Email list hosting service & mailing list manager

Re: New draft (#9) and last call for comments on SRFI 204: Wright-Cartwright-Shinn Pattern Matcher Marc Nieper-Wißkirchen (13 Oct 2020 09:40 UTC)

Re: New draft (#9) and last call for comments on SRFI 204: Wright-Cartwright-Shinn Pattern Matcher Marc Nieper-Wißkirchen 13 Oct 2020 09:40 UTC

(1) The description of the "failure" procedure indicates that it is a
continuation that does not return, which is what most people would
expect. In the same implementation, however, "failure" returns and
delivers the result of the match with the remaining clauses.

(2) The description of tree patterns isn't self-contained. What is a
tree? Can it contain vectors? Even records? Can it be safely applied
to cyclic data structures? My advice would be to make it optional as
some other extensions to the original matcher.

(3) Please add a pattern of the form (var <id>), which is just like
<id> except that <id> in (var <id>) will always be interpreted as a
pattern variable. This allows binding identifiers like $ or _ as
pattern variables, but more importantly, it allows to write robust
macros using the matcher by wrapping user-provided identifiers into
(var ...).

The implementation is almost trivial: Define "var" as auxiliary
syntax, add another clause in this macro
https://github.com/scheme-requests-for-implementation/srfi-204/blob/master/srfi/204/204.scm#L410
and extract https://github.com/scheme-requests-for-implementation/srfi-204/blob/master/srfi/204/204.scm#L486
into its own macro, say match-var so that it can be used by the "var"
clause as well.

Am Di., 13. Okt. 2020 um 08:08 Uhr schrieb Arthur A. Gleckler
<xxxxxx@speechcode.com>:
>
> I've just published draft #9 of SRFI 204. It was submitted by Felix Thibault, author of the SRFI.
>
> Felix has asked me to announce last call for this SRFI. He believes that it is ready for finalization, but would like to give reviewers one last chance to submit corrections and feedback before we finalize it.
>
> In particular, I appeal to anyone reading this to try the sample implementation, run the tests, and send feedback about your results.
>
> If you're interested in this SRFI, please give your feedback via the SRFI 204 mailing list before 2020-10-20. After that, assuming that no major revisions are required, we will declare it final. It is important that we get your feedback before 2020-10-20. If that deadline is too soon for you, but you would like to contribute, please let me know so that I can extend the last-call period.
>
> Here is the commit summary:
>
> update TODO/README.md
> integrate Chez into testing framework
> update TODO/README re Chez testing
> update to latest version (chibi match)
> update TODO/README
> added note to TODO/README about srfi-206
> update srfi to require srfi-206
> retry foment after response from Mike
> added make-pred to forum-topics.scm
> begin turn srfi examples -> tests
> finish making tests from srfi
> srfi tests passing w/guile 2 except empty body pattern
> get to pass with Gauche except empty body
> get to pass with Chez except empty body
> get srfi tests to pass loko except match w/o body
> test all srfi-64 implementations w/empty match screen
> fix mistakes found in srfi
> clean up, fix documentation
> add milestones for handling loose threads
> try make-setter and make-getter
> update TODO/README.md
> meet email milestone 1
> meet milestone 2
> meet email milestone 3
> finish email milestones
> get tests to run on false branch with util match
> fix commented out parts of srfi-test
> add record methods to cyclone
> remove c files from cyclone
> add Gerbil record forms, named record failing
> update TODO, add not implemented error for Gerbil records
> add equality benchmarks
> factor test out of srfi (breaks some includes)
> try Gerbil, can't handle (srfi 204)/(srfi srfi-204) vs (srfi-204)
> remove Gerbil
> get 206 working with chibi, gauche, larceny
> try/fail to make guile work with srfi-206
> put srfi-204 into r6rs form, srfi-206 not working
> try/fail to run chezscheme tests[bad includes]
> fix more Makefile environment/test versioning
> convert loko library to srfi form
> try different approaches to or ellipsis patterns
> bring TODO/README up to date
> document srfi/204.sld exports
> add make-match-xxx documentation, fix typos
> document sections of srfi-common.scm
> fix Makefile to work with Larceny, correct errors in srfi-test.scm
> put Guile module in r6rs form
> chezscheme to r6rs, clean up extra files
> bring cyclone up to srfi/204/204.scm
> take Guile out of 204.sld
> clean up code and get rid of not-working
> restore old srfi-common.scm as r6rs-srfi-common.scm, fix README
> update forum-topics.scm
> update documentation, move auxiliary-syntax to srfi/204
> get rid of extension helpers
> re-run match tests, all passing
> test examples, all passing except extract-imports
> update TODO/README
> get rid of extra logs
> corrections, add to TODOs
> copy edits
> add issues to srfi-204.html
> add extract-imports issue
> resolve conflicts between srfi-common and r6rs srfi-common
> run srfi tests for srfi-64, all passing except extract-imports, Larceny
> add forum to Acknowledgements section
> Fix errors reported by W3C HTML Validator.
> Update abstract.
> Fragment ID is missing and unnecessary.
> make corrections
> edits suggested by Grammarly
> Eliminate trailing whitespace.
> Publish ninth draft.
>
> Here's the diff:
>
> https://github.com/scheme-requests-for-implementation/srfi-204/compare/draft-8..draft-9
>
> Regards,
>
>
> SRFI Editor