[ORC-346] Bug in TimestampColumnReader (or Writer)

• Type: Bug
• Status: Closed
• Priority: Blocker
• Resolution: Fixed
• Fix Version/s: 1.5.0
• Component/s: C++
• Assignee: Renat Valiullin
• Reporter: Renat Valiullin
• Created: 18/Apr/18 21:28
• Updated: 14/May/18 22:24
• Resolved: 27/Apr/18 22:44

Description

void TimestampColumnReader::next(ColumnVectorBatch& rowBatch, uint64_t numValues, char *notNull)

has the following code:

c++\src\ColumnReader.cc:338

int64_t writerTime = secsBuffer[i] + epochOffset;
secsBuffer[i] = writerTimezone.convertToUTC(writerTime);
if (secsBuffer[i] < 0 && nanoBuffer[i] != 0) {
secsBuffer[i] -= 1;
}

which likely leads to read wrong seconds value for PRE_1970 dates

Issue Links

• links to: GitHub Pull Request #253