-
Notifications
You must be signed in to change notification settings - Fork 10
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
feat: Inflate_all() #204
feat: Inflate_all() #204
Conversation
tags: feat, wip Why? - The inflate_all() function is tested What? - We need to finalize the documentation and split the current flat file into two flat files Issues issue #96
tags: doc Why? - What? - Issues issue #96
tags: doc Why? - What? - Issues issue #96
tags: fix Why? - What? - Issues issue #96
tags: fix, test Why? - df_to_config() shouldnt be allowed to receive inflate parameters when working with "keep" What? - Edited the function, doc, and unit tests Issues issue #96
tags: feat, test Why? warnings comes with the call of the function it was called, here it was not really informative What? - Add call. = FALSE to clean the warning output
tags: feat, doc, test Why? - Stop also outputs the call by default What? - Clean stop call output to avoid do.call(xxx) message
tags: fix Stop the process if Cancel is checked issue #157
tags: feat, test
tags: fix, test Why? - Showing messages with {cli} for clarity What? - Update unit test - Add function to be tested for outputs only - Refactor unit test structure with multiple test_that calls for easier maintenance
tags: feat Why? - Need to document and check once only What? - Add check an document as in inflate
tags: refactor, test Why? - Unit tests output is long and multiple messages are useless for debugging - New version of R does not like to readlines with missing final line What? - Added some suppressMessages() - Added extra line in `cat()` created files
tags: fix, test Why? - What? - qpdf had to be installed on workbench - replaced some expect_equal by expect_true(all(...)) Issues issue #96
https://tech.lgbt/@debruine/110315357001043525
|
tags: ci Why? - Trying to resolve CI errors What? - Issues issue #96
tags: ci Why? - What? - Issues issue #96
tags: ci Why? - What? - Issues issue #96
tags: ci Why? - What? - Issues issue #96
tags: fix, feat, test, ci Why? - We had issues using devtools::check because of a missing MASS library What? - check() automatically builds a package before calling check_built(), as this is the recommended way to check packages. Note that this process runs in an independent R session, so nothing in your current workspace will affect the process. Under-the-hood, check() and check_built() rely on pkgbuild::build() and rcmdcheck::rcmdcheck(). Issues issue #96
tags: test Why? - What? - Issues issue #96
tags: fix Why? - What? - Issues issue #96
tags: test Why? - What? - Issues issue #96
tags: feat Why? - Logicals were written as yes/no, we want them to be stored as TRUE/FALSE What? - https://tech.lgbt/@debruine/110315357001043525 Issues issue #96
tags: feat Why? - The same operation was done in inflate() and inflate_all(), we needed a function What? - see document_and_check_pkg() Issues issue #96
Merci de continuer à avancer sur ce sujet ! Je vois que tu as retiré {devtools} pour mettre {rcmdcheck}. Je pense qu'il faut remettre {devtools} plutôt que {rmcdcheck} parce que les utilisateurs vont utiliser {fusen} dans un environnement interactif et non pas dans un contexte comme nos tests unitaires. Je te propose de regarder ce qu'il se passe dans les tests "inflate-part-1". Tu verras que j'ai mis des On pourrait discuter du vrai pourquoi. Pourquoi tu t'es senti obligé de déclarer MASS, et tous les packages r-base ? |
tags: fix, doc, ci Why? - What? - Issues issue #96
Merci pour ta réponse c'était très instructif ! J'ai fait des modifications en conséquent. J'ai par contre des soucis avec la github action R-CMD-check-devel.yaml Cf log https://github.com/ThinkR-open/fusen/actions/runs/5090267208/jobs/9148872503?pr=204 -> L'installation de |
Il semblerait que certaines dépendances systèmes nécessaire pour curl soient manquantes : On note un problème dans la récupération des dépendances systèmes dans le fichier github actions En théorie Ubuntu 20.04 est censé être encore supporté : Peut etre ouvrir une issue dédiée ? |
tags: chore, ci - Use the latest version of actions to apply to 'devel' action (the one that use the github version of parsermd) - Use ubuntu-devel
tags: fix, test path_foosen does not exist till it is not created. normalizePath need mustWork=FALSE to avoid sending a warning.
tags: fix, test - Add extra line in Rproj file.
tags: fix, test - It was anticipated a problem with new version of attachment, but it is finally not the case. We can verify it as it is now on CRAN
tags: refactor, test, doc - As clean phase is not ready, we keep it separated for later developments
tags: fix, test - `force` parameter should allow to add files in the config file even if they do not exist
tags: doc - Manage NEWS with unpublished 0.5.0 - Re-write inflate_all doc
tags: fix, test - Modification of files needs to be done using name instead of position as 'sort' changes in non-interactive environments - Fix 'sort' by 'order' in unit tests to properly select a row issue #96
tags: chore - Need to inflate a few flats before being able to use inflate_all() - Or define state
tags: fix, test - A dev version of fusen has no "state" in "keep" section issue #96
tags: feat, doc - Opening a vignette is an individual choice issue #96
tags: fix - check() outputs need to be printed in the console
tags: feat There is no general use to see the path of the current package being inflated. It only matters for unit tests
TODO:
->
devtools::check()
performed by @ymansiaux on 2023-05-11All like:
Test
inflate_all_no_check()
to verify it does not check ?done by @ymansiaux on 2023-05-10 : a39a1a5
Test what happens with "keep" section: test driven
done by @ymansiaux on 2023-05-10 : 8e4901d
Edit inflate-params writing in config_fusen.yaml to deal appropriately with logicals.
Logicals were written as yes/no, we want them to be stored as TRUE/FALSE
done by @ymansiaux on 2023-05-11 : a253b4e
See if we can put in function the document & check part used similarly in
ìnflate()
andinflate_all()
done by @ymansiaux on 2023-05-11 ba45a5a
See if there is still a need to separate
test_that()
calls for easier maintenanceOpen issue for documentation refactor
Faire tourner
inflate_all()
etinflate_all_no_check()
sur {fusen} lui-même pour voir. Gérer les messages d'erreurs plus explicites pour l'utilisateurissue #96
close #157