BigDatalex
commented
1 year ago If we use this feature now, how does it affect the local development process? For example if I develop locally in the scraping package and add lets say a new start_URL in the start_scripts from a shop with a new product category e.g. HAT. In order to make this work I would need to touch the core package and add the category HAT to the enum located in domain.py. If I then run the scraping tests, it will throw an error, because poetry is using the core package from the main branch that does not yet include my local changes. This is just one very simple example, there are probably even more dependencies also with the database package etc.
Did I missed something? If not, how can we solve this? Or should we always change in such a case the poetry specification to use the branch in which we are developing currently and then switch it back to main if everything is working?
This is relevant for #136
Finally, poetry allows to use git repo's subpath as dependency: https://python-poetry.org/docs/dependency-specification/#git-dependencies
We should switch to this feature for all packages because this would make it much easier to add, e.g.,
databasepackage as dependency in repositories that are not part of this code.Steps:
core = {git = "https://github.com/calgo-lab/green-db", branch = "main", subdirectory = "core"}