Age | Commit message (Collapse) | Author |
|
star column.
|
|
|
|
|
|
|
|
* Fix an issue that causes calls to splitAndTransferTo malfunction on repeated vector types
|
|
|
|
|
|
|
|
ii) also display max record size allowed
|
|
incoming batch has empty values for repeated fields
|
|
|
|
expression that returns a repeated scalar value.
|
|
|
|
dynamically
|
|
|
|
|
|
up resources ii) FragmentExecutor should clean up gracefully at faulty & non-faulty runs regardless
|
|
the current container.
|
|
|
|
other vectors
|
|
partition names
|
|
print millis
|
|
->decimal28/decimal38
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Hive UDFs
|
|
|
|
converted existing binary files in the test-data directory.
|
|
|
|
Added Gantt chart
Improved organization
Decrease number of tables and overall page size
Implement general javascript functions for profile processing
|
|
the expressions appropriately based on the annotated expressions created by planner.
|
|
|
|
|
|
|
|
|
|
Includes a fairly substantial refactoring of the overall reader structure.
Fix records counts expected in parquet read tests, previously the wrong records counts were being expected and messages were sent to the logger, but in the form of debug messages. not errors, so they were not flagging when the tests were failing.
After review: Rename PageReadStatus to PageReader, removed underscore from package name column_readers, address review comment from parth in VarLengthColumnReaders.
Fix regression in nullable varlength reader for parquet. Change output names in parquet writer/reader tests to prevent issues with files not being deleted correctly during tests.
|
|
unnecessary exchange operator.
|
|
Optiq).
|
|
|
|
|
|
class and update FilterPrule to use all instead of best to work with Optiq 0.9.
|
|
|
|
while splitting JSON records
|
|
output type
|
|
rather than from the file system
|