How to pass Lazy input to ShellCommandTask?

[adi611]

@djarecka - I am attempting to implement the worker from the PSI/J site, which consists of two parts: 1. Extracting text from a PDF (pdftotext INPUT.pdf OUTPUT.txt) and 2. Creating a word cloud image from the extracted text (wordcloud_cli --text OUTPUT.txt --imagefile IMAGE.png).

My approach involves creating ShellCommandTasks for the above commands and adding them to the workflow to execute them. Here is my working code:

cmd1 = 'pdftotext'
args = ['/content/Lorem_ipsum.pdf', '/content/Output.txt']

shelly1 = pydra.ShellCommandTask(name='shelly1', executable=cmd1, args=args)
print('cmndline = ', shelly1.cmdline)

cmd2 = 'wordcloud_cli'
args2 = ['--text', '/content/Output.txt', '--imagefile', '/content/Image.png']

shelly2 = pydra.ShellCommandTask(name='shelly2', executable=cmd2, args=args2)
print('cmndline = ', shelly2.cmdline)

wf1 = pydra.Workflow(name = 'wf1', input_spec=['x'], x=3)
wf1.add(shelly1)
wf1.add(shelly2)

wf1.set_output([('out_s1', wf1.shelly1.lzout.stdout), ('out_s2', wf1.shelly2.lzout.stdout)])

with pydra.Submitter(plugin='cf') as sub:
    sub(wf1)

wf1.result()

One observation I made is that it is necessary to provide the input_spec and set_output to the workflow. In the given code, I provided a dummy input_spec because I only wanted to execute static shell commands where the filenames were already defined in the args parameter. However, let's say I do want to provide my PDF filename as input to the workflow. In that case, how can I use it as a 'Lazy Input' for my shelly1 task?

[djarecka]

you can point argument of your task to workflow input like this wf.lzin.x, see example here

An example that is closely matching your case you can find in the tests, e.g. here