nastra commented on a change in pull request #2826:
URL: https://github.com/apache/iceberg/pull/2826#discussion_r677161667
##########
File path: build.gradle
##########
@@ -247,13 +250,30 @@ project(':iceberg-data') {
exclude group: 'org.slf4j', module: 'slf4j-log4j12'
}
- testCompile("org.apache.hadoop:hadoop-client") {
+ implementation("org.apache.orc:orc-core::nohive") {
Review comment:
> These are no longer pulled in via iceberg-orc and iceberg-core?
correct, because `implementation` allows now more fine-grained control over
dependencies being pulled in, which wasn't the case previously with `compile`
(you got more dependencies than you were asking for). Now one could probably
achieve the same thing by pulling in certain dependencies via `api`, but that
means you're making those dependencies explicitly part of your API. The Gradle
docs also state that one should generally prefer `implementation` over `api`,
meaning it is better to explicitly state all the dependencies that are really
required, so that you don't get surprises with non-expected dependencies and
such.
> Do we need an implementation dependency on Parquet as well?
already added it in
https://github.com/nastra/iceberg/blob/6c308fd66a9f33ce31b5abf8de61dd8269fda4b5/build.gradle#L261
--
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]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
