I have a project working, and then I want to do a copy. Unfortunately, this is not as simple as it should be :-(. In order to copy a project with all the settings, some knowledge about the internals of the project structure is required, which is the topic of this post.

In “Copy my Project” I was using a ColdFire V2 project which is not an easy case, as is using a Target Task to flash the microcontroller. Fortunately, all other targets in CodeWarrior for MCU are *not* using target tasks :-). With little help and preparation, a copy a project is not that difficult to do.  I’m showing how to do this with the FRDM-KL25Z project I have created in this post.

Clone with Copy-Paste

Eclipse projects can be easily copied with a copy-paste operation. That works pretty well, except for the CodeWarrior debugger settings. They are not updated/copied all the way through, at least not up to CodeWarrior for MCU10.3 beta.

There are a few things tied to the project name, and at least in 10.3 beta some are ‘hard coded’. So changing the ‘base’ project a bit will help me to copy it later. So I explain first what the (file) dependencies are, followed by some project adjustments which make the copy easier later one. And finally I’ll show how to do this with an actual project :-).

Project Name Dependencies

There are a few files in the project which are sharing the project name. Namely the output files (Linker ELF file and related output files) and the debugger/launch configuration files:

The *.pemicro file is a text file which keeps the last P&E connection settings. As long as I go with the defaults, I can ignore that file for now.

There is one more file in the Debugger folder: the *.launch file which is filtered by default. To show it inside the project structure, I use the Filters… menu:

That way I can disable the filter for Launch Configurations in the project view:

Now that file shows up (I need to press F5 or do a ‘Refresh’ on the project):

💡 I can export any Eclipse Launch Configuration to a file using the menu File > Export > Run/Debug > Launch Configurations.

While I have now the launch configuration visible, two other things are missing: the RSE (Remote System Explorer) target and connection settings (see “Flashing with a Button (and a Magic Wand)”). To enable the settings file, I need to remove the filter for ‘Referenced Remote System Settings’:

Now the ReferencedRSESystems.xml file shows up (a F5/Refresh might be needed):

➡ In earlier CodeWarrior/Eclipse versions this file is named as ‘rseHostSettingsCache.xml’

❗ Important Note: This .xml file is a ‘cache’ file, and has the debugger target and connection settings used by that project. It is updated automatically in the background by Eclipse. To make sure that the cache is flushed, I close the project to make sure the latest information is stored in it. Modifications in this file should only be done with having the project closed in Eclipse.

Now I have at least all the dependencies with the project name shown in my project structure.

💡 As the .lauch and ReferencedRSESystems.xml files are needed debugger files for my projects, I always store them in my version control system (e.g. SVN). If I don’t have them in my version control system, then I do *not* have any debugging settings for my project.

😥 The fact that there is no way in the Eclipse RSE system to store the connection and target settings in a file inside the project and loaded automatically is a major problem in my view. Having the RSE cache file is better than nothing, but really painful. I hope that in future Eclipse versions this will be improved. So for now it means dealing with this ReferencedRSESystems.xml file.

Building: Using ${ProjectName} for Artifact

The project created used by the standard CodeWarrior projects (at least in MCU10.3beta) has the output (artifact) name defined like this in the project properties:

I can give my output/artifact file any name, but it makes sense to have it named after the project name. But with above, I have to change that name after I have copied it. A better way is to use the variable ${ProjName}:

Debugging: Using Variables to Name the Application

In a similar way, the Application name is used in the debug/launch configuration:

Instead of using the hard-coded naming, I can use ${ConfigName}/${BuildArtifactFileName}:

Copy/Clone of a Project – Step by Step

To copy or clone a project is simple:

a) select the project, then use copy in the context menu (or CTRL+C):

b) and paste (or CTRL+V) it in the workspace. It will ask me to specify the new project name:

This creates a new project for me.

The good thing: the content of the .launch file and the *.pemicro file are OK and updated as needed. The not so good thing: the files names are still the same:

The file name of the launch file (and consequently the one of the *.pemicro) does not really matter. The Eclipse IDE will load any *.launch file inside the project and add it to the launch configurations. But to be consistent, I rename the *.launch and *.pemicro files to match my project name:

What is now an issue is that my copied project still references the connection and target settings of the base/source project. This is obvious when inspect my Debug launch configuration:

This means, my copied project will all use the same (shared) connection settings of the base project:

That might not be a problem at all. If I copy a project, then I might want to use the same shared settings. However, in my opinion I would like to have the copy a full (deep) copy. I want to have the connection settings copied too:

It is possible to copy the Connection and Target settings with the Remote Systems View. That view is accessible with the menu Window > Show View > Remote Systems:

To copy the Connection and Target Settings, I can select both entries and use the ‘Copy…’ context menu:

It will ask me which profile to use (on my machine) which I simply can confirm:

Then it allows me to give the Connection Settings a new name:

And the same for the Target Settings:

Now I have both the Target and the Connection Settings duplicated:

Now I can use the context menu on the Connection and apply it to my project debugger launch configuration:

Unfortunately such a ‘apply’ menu does not exist for the connection. So I need to make the linkage from the Target to the new Connection from within the Connection properties. For this I select the Connection and use the Properties menu and then assign the new Target:

Now I have my copy of the Target and Connection Settings for my new project. An alternative way to accomplish this would be to change the ReferencedRSESystems.xml directly:

1. Close the project so the cache is updated in the file
2. Open the file and edit the XML file. In our example to replace ‘Freedom_LED’ with ‘Copy of Freedom_LED’. Save the file.
3. Re-Open the project in Eclipse again.

💡 Modifying that XML file could be part of a script for automatic copy of a project.

Summary

In summary, here is how copy or clone a project:

First, I make sure my base/source project is set up like this so do not need to update the artifact or application file name for the debugger:

1. Use ${ProjectName} for the Artifact name in the project build settings.
2. Use ${ConfigName}/{BuildArtifactFileName} for the Application name to debug.
3. Disable the project filters for the *.launch and the RSE *.xml cache file.

To copy the project and as well the Connection and Target settings:

1. Select the project, copy (CTRL-C) and paste it (CTRL-V) and give it a new name.
2. Rename the .launch and .pemicro files
3. In the Remote Systems view copy and rename the Connection and Target Settings
4. Apply the Connection setting to the new project
5. Go into the Connection properties and assign the new Target Settings

➡ If I have copied the project with existing make files and object files, it is a good idea to do a Project > Clean, see this post.

Happy Cloning 🙂