custom recipe not working in 1.9.8

,
1

Wonder it's a regression.
Custom recipe doesn't work any more. The same recipe worked with 1.9.6. This is the error I got -

    ✘ Can't find recipe or command 'custom.textcat.teach'.

Here is command I used -

    PRODIGY_PORT=8000 prodigy custom.textcat.teach dataset blank:en 
    data.jsonl –F custom_textcat.py

Here is the custom_textcat.py

    import prodigy
    from prodigy.recipes.textcat import teach
   @prodigy.recipe('custom.textcat.teach',
    dataset=prodigy.recipe_args['dataset'],
    spacy_model=prodigy.recipe_args['spacy_model'],
    source=prodigy.recipe_args['source'],  
    patterns=prodigy.recipe_args['patterns'],
    label=prodigy.recipe_args['label_set'])
def custom_textcat_teach(dataset, spacy_model, source, patterns, label=None):
    components = teach(dataset=dataset, spacy_model=spacy_model,
                       source=source, patterns=patterns, label=label)
    with open('/home/ec2-user/anaconda3/lib/python3.7/site-packages/prodigy/extension/template.html', 'r') as f:
        template = f.read()
    components['config']['html_template'] = template
    components['view_id'] = 'html'
    return components
2

Ha, this was confusing for a second, but I think I found the problem: that dash in –F in your command isn't a regular dash, it's an n-dash. It looks just like -F, but it's not, so Prodigy doesn't read the recipe path or load the file and then exits early because it doesn't have a recipe for that name.

Not sure how we could check for something like this to catch it and show a warning to the user :thinking: (Because in theory, the –F could be part of some other command line input that we don't yet know the meaning of because we don't yet know the recipe.)

3

You're spot-on. I am saving all the commands in a power point file. Power point is altering the dashes.

I'm sure that you have heard from many people. You're the best!