Honour module's `config.yml` when dynamically importing modules into Snakemake workflow - snakemake

Question

Background
I'm new to scientific workflows, and am building my first more complex Snakemake workflow.
Setup & what I've tried
I'm importing modules dynamically. The scenario is that I have a root config.yml file that includes:
subworkflows:
- subwf-1
- subwf-2
- other-subwf
In the root Snakefile I'm doing:
configfile: 'config.yml'
ALL_SUBWFS = config['subworkflows']
# <omitted> Check for min version 6.0
for MODULE in ALL_SUBWFS:
module MODULE:
snakefile:
f'subworkflows/{MODULE}/Snakefile'
use rule * from MODULE as f'{MODULE}_*'
This works fine so far. However, I'd like to be able to configure the different submodules each in their own config.yml.
Assume the following directory structure:
.
├── Snakefile
├── config.yml
└── subworkflows/
├── subwf-1/
│   ├── Snakefile
│   └── config.yml
├── subwf-2/
│ ├── Snakefile
│ └── config.yml
└── other-subwf/
├── Snakefile
└── config.yml
As far as I understand, this isn't supported, and neither of these options work:
Define configfile: config.yml in main workflow and configfile: cf in subworkflows, where I've tried three options for cf:
cf = str(workflow.source_path('config.yml')) # 1
cf = f'{workflow.basedir}/config.yml' # 2
cf = 'config.yml' # 3
# With each of these options
configfile: cf
All give me KeyError in <...>/Snakefile: 'config'.
Using config: config in module import, and subworkflows' Snakefiles including something like VALUES = config['values'] gives me KeyError in <...>/Snakefile: 'values' using each option.
Actual question
Am I right in assuming that it isn't possible to honour the configfiles for modules at all, and that instead, I need to use a global config file, e.g. with keys for each subworkflow imported as config: config['<key-for-subworkflow-config-YAML-map>']?

Answer 1

Further experimentation seems to show that my dynamic import of modules doesn't work: While print()ing from the subworkflow Snakefiles works fine, onyl one of the subworkflows gets executed when rules are defined.
This renders the question unanswerable at this point.

Answer 2

Regarding your Actual question:
Rather than trying to pass or overwrite configfile you can try the intended approach for module of passing/overwriting the config dictionary used by the module.
The keyword here is config: <dict> in the module <name>: block which you might be able to combine with the load_config method of `snakemake:
from snakemake.io import load_configfile
for MODULE in ALL_SUBWFS:
module MODULE:
snakefile: f'subworkflows/{MODULE}/Snakefile'
config: load_configfile(f'subworkflows/{MODULE}/config.yml'
Note: Untested as you report other issues with your idea for loading external workflows dynamically.