re-run diagnostic script stuck

[tomaslovato]

After the successful execution of a recipe, re-running the diagnostic script stop with the following message:

Script will abort to prevent accidentally overwriting your data in these directories
Use -f or --force to force emptying the output directories or use -i or --ignore-existing to ignore existing output directories.
Script will abort to prevent accidentally overwriting your data in these directories

This can be easily solved by adding the suggested options -f or -i, but I fell like that the re-run message produced by code at the end of diagnostic function should more clear, e.g. by including at least -i.

In particular, -i can be added in rerun_msg variable within the following piece of code:

ESMValCore/esmvalcore/_task.py

Lines 409 to 414 in 82178a9

	rerun_msg = 'cd {}; '.format(cwd)
	if env:
	rerun_msg += ' '.join('{}="{}"'.format(k, env[k]) for k in env
	if k not in os.environ)
	rerun_msg += ' ' + ' '.join(cmd)
	logger.info("To re-run this diagnostic script, run:\n%s", rerun_msg)

It could be also an idea to add in the log informations the need to specify -f or -i options.

[bouweandela]

Note that those options are specific for Python diagnostics, they are not implemented in NCL, R, or Julia.

[tomaslovato]

@bouweandela I recognize it is rather trivial, although not immediate, so I'll close this issue if you agree...

[bouweandela]

It's up to you, I think it's a good suggestion for improving the user friendliness of the tool. Unfortunately we have limited time available to implement all of those suggestions...

[tomaslovato]

@bouweandela I report here below a tiny change to the above referred code to incldue the -i options in the log info when using python diags ...

@@ -411,6 +411,8 @@ class DiagnosticTask(BaseTask):
             rerun_msg += ' '.join('{}="{}"'.format(k, env[k]) for k in env
                                   if k not in os.environ)
         rerun_msg += ' ' + ' '.join(cmd)
+        if cmd[1].endswith('.py'):
+            rerun_msg = rerun_msg.replace('.py', '.py -i', 1)
         logger.info("To re-run this diagnostic script, run:\n%s", rerun_msg)

         try:

If it looks ok, should I open a PR or you want to add it in the code ....

[tomaslovato]

This was solved via #326