You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
Commit 3398c18
Browse filesBrowse the repository at this point in the historyBrowse files
Say that the dictionary id section is empty, and add the page and column once
- An empty dictionary id section now fails with "the dictionary id section is empty, not even the bit width", checked once before reading groups.
- Remove the inner `IOException` wrapper in `initDataReader`, since `readPageV1` and `readPageV2` already add the page and column, so both exception types get one layer.
- Check that the dictionary page error includes the original message instead of parquet-java's wording, and use a plain `val` for the test thread pool.
Co-authored-by: Isaac <[email protected]>
Copy file name to clipboardExpand all lines: sql/core/src/test/scala/org/apache/spark/sql/execution/datasources/parquet/VectorizedRleValuesReaderSuite.scala
+2-2Lines changed: 2 additions & 2 deletions
Original file line number
Diff line number
Diff line change
@@ -54,7 +54,7 @@ class VectorizedRleValuesReaderSuite extends SparkFunSuite {
54
54
// stuck in a loop does not block the following ones. The decode loops do not check for
55
55
// interruption, so a stuck read keeps its thread busy until the JVM exits. After the first
56
56
// timeout, later reads fail without running, so that a regression leaves one busy thread.
0 commit comments