Is there a way to import rascal modules from other rascal projects in your eclipse work-space? If so, how?
Yes, like normal eclipse projects, you have to put this dependency in your build path. Internally, Rascal projects are normal java projects with rascal added on top. This is so that you can mix java and rascal code.
Open your project's settings, and open: Java > Compiler > Building, and in the Projects tab, add the other project.
Remember to restart the REPL/Terminal for this to take effect
Related
I have two Eclipse plugin Xtext(2.11) projects.(Called Project A and B) Project B grammar file is dependent on Project A grammar file. Project A's grammar file extension is a and Project B is b.
I am able to generate the language server for Project B. I want to generate the single language server for both projects.(Currently, in project B it includes project A.)
Depend on extension type it needs invoke respective Xtext grammar functionality.
In above scenario, I have below questions:
Does Xtext can handle this kind of scenario seamlessly? Do I need to generate the language server for Project A also and need to add in Project B? What is the best way to do this?
What kind of Xtext changes required?
What kind of changes required at client side also? like in Visual code/ Che?
Xtext LS can be used with multiple Xtext languages. It uses java service loading to look up language setups: https://github.com/eclipse/xtext-core/blob/master/org.eclipse.xtext/src/org/eclipse/xtext/resource/ResourceServiceProviderServiceLoader.xtend#L25. In ide project of your language you should be able to find META-INF/services/org.ecalipse.xtext.ISetup file that says java what implementations should be used for ISetup interface.
Having jars for ide projects on the classpath should be enough. No changes are required.
VS code and Che does not care about implementation details of Xtext LS, but both languages should be registered for them
I develop Dart with WebStorm. I have a server-project, a client-project and some shared code.
Is it possible to create a project solution, containing several Dart projects (like you can do in VisualStudio)? Or is it common to develop each project in it's own WebStorm window?
You can configure as many content roots as you want in one WebStorm project, see Settings (Preferences) | Directories.
See also https://www.jetbrains.com/help/webstorm/directories.html and https://www.jetbrains.com/help/webstorm/content-root.html
You can put the project directories together in a directory and then open that directory instead.
There are also ways with project root configurations to do that with projects that are not in the same directory but I did't bother to try to understand how this exactly work.
I moved away from this setup and use different windows for different projects a while ago because I don't want to maintain project configurations. I have many Dart projects that are reusable packages that I use in many other projects.
Using one window for one Dart package seems to be the easiest setup.
I configured keyboard shortcuts (there are probably defaults anyway) for easier switching between WebStorm windows.
By default, Maven standard directory layout has two Java source folders:
src/main/java
src/test/java
For my purposes, I need a third one src/junit/java which should be packaged into a JAR with the classifier junit.
If possible, the new source folder should have it's own classpath (compile + everything with scope junit).
My guess is that for this, I will have to modify at least the resource and compile plugins.
Or is there an easier way?
I have a workaround as explained here but for that, I have to put things like Mockito and JUnit on the compile classpath which violates my sense of purity.
For all people who doubt the wisdom of my approach: I have support code that help to write unit tests when you work with code from src/main/java. Since I'm using the same support code in the tests for the project itself, this code needs to be compiled after src/main/java and before src/test/java.
Specifically, my support code needs to import code from src/main/java and the tests need to be able to import the support code.
I've seen a couple of Maven setups, which bundle test code in an own Maven module. You could then create a simple main-module <- support-module <- test-module dependency chain with that. But then main-module would compile fine, if you build it on it's own without test-module. Ofc you could aggreate them together with a reactor-pom and just build the project via this pom.
Edit:
If you have problems with this setup regarding code coverage, you can use the Jacoco Maven plugin to aggregate the test coverage generated by test-module to main-module. See this for further information: http://www.petrikainulainen.net/programming/maven/creating-code-coverage-reports-for-unit-and-integration-tests-with-the-jacoco-maven-plugin/
After a recent juggling with our ant scripts I've started to wonder if something better is possible.
I need a builder that will know to recompile all required .java files for me.
For ex. for this structure
public class A { ]
public class B extends A {}
public class C {
B b;
}
For: Compile('C') Will know to compile A, B, C.
For: B changed, Compile('C') will know to recompile just B.
I know of several alternatives, Ivy which seems like an extension of ant which is our current java builder. Scons which we are currently using for building C++ code, scons is excellent in doing the above described behavior for C code. Then there are reports of Maven being almost but not quite there.
What would you suggest? What tools are you using Free Software / Commercial for you build system?
Thank you,
Maxim.
Ant, with 'depend' task and with 'closure' option turned on
'make', from IDEA ide
None of ivy, scons or maven will help you with your problem as stated.
What do you mean by "for Compile('C')"? I don't think this is what you have in your ant file.
For this case, Ant should be working as desired: you have described its default behaviour. In the same javac element, Ant will only recompile changed classes. See the Ant manual entry for the javac task, especially the 'includeDestClasses' attribute.
You should probably post an example ant file that you are finding inadequate.
maven, both for my personal and my commercial products
In your question you describe inter-class dependencies. Most build systems, in particular Maven, are aimed more at inter-project dependencies. I believe most systems just recompile all the classes in a project and most of the benefits of these build systems is in building as few projects as possible.
Both Maven and Ivy will allow you to easily specify both external and internal dependencies of your project, including which version of the project you depend on. They will both also automatically download external libraries (such as apache commons) to your local machine as part of the build process if they are not already locally cached, saving a lot of work manually downloading and organizing third party jar files.
Ivy is an extension of ant, like you mention. I recommend Maven. It is a convention oriented build system that I've used successfully and feel is quite mature. Maven requires far less up front effort to start using and is quite extensible.
I'm curious about everyones practices when it comes to using or distributing libraries for an application that you write.
First of all, when developing your application do you link the debug or release version of the libraries? (For when you run your application in debug mode)
Then when you run your app in release mode just before deploying, which build of the libraries do you use?
How do you perform the switch between your debug and release version of the libraries? Do you do it manually, do you use macros, or whatever else is it that you do?
I would first determine what requirements are needed from the library:
Debug/Release
Unicode support
And so on..
With that determined you can then create configurations for each combination required by yourself or other library users.
When compiling and linking it is very important that you keep that libraries and executable consistent with respect to configurations used i.e. don't mix release & debug when linking.
I know on the Windows/VS platform this can cause subtle memory issues if debug & release libs are mixed within an executable.
As Brian has mentioned to Visual Studio it's best to use the Configuration Manager to setup how you want each configuration you require to be built.
For example our projects require the following configurations to be available depending on the executable being built.
Debug+Unicode
Debug+ASCII
Release+Unicode
Release+ASCII
The users of this particular project use the Configuration Manager to match their executable requirements with the project's available configurations.
Regarding the use of macros, they are used extensively in implementing compile time decisions for requirements like if the debug or release version of a function is to be linked. If you're using VS you can view the pre-processor definitions attribute to see how the various macros are defined e.g. _DEBUG _RELEASE, this is how the configuration controls whats compiled.
What platform are you using to compile/link your projects?
EDIT: Expanding on your updated comment..
If the Configuration Manager option is not available to you then I recommend using the following properties from the project:
Linker->Additional Library Directories or Linker->Input
Use the macro $(ConfigurationName) to link with the appropriate library configuration e.g. Debug/Release.
$(ProjectDir)\..\third-party-prj\$(ConfigurationName)\third-party.lib
Build Events or Custom Build Step configuration property
Execute a copy of the required library file(s) from the dependent project prior (or after) to the build occurring.
xcopy $(ProjectDir)\..\third-party-prj\$(ConfigurationName)\third-party.dll $(IntDir)
The macro $(ProjectDir) will be substituted for the current project's location and causes the operation to occur relative to the current project.
The macro $(ConfigurationName) will be substituted for the currently selected configuration (default is Debug or Release) which allows the correct items to be copied depending on what configuration is being built currently.
If you use a regular naming convention for your project configurations it will help, as you can use the $(ConfigurationName) macro, otherwise you can simply use a fixed string.
I use VS. The way that I do it is that the libraries I need through the references of the project. Which basically just says in what folder to look for a specific library at project load time. I develop my libraries to be as project independent or reusable as possible. Therefore they are all projects of their own. So of the libraries that I need for a specific project, I create a "3rdParty" or "libs" folder at the same level as my "src" folder in my svn folder tree. I tend to only use released libraries, but when I get some unknown issues and want to switch to debug, I manually copy a debug version of the files in the "lib" folder and reload the project.
I am unsure wether I should be keeping both debug and released versions in my svn tree. Although since they are projects of their own, keeping them in the svn tree of another project doesn't right. They can be built again without an hitch at any moment.
And then I wanted to find a way of making the switch more...hmmm...well basically automatic if you while, but that's not what I really mean. It just feels that switching the files manually between released and debug isn't right. Maybe I haven't found it yet, but what I would like is an option that would do like:
For library "stack.dll" look in "......\3rdParty\" for release and "......\3rdPartyD\" for debug.
Anything that those something like I don't know. What do you suggest?
Remember libraries are external projects. There the built files are totally elsewhere. In fact think of it as you have to check out another project, build it, and copy the built library if you want another copy. How would you set that up?