Hi folks, I have updated PEP 642 significantly based on the feedback received over the past week.
Since the Discourse thread hadn't received any comments on it, I amended the existing thread in place rather than making a new one: https://discuss.python.org/t/pep-642-constraint-pattern-syntax-for-structural-pattern-matching/5614 The full text of the PEP is quoted in the Discourse thread. The rendered version of the updated PEP can be found in the usual place: https://www.python.org/dev/peps/pep-0642/ Relative to the initially published version, the major changes are based primarily on the counterproposal that Steven D'Aprano posted: * Explicit equality constraints become "== EXPR" (replacing v1's "?EXPR") * Explicit identity constraints become "is EXPR" (replacing v1's "?is EXPR") * Inferred constraints are introduced to deal with the resulting verbosity problems. These mostly follow the same syntax as PEP 634 proposes for value patterns and literal patterns, but are defined in terms of the explicit constraints, rather than being the only way to request the desired semantics. While I considered ceding the point on using `_` as the wildcard marker (and had updated the PEP to say that), I couldn't bring myself to actually do it, especially after reading Thomas Wouters review of the current state of the pattern matching proposals. So instead this version proposes the use of `__` in that role, with an explanation of how that opens a path towards restoring language self-consistency in a future PEP by making `__` special everywhere, rather than only having it be special in match patterns. There are a few other changes that get down into the weeds of actually implementing these features in the compiler (like introducing a SkippedBinding node into the AST so most of the compiler doesn't need to care how the wildcard is spelled in the surface syntax), but I doubt those are going to matter to most folks in deciding which of the proposals on offer they prefer (or if they'd prefer continue without any pattern matching syntax at all). Cheers, Nick. -- Nick Coghlan | ncogh...@gmail.com | Brisbane, Australia _______________________________________________ Python-Dev mailing list -- python-dev@python.org To unsubscribe send an email to python-dev-le...@python.org https://mail.python.org/mailman3/lists/python-dev.python.org/ Message archived at https://mail.python.org/archives/list/python-dev@python.org/message/IKAMHN3FVNHZEAEIMDRWFIAD4HXFQ24X/ Code of Conduct: http://python.org/psf/codeofconduct/
