Providing additional init_args in later config overwrites existing namespace instead of updating

[florianblume]

🐛 Bug report

With the new version 4.10.0, providing only init_args in a yaml config for a subclass is possible when class_path is set in some other config. Unfortunately, when I pass a model config using --model and then in a later config using --config additional init_args like so:

model:
    init_args:
        batch_size: 32

then this dictionary overwrites the existing, previously loaded model config completely. This means that there's not class_path entry anymore and the CLI crashes.

I tracked it down to the update function of class Namespace:

def update(self, value: Union['Namespace', Any], key: Optional[str] = None, only_unset: bool = False) -> 'Namespace':
    """Sets or replaces all items from the given nested namespace.

    Args:
        value: A namespace to update multiple values or other type to set in a single key.
        key: Branch key where to set the value. Required if value is not namespace.
        only_unset: Whether to only set the value if not set in namespace.
    """
    if not isinstance(value, Namespace):
        if not key:
            raise KeyError('Key is required if value not a Namespace.')
        if not only_unset or key not in self:
            self[key] = value
    else:
        prefix = key+'.' if key else ''
        for key, val in value.items():
            if not only_unset or prefix+key not in self:
                self[prefix+key] = val
    return self

value is not a Namespace in case of the update dict coming from the second config which then goes into the first part of the if statment: if not isinstance(value, Namespace):. This is the part where it simply overwrites the existing and properly loaded namespace of the model.

To reproduce

from jsonargparse import ActionConfigFile, ArgumentParser
import yaml

class ModelBaseClass():
    def __init__(self, batch_size: int=0):
        self.batch_size = batch_size

class ModelClass(ModelBaseClass):
    def __init__(self, image_size: int=0, **kwargs):
        super().__init__(**kwargs)
        self.image_size = image_size

parser = ArgumentParser(
    prog='app',
    description='Description for my app.'
)

parser.add_subclass_arguments(ModelBaseClass, 'model')
parser.add_argument('--config', action=ActionConfigFile) 

model = yaml.safe_dump({
    'class_path': 'ModelClass',
    'init_args': {
        'image_size': 10
    }
})

config = yaml.safe_dump({
    'model': {
        'init_args': {
            'batch_size': 5
        }
    }
})

parser.parse_args([f'--model={model}', f'--config={config}'])

Output:
app: error: Configuration check failed :: Parser key "model": Type <class 'model.ModelBaseClass'> expects: a class path (str); or a dict with a class_path entry; or a dict with init_args (if class path given previously). Got "{'init_args': {'batch_size': 5}}".

[mauvilsa]

Fixed in commit c4ea888.