[MPH-168] effective-pom should support multi-module project #76
Conversation
broodjetom
changed the title
|
I have tried some things with the I get the following output: Therefore, I currently have the output path hardcoded set to the target directory of the relevant module. We could go for the option of making the We could also choose to allow to specify a directory as Please, let me know what your preferred solution is. |
There was a problem hiding this comment.
Thanks for this effort!
I'd appreciate if the new property individual would not automatically set output.
Could you make it in such a way that individual would output to standard out, unless output was set? If individual=true and output are set, I would interpret output as relative to the base directory of the module for which each individual effective pom was generated.
src/main/java/org/apache/maven/plugins/help/EffectivePomMojo.java
Outdated
Show resolved
Hide resolved
src/main/java/org/apache/maven/plugins/help/EffectivePomMojo.java
Outdated
Show resolved
Hide resolved
When the user only specifies the individual flag, it will create individual effective POMs and print it to standard out. When the output parameter is specified, it is used to write the ouput to files, relative to the (sub)project the effective POM is for. The output parameter is parsed as a File so contains the full path. We replace the basedir of that path to get the actual location where the effective POM should be written to. That would fail in case of a relative path going up the tree. Therefore, if the basedir cannot be found, we take the name of the output path and use only that.
|
If have changed it to where the effective POMs are written to standard out when the output flag is not set. When the output flag is set, we need to interpret that as relative to the build directory of the (sub)project. A problem is actually that output flag will be parsed as a file, and thus contains the full absolute path (including the drive on Windows). Luckily, the project where we run the command from is available and provides a basedir. I have used this to do some testing: If you run the following command from the main module:
To then get the relative output I made it to remove the basedir from the output path, meaning you are left with Now run the following command from the main module:
Now the As a conclusion, you are now able to specify a relative path, but there is a constraint on relative paths going up the file tree. Please let me know if you see another solution? |
| { | ||
| String rawOutputPath = output.getPath(); | ||
| String rawBasedirPath = project.getBasedir().getPath(); | ||
| String result = rawOutputPath.contains( rawBasedirPath ) |
There was a problem hiding this comment.
This brute-force searching of strings is not the correct way to process paths. Even in the best of worlds, there are all sorts of cases this won't work with (e.g. if the paths are not normalized, and contains ../, etc. In addition you're using contains(), which even with this approach isn't correct (it looks even in the middle of the string).
Instead of searching raw strings, you should be using path manipulation methods for semantic Path objects. For example, there's a method made to do exactly what you want: Path.relativize(). And if you don't have a Path, you can create one from a string. You can also can convert from the old File model to the new Java NIO Path model and back.
If you're not used to working with semantic path objects, you might start at Oracle's The Path Class tutorial, which includes Path Operations explanations..
Following this checklist to help us incorporate your
contribution quickly and easily:
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.
[MPH-XXX] - Fixes bug in ApproximateQuantiles,where you replace
MPH-XXXwith the appropriate JIRA issue. Best practiceis to use the JIRA issue title in the pull request title and in the first line of the
commit message.
mvn clean verifyto make sure basic checks pass. A more thorough check willbe performed on your pull request automatically.
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. The ICLA has been signed and is sent to Apache secretary.
This PR closes https://issues.apache.org/jira/browse/MPH-168.
I have chosen to solve this issue using a new parameter
individual. I chose this name because you are generating an effective POM for each project/module individually. It is already possible to generate an effective POM across for each project/module, but it will put it in a single result/file. Please, advice on this naming if you like it or what it should be changed to.I have not yet built the support of specifying
outputin combination withindividual. For now, it will always output the effective POM to${projectDirectory}/target/effective.pom.xml. In the issue it mentions setting theoutputto${project.build.directory}/effective-pom.xml, but I'm not sure how to transform the variable part in it per project.IMO the code does not look very clean, I think some of it could be refactored, but I'm not sure in what way.