Diff line number	Diff line change
`@@ -45,7 +45,7 @@ std::vector<i64> TSimpleSerializationStat::SplitRecords(`
`45`	`45`	`}`
`46`	`46`
`47`	`47`	`std::vector<i64> TBatchSerializationStat::SplitRecordsForBlobSize(const i64 recordsCount, const ui64 blobSize) const {`
`48`		`- if (!SerializedBytes) {`
	`48`	`+ if (!SerializedBytes \|\| blobSize < GetSerializedBytesPerRecord()) {`
`49`	`49`	`return { recordsCount };`
`50`	`50`	`}`
`51`	`51`	`const ui32 recordsCountPerBlob = blobSize / GetSerializedBytesPerRecord();`