Passing arguments to tasks: need to specify at least one of short or long?

Hey, thanks for this great tool! I just came across the following issue: For passing arguments to tasks, the documentation specifies that “short” and “long” names are optional. However, in my setup it seems as if omitting both gives an error.

dodo.py:

def task_py_param():
    def show_param(param):
        print(param)
    return {'actions':[(show_param,)],
            'params':[{'name':'param',
                       'default':'default value'}],
            'verbosity':2,
            }

Running the task:

$ doit py_param --param=test
ERROR: Error parsing Task: option --param not recognized (parsing options: ['--param=test'])
$ doit help py_param
py_param  

Versions:

$ python --version
Python 3.4.0
$ pip list
doit (0.28.0)
pip (1.5.4)
pyinotify (0.9.6)
setuptools (2.1)
six (1.9.0)

[mitchbregs]

@schettino72 - I am running into a similar situation when passing parameters through a yield statement instead of return. The documentation only gives examples of passing parameters in one-off tasks with return statement, whereas I am running a task with for loop through a bunch of data and using a yield statement in order to capture the task actions, targets, dependencies, etc. However, I am now trying to add parameters to this task:

doit -f wp/dodo.py external_report --week 1      
ERROR: Error parsing Task: option --week not recognized (parsing options: ['--week', '1'])

This is my yield definition:

            yield {'value': dict(name=output.stem,
                                 actions=[cmd],
                                 file_dep=[(template), (data), (geo)],
                                 targets=[output],
                                 uptodate=[False],
                                 params=[{'name': 'week',
                                          'short': 'wk',
                                          'long': 'week',
                                          'default': '1',
                                          'help': 'Week of reporting.'}]),
                   'log_status': 1}

You can disregard the log_status, I have built a custom logger to log tasks. Read the yield statement as the sub-dictionary of value.

[schettino72]

Thanks a bunch @schettino72 - loving your work on pydoit! smile

Cool, so give it a 🌟 😬