[gentoo-soc] Weekly Report: Big Data Infrastructure and Maven Overlay in Week 2

[Zhang Zongyu]

Hello,

During the past week, I have modified 'tree.sh' to fit in the new pipeline.
There are some new features.
1. Automatically detect the newest maven artifact version by parsing
metadata.xml of the artifact.
2. Compare the required maven version range and the artifact version
mentioned above, which makes sure the package satisfies the
requirement.
3. Force tree.sh to use a specified ${PN}, because artifacts sometimes
change their groupId and artifactId during version bump.
The related code is located in my overlay.[1]
A new description of Makefile Phases can be found in [2].

There are few weird pom.xml files, e.g. 'jaxb-*' usually cannot
produce a proper XML file and 'jruby-stdlib' depends on some
non-existing artifacts.
I decided to write them a proper ebuild by hand, which prevents us
from adding strange code into java-ebuilder.

And there some potential feature enhancement of java-ebuilder.
A list of features is located in [4].

Finally, an overlay to emerge spark-core is located in [5].
It has not been tested, so I am sure there will be tons of problems.
I will try to fix them next week:)

Regards,
Zhang Zongyu


[1] my overlay that contains patches of java-ebuilder
https://github.com/6-6-6/JEbuilder
[2] Google Sheet showing different stages of Makefile
https://docs.google.com/spreadsheets/d/1UwyjzuP1zjC_M1OcpPCSTMJmjhigvWAKKl9IG3h4AMg/edit?usp=sharing
[3] Time Table
https://docs.google.com/spreadsheets/d/1R36-bjkmEy0uhigpWTNAQcqy_slO3B1q_YLX1kbEE_c/edit?usp=sharing
[4] Memo of Features
https://github.com/6-6-6/JEbuilder/blob/master/Memo.md
[5] spark-overlay
https://github.com/6-6-6/spark-overlay