Heating scheduler #37

firstdayofjune · 2021-02-18T12:43:01Z

Added a client to communicate with the IOLITE heating API (plus tests)
Added some helpers to the Discovered
Added a pyproject.toml file to configure isort & black
Minor refactorings to code and test
Readme updated

- added the iolite directory as an editable dev-dependency, so iolite can be imported without being installed (and without a need to manipulate the path) - changed the environment-accesses to use environs instead of os.getenv

- added black and isort as dev-dependencies to use them in file watchers - as all black releases are pre-releases, a specific version number is required to avoid an error when locking (I chose the latest version)

- added black and isort as dev-dependencies to use them in file watchers - chose isort 4.3.21 to match the pre-commit setup - chose black 20.8.b to match the pre-commit setup - also all black releases are pre-releases, a specific version number is required to avoid an error when locking - added a pyproject.toml to configure isort

- moved place and device response handling into separate methods

- added toml as a pre-commit dependency - added toml to pipenv

- added a Heating Scheduler object for interaction with the heating intervals endpoint - added helpers to get rooms and entities to the Discovered class, to use these rooms with the Heating Scheduler - added a jupyter notebook to showcase the Heating Scheduler - added tests and extended client tests, for safer refactoring - updated the readme to introduce the Heating Scheduler

# Conflicts: # Pipfile.lock

- updated lock file after merge

- updated lock file afterwards

- added missing spaces to bulletpoints in README.md - suppressed security warning in test_heating_scheduler.py, as the password is hard-coded deliberately and only used to verify that the client encodes a given password correctly

inverse

Minor comments but looking good 👍🏼

inverse · 2021-02-19T20:36:01Z

Pipfile

@@ -10,6 +10,11 @@ pre-commit = "*"
 coverage = "*"
 pytest-cov = "*"
 codacy-coverage = "*"
+iolite = {editable = true,path = "."}


Why do we need to declare this?

For me, the example.py threw an error because iolite was not an installed package. So I added the iolite directory as an editable dev-dependency. This way you can import iolite without having the package actually installed in the environment.

Ahh now I see it. It feels weird having the package add itself as a dependency but I am not sure of a better way.

Another solution would be to turn scripts into a module (add __init__.py) and then it can be invoked as python -m scripts.example

That would work for the scripts, yes. I just realized the iolite = {editable = true,path = "."} is also needed to import iolite inside the notebooks, though. So I would like to keep this change.

iolite/client.py

iolite/heating_scheduler.py

README.md

notebooks/Heating Scheduler.ipynb

inverse · 2021-02-20T21:04:07Z

@firstdayofjune sorry for the conflict in the Pipfile - shouldn't be hard to fix and regenerate the pipfile.lock

I also fixed the trigger for CI to work against your branch so if you merge/rebase against master that should work now 👍🏼

- renamed discovered into discovered_rooms, as only rooms will be stored - refactored Interval min/max temperatures into constatnts - detailed the instructions on how to install jupyter notebook and how to fix a ModuleNotFoundError that may occur when using jupyter from within the venv

- Changed black version to 19.10b to match pipenv-setup's dependency - Regenerated lock file # Conflicts: # Pipfile # Pipfile.lock

firstdayofjune · 2021-03-22T10:32:25Z

Applied all the changes as discussed. It took me some time to recreate the lockfile until I found out that pipenv-setup requires black at version ==19.10b0, when I had ==20.8b1 in dev packages 🤦‍♀️
Still, finally recreated the lock and pushed my latest changes.
Only issue left seems that CI test 3.9 cannot find the CODACY_PROJECT_TOKEN when uploading the report (and I guess this also stopped test 3.8 from continuing the test execution?).

inverse · 2021-03-22T21:32:30Z

Something must be up with CI - it's happening with other PR's #46

inverse

I updated the pipeline that should only upload coverage on master now.

Going forward I think we should swap out Codacy for coverage and use an alternative platform.

inverse · 2021-03-22T21:35:43Z

Pipfile

 pytest-socket = "*"
 freezegun = "*"

 [packages]
 websockets = "*"
 environs = "*"
 requests = "*"
+
+[pipenv]
+allow_prereleases = true


Why was this declared? Do we really need pre-release dependencies?

Black is only available as a pre-release package and pipenv does not install pre-releases without actively allowing it to do so. When installing it with pipenv install black --dev --pre the "allow_prereleases" part is automatically added by pipenv.
Once black becomes available as a major release, this can be removed again. I guess this can take some time, though: psf/black#1746

What's the need for black? I've set it up as a pre-commit hook which you can install via pre-commit install or just run manually with pre-commit run --all?

Which is also called on CI

Having black, isort, and flake8 installed in the environment, I use file watchers to run them on every file save. I prefer this setup to running the pre-commit hook once in a while, because I immediately see the changes made by black and isort. Also, this way I can fix the flake8 warnings right away.

I could probably also just run the pre-commit hook, this feels slightly slower, though (and runs on all files, not just the file I am working on).

So I added the scripts as dev dependencies, cause I though it does no harm 🙂

Okays - that makes sense.

I think pre-commit runs on the diff by default but you can also tell it to run on all changes.

We can always remove the dev dependencies if it becomes too much.

inverse · 2021-03-23T21:08:41Z

@firstdayofjune fixed CI 👍🏼

inverse · 2021-03-24T07:52:04Z

Thanks 👍🏼

firstdayofjune added 10 commits January 29, 2021 18:29

Made example.py run without errors

42c7905

- added the iolite directory as an editable dev-dependency, so iolite can be imported without being installed (and without a need to manipulate the path) - changed the environment-accesses to use environs instead of os.getenv

Added black & isort

68a395e

- added black and isort as dev-dependencies to use them in file watchers - as all black releases are pre-releases, a specific version number is required to avoid an error when locking (I chose the latest version)

Minor refactoring to client

8d9f8d4

- moved place and device response handling into separate methods

Fix isort in pre-commit

d1825b2

- added toml as a pre-commit dependency - added toml to pipenv

Merge remote-tracking branch 'origin/master' into heating_scheduler

4fb95e2

# Conflicts: # Pipfile.lock

Updated lockfile

9576e6d

- updated lock file after merge

Removed isort duplicate from pipfile

3ae2b81

- updated lock file afterwards

Fixed Code Quality issues

ea8f54c

- added missing spaces to bulletpoints in README.md - suppressed security warning in test_heating_scheduler.py, as the password is hard-coded deliberately and only used to verify that the client encodes a given password correctly

inverse reviewed Feb 19, 2021

View reviewed changes

inverse self-assigned this Feb 27, 2021

firstdayofjune added 3 commits March 18, 2021 18:59

Merge remote-tracking branch 'origin/master' into heating_scheduler

e97ac8e

- Changed black version to 19.10b to match pipenv-setup's dependency - Regenerated lock file # Conflicts: # Pipfile # Pipfile.lock

Removed fixed python version from pipfile, as it is not available on CI.

67ee878

inverse reviewed Mar 23, 2021

View reviewed changes

inverse added 2 commits March 23, 2021 21:24

Merge branch 'master' into heating_scheduler

88995e5

Merge branch 'master' into heating_scheduler

b6abf52

inverse approved these changes Mar 24, 2021

View reviewed changes

inverse merged commit e9bee7b into inverse:master Mar 24, 2021

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Heating scheduler #37

Heating scheduler #37

firstdayofjune commented Feb 18, 2021

inverse left a comment

inverse Feb 19, 2021

firstdayofjune Mar 13, 2021

inverse Mar 23, 2021

inverse Mar 23, 2021

firstdayofjune Mar 24, 2021

inverse commented Feb 20, 2021

firstdayofjune commented Mar 22, 2021

inverse commented Mar 22, 2021

inverse left a comment •

edited

inverse Mar 22, 2021

firstdayofjune Mar 23, 2021

inverse Mar 23, 2021

firstdayofjune Mar 24, 2021

inverse Mar 24, 2021

inverse commented Mar 23, 2021

inverse commented Mar 24, 2021

Heating scheduler #37

Heating scheduler #37

Conversation

firstdayofjune commented Feb 18, 2021

inverse left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

inverse commented Feb 20, 2021

firstdayofjune commented Mar 22, 2021

inverse commented Mar 22, 2021

inverse left a comment • edited

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

inverse commented Mar 23, 2021

inverse commented Mar 24, 2021

inverse left a comment •

edited