[MSHADE-323] make shade plugin compatible with the flatten plugin

[gzm55]

The runtime injected pom will be set via MavenProject::setFile orMavenProject::setPomFile, including this plugin and the flatten plugin. So to chain these kind of plugins, when generating the d-r-p pom, this pr first tries to parse a model directly from MavenProject::getFile and fallback to a clone of MavenProject::getOriginalModel if no pom file set.

Following this checklist to help us incorporate your contribution quickly and easily:

• Make sure there is a JIRA issue filed
  for the change (usually before you start working on it). Trivial changes like typos do not
  require a JIRA issue. Your pull request should address just this issue, without
  pulling in other changes.
• Each commit in the pull request should have a meaningful subject line and body.
• Format the pull request title like [MSHADE-XXX] - Fixes bug in ApproximateQuantiles,
  where you replace MSHADE-XXX with the appropriate JIRA issue. Best practice
  is to use the JIRA issue title in the pull request title and in the first line of the
  commit message.
• Write a pull request description that is detailed enough to understand what the pull request does, how, and why.
• Run mvn clean verify to make sure basic checks pass. A more thorough check will
  be performed on your pull request automatically.
• You have run the integration tests successfully (mvn -Prun-its clean verify).

If your pull request is about ~20 lines of code you don't need to sign an
Individual Contributor License Agreement if you are unsure
please ask on the developers list.

To make clear that you license your contribution under
the Apache License Version 2.0, January 2004
you have to acknowledge this by using the following check-box.

• I hereby declare this contribution to be licenced under the Apache License Version 2.0, January 2004

• In any other case, please file an Apache Individual Contributor License Agreement.

[rmannibucau]

looks like an issue any plugin consuming the pom will get so it should likely be fixed in flatten plugin using setOriginalModel to be a global fix and not a local one.

[gzm55]

hi @rmannibucau , the in Maven, what is originalModel defined to be, a model of the original pom or a model of the current pom file? If defined as the latter one, the global solution of updating originalModel could be done in MavenProject::setPomFile or MavenProject::setFile.

[rmannibucau]

My guess is that it was the really original file but since maven is a chain (per module) it is in practise the current pom file - otherwise it can't really be used by any plugins and some do.
So guess the choice is:

1. keep original untouched but means no plugin should use it except help/debug ones (so fix is to always read getFile())
2. ensure flatten plugin updates original model to let downstream plugins consumes the post processing model

Personally I think the option 2 is what makes the most sense for the ecosystem (inter-plugin operability).

[gzm55]

ok, this pr is the first choice, which prefer a practice single source to get current raw model.

[rmannibucau]

@gzm55 this pr does not impl 1 since it should never use original model in case 1 but always file to enable to chain plugins.

[gzm55]

it is ok to not fall back to original model when getFile() is null, throwing an exception instead.

[rmannibucau]

Hmm, I just checked flatten code and it only works if updatePomFile is set to true no? So long story short, if you want it to work reliably you must ensure originalModel is set before shade plugin is executed. From a design point of view it is the responsability of flatten plugin to do that since it is the one hacking the user pom but an intermediary plugin setting it from a file when flatten plugin dumped its value would work too but I don't think there is any bug in shade there.

[gzm55]

it seems better to have a convention or rule that the original model should be equal to the pom file. the shade plugin also breaking this convention, it reverts the parent.relativePath after write the model to the file.

[rmannibucau]

shade plugins reset it cause it just needs to change it while dumping the pom (that said good point it breaks thread safety in concurrent builds).
agree shade plugin breaks this rules and happy to get it fixed, it is just not that important cause shade plugin is generally in the end of the chain (so dowstream consumers list is empty) and it also handles itself the pom rewriting but you are right.
long story short: I don't think we should re-read from the filesystem any pom model anytime in maven build otherwise the design itself is broken so I'm in favor of trying to not use the reader more than once per file.

hope it makes sense.