gh-146436: Implementation of PEP 786#146437
Conversation
|
I managed to achieve the implementation of |
|
Note for myself: "PEP 786: Precision and Modulo-Precision Flag format specifiers for integer fields" is still a draft: python/peps#4416. |
|
Can we avoid creating reference implementations on the main repo please? use a fork please. In addition, issues should be created once it's been discussed and accepted, not before. |
|
I already thought of setting the PEP reference implementation link to a branch on my fork, but then I realized that the upstream python/cpython repo serves as a better permalink; that branch on my fork might get deleted once merged! |
|
In this case, you can create your own repository. We usually avoid creating PRs for features that are still under discussions. |
You can create a pr in your fork and point to that, not to the branch. Then (unless you delete the fork) - you can re-create removed branches. And people still be able to see your patch even if the branch itself will be deleted. See e.g.: skirpichev#1 Though, as this pr was already opened - lets keep this work here (as a draft). |
Disallow precision and two's complements Also add (missing?) tests do disallow 'c' with #, -, and ' '
Forbid '', 'd', and 'n' presentation types from using two's complements as per formatter_unicode::format_long_internal
I'll remember that for the future! 😅 Rebased successfully with no conflicts, pushing now |
I do wish the developers guide (documentation section) would say to `make -C Doc/ check` to avoid the embarassment of pushing broken docs. I tracked down the failing test to a 'sphinx-lint' in `.pre-commit-config.yaml` but how that links up to the 'check' target in Docs/Makefile isn't clear.
|
This PR is stale because it has been open for 30 days with no activity. |
Reference implementation of PEP 786
📚 Documentation preview 📚: https://cpython-previews--146437.org.readthedocs.build/