voonhous commented on code in PR #19510:
URL: https://github.com/apache/hudi/pull/19510#discussion_r3871961074
##########
hudi-hadoop-mr/src/main/java/org/apache/hudi/hadoop/HoodieParquetInputFormat.java:
##########
@@ -210,10 +211,10 @@ private RecordReader<NullWritable, ArrayWritable>
createBootstrappingRecordReade
LOG.info("colNameWithTypes ={}, Num Entries ={}", colNameWithTypes,
colNameWithTypes.size());
- if (hoodieColsProjected.isEmpty()) {
- return getRecordReaderInternal(eSplit.getBootstrapFileSplit(), job,
reporter);
- } else if (externalColsProjected.isEmpty()) {
- return getRecordReaderInternal(split, job, reporter);
+ Option<FileSplit> singleSplit = resolveSingleFileSplit(eSplit,
!hoodieColsProjected.isEmpty(),
Review Comment:
Non-blocking, since this was optional to begin with. Flagging only because
the reply and the PR body describe it as done and it is not in the branch: at
head `1d6d8bb` (and at `e5190c9`, the head when this reply was posted) lines
193-205 still build `projectedColsWithIndex` from the `rawColIds`/`rawColNames`
zip, and line 214 passes `!hoodieColsProjected.isEmpty()` /
`!externalColsProjected.isEmpty()` into `resolveSingleFileSplit`.
Either is fine: push the names-only derivation, or leave the code as is and
drop the "deleting the id/name zip" / "Why the id/name zip went" / Risk Level
lines from the body so it matches the diff. Could we do one of the two?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
