Release assembly jar - Githubissues

eto-ai / rikai

Parquet-based ML data format optimized for working with unstructured data

https://rikai.readthedocs.io/en/latest/

Apache License 2.0

136 stars 19 forks source link

Release assembly jar #625

Closed da-tubi closed 2 years ago

da-tubi commented 2 years ago

closes #386

sbt publishLocal
- $HOME/.ivy2/local/ai.eto/rikai_2.12/0.1.9-SNAPSHOT/jars/rikai-spark312-assembly_2.12.jar
SPARK_VERSION=3.2.1 sbt publishLocal
- $HOME/.ivy2/local/ai.eto/rikai_2.12/0.1.9-SNAPSHOT/jars/rikai-spark321-assembly_2.12.jar
sbt clean assembly
- rikai-spark312-assembly_2.12-0.1.9-SNAPSHOT.jar
SPARK_VERSION=3.2.1 sbt clean assembly
- rikai-spark321-assembly_2.12-0.1.9-SNAPSHOT.jar

Renkai commented 2 years ago

Would other people get it from maven central? If do so, a sbt definition like "ai.eto" %% "rikai-spark321-assembly" % "0.1.8" would be translated to 0.1.8/jars/rikai-spark321-assembly_2.12.jar. (Not very sure, I guess so) If we don't need it on maven central, then the current name format should be ok.

changhiskhan commented 2 years ago

Would other people get it from maven central? If do so, a sbt definition like "ai.eto" %% "rikai-spark321-assembly" % "0.1.8" would be translated to 0.1.8/jars/rikai-spark321-assembly_2.12.jar. (Not very sure, I guess so) If we don't need it on maven central, then the current name format should be ok.

I think we should keep this option open (for people to get it from maven central)

da-tubi commented 2 years ago

@changhiskhan please review, I've updated it for a while.