March 2020 - Esug-list - lists.esug.org

Re: [Esug-list] destructDo:

by Bernhard Pieber

Hi Christian, Thank you for the example. I like it (except for the need of ignore maybe). So #bind: from Pharo 8 would not work in that case, right? I see three possible semantics: 1. block arguments size must match receiver collection's size: like Pharo 8's #bind or PetitParsers's #map:: 2. arguments size must be equal or less than receiver collection's size: like #destructDo: 3. Like 2 but with using rest of collection (like in Nobert's JavaScript example): could be #destructDo:rest: Bernhard > Am 28.03.2020 um 19:20 schrieb Christian Haider <christian.haider(a)smalltalked-visuals.com>: > > > Hi Bernhard, > > yes, it is very handy for csv tables. I deal a lot with them lately (budgets, CoViD19). > I read a line, separate the fields and use that list for #destructDo: . > Now I can just pick the first n columns I am interested in. > (Mostly, identifying information is in the first columns.) > > Real world example: > ^stringsOfLine destructDo: [:ignore :string2 :string3 :string4 | > self produktbereich: string2 profitcenter: string3 bezeichnung: string4] > > I think that the code becomes much more readable. > Except the name ... :-) > > Stay healthy, > Christian > > -----Ursprüngliche Nachricht----- > Von: Bernhard Pieber <bernhard(a)pieber.com> > Gesendet: Samstag, 28. März 2020 18:53 > An: Martin McClure <martin.mcclure(a)gemtalksystems.com> > Cc: Christian Haider <christian.haider(a)smalltalked-visuals.com>; Christophe Dony <dony(a)lirmm.fr>; esug-list(a)lists.esug.org > Betreff: Re: [Esug-list] destructDo: > > Hi everyone, > > I find Martin's arguments regarding "destruct" and "do" rather convincing. Which brings me to: > > #(1 2 3) destructured: [:a :b :c | a + b + c] > > Christian, have you ever used this method in a case in which the number of block arguments was smaller than the receiver collection's size? > > Bernhard > >> Am 28.03.2020 um 16:13 schrieb Martin McClure <martin.mcclure(a)gemtalksystems.com>: >> >> Hi Christian and all, >> >> #destructDo: is probably not the best name. "Destruct" can be interpreted as either short for "destructured" (which is what you want) or more like "destroy" as in "self-destruct." I initially took "destruct" in the "destroy" sense and guessed that #destructDo: might remove each element from the collection as it iterated. You could lengthen it to #destructured: or #destructuredDo:, but if there's a good short name that would be better. >> >> #tupleDo: feels better. >> >> But both #tupleDo: and #destructDo: contain "do" which implies >> iteration. Unless I'm missing something, your #destructDo: does not >> iterate. I'd really expect a #tupleDo: to iterate -- I'd expect with >> >> #(1 2 3 4) tupleDo: [:a :b: | a + b] >> >> that the block would be evaluated twice, once with 1 and 2 and once with 3 and 4. This would be a nice pattern; you could also have a tupleCollect: a tupleDetect: and so on. >> >> #mapTo: is quite similar to #map: in PetitParser, except IIRC PetitParser's #map: expects an exact match in argument count. But maybe #map: is a good name for consistency. >> >> Hope this helps. >> >> Regards, >> -Martin >> >> On 3/28/20 7:24 AM, Christian Haider wrote: >>> Yes, that would be nice. Unfortunately a bit more involved than this >>> 1-liner… >>> >>> Von: Esug-list <esug-list-bounces(a)lists.esug.org> Im Auftrag von >>> Christophe Dony >>> Gesendet: Samstag, 28. März 2020 15:17 >>> An: esug-list(a)lists.esug.org >>> Betreff: Re: [Esug-list] destructDo: >>> >>> Hi. >>> I like tupleDo: >>> >>> or matchDo: >>> >>> and what about installing a full pattern matching, as in many >>> languages e.g. Scheme : >>> https://docs.racket-lang.org/reference/match.html >>> >>> #(1 #(2 3 6 7) 4) matchDo: [:a (:b :@c) :d] | …] here c would be #(3 >>> 6 7) >>> >>> may be it already exist? >>> if not, would really be a + >>> >>> Christophe >>> >>> >>> >>> Le 28/03/2020 à 14:15, Mark Plas a écrit : >>> Hello Christian, >>> >>> " I don’t really like the name," >>> >>> Maybe you could call it #tupleDo:? >>> >>> #(1 2 3) tupleDo: [:a :b :c | …] >>> >>> Mark >>> >>> >>> >>> Mark Plas >>> >>> Software Engineer >>> T +32 2 467 34 30 >>> mark.plas(a)mediagenix.tv >>> >>> >>> >>> <image001.png> >>> >>> >>> Nieuwe Gentsesteenweg 21/1 >>> >>> 1702 Groot-Bijgaarden - Belgium >>> >>> >>> <image002.png> <image003.png><image004.png>www.mediagenix.tv >>> >>> >>> Directions to MEDIAGENIX >>> >>> >>> >>> This e-mail and any files attached to it are confidential and intended only for the use of the individuals or entity to whom they are addressed. If you have received this e-mail in error, please notify the sender immediately. >>> >>> From: Esug-list <esug-list-bounces(a)lists.esug.org> On Behalf Of >>> Christian Haider >>> Sent: zaterdag 28 maart 2020 13:43 >>> To: vwnc(a)cs.uiuc.edu; esug-list(a)lists.esug.org; >>> amber-lang(a)googlegroups.com >>> Subject: [Esug-list] destructDo: >>> >>> Hi, >>> >>> I am using a nice little method #destructDo: for a while now and it feels good. >>> In VisualWorks: >>> SequenceableCollection>>destructDo: aBlock >>> "Evaluate aBlock with the receiver's elements as parameters. >>> aBlock takes its arguments from the receiver. >>> 'ok' >>> #(1 2 3) destructDo: [:a :b :c | a + b + c] >>> #(1 2 3) destructDo: [:a :b | a + b] >>> #(1 2 3) destructDo: [:a | a] >>> #(1 2 3) destructDo: [42] >>> 'not ok' >>> #(1 2 3) destructDo: [:a :b :c :d | a + b + c + d] >>> " >>> ^aBlock cullWithArguments: self >>> asArray >>> >>> In Amber: >>> SequenceableCollection>>destructDo: aBlock >>> ^aBlock valueWithPossibleArguments: >>> self >>> >>> In Pharo and other dialects, I don’t know, but should be as easy. >>> >>> For example you can do >>> (('2020-03-28' tokensBasedOn: $-) collect: #asNumber) destructDo: [:year :month :day | >>> Date newDay: day monthNumber: month >>> year: year] >>> >>> I like that the block is not the receiver (like with #valueWithArguments or #cullWithArguments), but the last argument. >>> >>> Now the questions: >>> • I am sure that others came up with this. Anybody knows? >>> • What are you using for this pattern? >>> • I don’t really like the name, but haven’t found anything better yet. Maybe #destructedDo: or just #destructed: or: #withPartsDo:… maybe something shorter? Ideas? >>> • What are you thinking about this? >>> >>> Happy hacking, >>> Christian >>> >>> >>> >>> >>> _______________________________________________ >>> Esug-list mailing list >>> Esug-list(a)lists.esug.org >>> http://lists.esug.org/mailman/listinfo/esug-list_lists.esug.org >>> >>> >>> >>> >>> _______________________________________________ >>> Esug-list mailing list >>> >>> Esug-list(a)lists.esug.org >>> http://lists.esug.org/mailman/listinfo/esug-list_lists.esug.org > > >

March 28, 2020

Re: [Esug-list] destructDo:

by Pape, Tobias

> On 28.03.2020, at 19:20, Christian Haider <christian.haider(a)smalltalked-visuals.com> wrote: > > Hi Bernhard, > > yes, it is very handy for csv tables. I deal a lot with them lately (budgets, CoViD19). > I read a line, separate the fields and use that list for #destructDo: . > Now I can just pick the first n columns I am interested in. > (Mostly, identifying information is in the first columns.) > > Real world example: > ^stringsOfLine destructDo: [:ignore :string2 :string3 :string4 | > self produktbereich: string2 profitcenter: string3 bezeichnung: string4] > > I think that the code becomes much more readable. What about ^stringsOfLine asComponentsIn: [:ignore :string2 :string3 :string4 | self produktbereich: string2 profitcenter: string3 bezeichnung: string4] ? Best regards -Tobias > Except the name ... :-) > > Stay healthy, > Christian > > -----Ursprüngliche Nachricht----- > Von: Bernhard Pieber <bernhard(a)pieber.com> > Gesendet: Samstag, 28. März 2020 18:53 > An: Martin McClure <martin.mcclure(a)gemtalksystems.com> > Cc: Christian Haider <christian.haider(a)smalltalked-visuals.com>; Christophe Dony <dony(a)lirmm.fr>; esug-list(a)lists.esug.org > Betreff: Re: [Esug-list] destructDo: > > Hi everyone, > > I find Martin's arguments regarding "destruct" and "do" rather convincing. Which brings me to: > > #(1 2 3) destructured: [:a :b :c | a + b + c] > > Christian, have you ever used this method in a case in which the number of block arguments was smaller than the receiver collection's size? > > Bernhard > >> Am 28.03.2020 um 16:13 schrieb Martin McClure <martin.mcclure(a)gemtalksystems.com>: >> >> Hi Christian and all, >> >> #destructDo: is probably not the best name. "Destruct" can be interpreted as either short for "destructured" (which is what you want) or more like "destroy" as in "self-destruct." I initially took "destruct" in the "destroy" sense and guessed that #destructDo: might remove each element from the collection as it iterated. You could lengthen it to #destructured: or #destructuredDo:, but if there's a good short name that would be better. >> >> #tupleDo: feels better. >> >> But both #tupleDo: and #destructDo: contain "do" which implies >> iteration. Unless I'm missing something, your #destructDo: does not >> iterate. I'd really expect a #tupleDo: to iterate -- I'd expect with >> >> #(1 2 3 4) tupleDo: [:a :b: | a + b] >> >> that the block would be evaluated twice, once with 1 and 2 and once with 3 and 4. This would be a nice pattern; you could also have a tupleCollect: a tupleDetect: and so on. >> >> #mapTo: is quite similar to #map: in PetitParser, except IIRC PetitParser's #map: expects an exact match in argument count. But maybe #map: is a good name for consistency. >> >> Hope this helps. >> >> Regards, >> -Martin >> >> On 3/28/20 7:24 AM, Christian Haider wrote: >>> Yes, that would be nice. Unfortunately a bit more involved than this >>> 1-liner… >>> >>> Von: Esug-list <esug-list-bounces(a)lists.esug.org> Im Auftrag von >>> Christophe Dony >>> Gesendet: Samstag, 28. März 2020 15:17 >>> An: esug-list(a)lists.esug.org >>> Betreff: Re: [Esug-list] destructDo: >>> >>> Hi. >>> I like tupleDo: >>> >>> or matchDo: >>> >>> and what about installing a full pattern matching, as in many >>> languages e.g. Scheme : >>> https://docs.racket-lang.org/reference/match.html >>> >>> #(1 #(2 3 6 7) 4) matchDo: [:a (:b :@c) :d] | …] here c would be #(3 >>> 6 7) >>> >>> may be it already exist? >>> if not, would really be a + >>> >>> Christophe >>> >>> >>> >>> Le 28/03/2020 à 14:15, Mark Plas a écrit : >>> Hello Christian, >>> >>> " I don’t really like the name," >>> >>> Maybe you could call it #tupleDo:? >>> >>> #(1 2 3) tupleDo: [:a :b :c | …] >>> >>> Mark >>> >>> >>> >>> Mark Plas >>> >>> Software Engineer >>> T +32 2 467 34 30 >>> mark.plas(a)mediagenix.tv >>> >>> >>> >>> <image001.png> >>> >>> >>> Nieuwe Gentsesteenweg 21/1 >>> >>> 1702 Groot-Bijgaarden - Belgium >>> >>> >>> <image002.png> <image003.png><image004.png>www.mediagenix.tv >>> >>> >>> Directions to MEDIAGENIX >>> >>> >>> >>> This e-mail and any files attached to it are confidential and intended only for the use of the individuals or entity to whom they are addressed. If you have received this e-mail in error, please notify the sender immediately. >>> >>> From: Esug-list <esug-list-bounces(a)lists.esug.org> On Behalf Of >>> Christian Haider >>> Sent: zaterdag 28 maart 2020 13:43 >>> To: vwnc(a)cs.uiuc.edu; esug-list(a)lists.esug.org; >>> amber-lang(a)googlegroups.com >>> Subject: [Esug-list] destructDo: >>> >>> Hi, >>> >>> I am using a nice little method #destructDo: for a while now and it feels good. >>> In VisualWorks: >>> SequenceableCollection>>destructDo: aBlock >>> "Evaluate aBlock with the receiver's elements as parameters. >>> aBlock takes its arguments from the receiver. >>> 'ok' >>> #(1 2 3) destructDo: [:a :b :c | a + b + c] >>> #(1 2 3) destructDo: [:a :b | a + b] >>> #(1 2 3) destructDo: [:a | a] >>> #(1 2 3) destructDo: [42] >>> 'not ok' >>> #(1 2 3) destructDo: [:a :b :c :d | a + b + c + d] >>> " >>> ^aBlock cullWithArguments: self >>> asArray >>> >>> In Amber: >>> SequenceableCollection>>destructDo: aBlock >>> ^aBlock valueWithPossibleArguments: >>> self >>> >>> In Pharo and other dialects, I don’t know, but should be as easy. >>> >>> For example you can do >>> (('2020-03-28' tokensBasedOn: $-) collect: #asNumber) destructDo: [:year :month :day | >>> Date newDay: day monthNumber: month >>> year: year] >>> >>> I like that the block is not the receiver (like with #valueWithArguments or #cullWithArguments), but the last argument. >>> >>> Now the questions: >>> • I am sure that others came up with this. Anybody knows? >>> • What are you using for this pattern? >>> • I don’t really like the name, but haven’t found anything better yet. Maybe #destructedDo: or just #destructed: or: #withPartsDo:… maybe something shorter? Ideas? >>> • What are you thinking about this? >>> >>> Happy hacking, >>> Christian >>> >>> >>> >>> >>> _______________________________________________ >>> Esug-list mailing list >>> Esug-list(a)lists.esug.org >>> http://lists.esug.org/mailman/listinfo/esug-list_lists.esug.org >>> >>> >>> >>> >>> _______________________________________________ >>> Esug-list mailing list >>> >>> Esug-list(a)lists.esug.org >>> http://lists.esug.org/mailman/listinfo/esug-list_lists.esug.org > > > > > _______________________________________________ > Esug-list mailing list > Esug-list(a)lists.esug.org > http://lists.esug.org/mailman/listinfo/esug-list_lists.esug.org -- Tobias Pape Software Architecture Group | http://www.hpi.uni-potsdam.de/swa/ Future SOC Lab | https://hpi.de/future-soc-lab Hasso-Plattner-Institut für Digital Engineering gGmbH | Universität Potsdam Prof.-Dr.-Helmert-Str. 2-3, D-14482 Potsdam, Germany Amtsgericht Potsdam, HRB 12184 | Geschäftsführung: Prof. Dr. Christoph Meinel

March 28, 2020

Re: [Esug-list] destructDo:

by Christian Haider

Hi Bernhard, yes, it is very handy for csv tables. I deal a lot with them lately (budgets, CoViD19). I read a line, separate the fields and use that list for #destructDo: . Now I can just pick the first n columns I am interested in. (Mostly, identifying information is in the first columns.) Real world example: ^stringsOfLine destructDo: [:ignore :string2 :string3 :string4 | self produktbereich: string2 profitcenter: string3 bezeichnung: string4] I think that the code becomes much more readable. Except the name ... :-) Stay healthy, Christian -----Ursprüngliche Nachricht----- Von: Bernhard Pieber <bernhard(a)pieber.com> Gesendet: Samstag, 28. März 2020 18:53 An: Martin McClure <martin.mcclure(a)gemtalksystems.com> Cc: Christian Haider <christian.haider(a)smalltalked-visuals.com>; Christophe Dony <dony(a)lirmm.fr>; esug-list(a)lists.esug.org Betreff: Re: [Esug-list] destructDo: Hi everyone, I find Martin's arguments regarding "destruct" and "do" rather convincing. Which brings me to: #(1 2 3) destructured: [:a :b :c | a + b + c] Christian, have you ever used this method in a case in which the number of block arguments was smaller than the receiver collection's size? Bernhard > Am 28.03.2020 um 16:13 schrieb Martin McClure <martin.mcclure(a)gemtalksystems.com>: > > Hi Christian and all, > > #destructDo: is probably not the best name. "Destruct" can be interpreted as either short for "destructured" (which is what you want) or more like "destroy" as in "self-destruct." I initially took "destruct" in the "destroy" sense and guessed that #destructDo: might remove each element from the collection as it iterated. You could lengthen it to #destructured: or #destructuredDo:, but if there's a good short name that would be better. > > #tupleDo: feels better. > > But both #tupleDo: and #destructDo: contain "do" which implies > iteration. Unless I'm missing something, your #destructDo: does not > iterate. I'd really expect a #tupleDo: to iterate -- I'd expect with > > #(1 2 3 4) tupleDo: [:a :b: | a + b] > > that the block would be evaluated twice, once with 1 and 2 and once with 3 and 4. This would be a nice pattern; you could also have a tupleCollect: a tupleDetect: and so on. > > #mapTo: is quite similar to #map: in PetitParser, except IIRC PetitParser's #map: expects an exact match in argument count. But maybe #map: is a good name for consistency. > > Hope this helps. > > Regards, > -Martin > > On 3/28/20 7:24 AM, Christian Haider wrote: >> Yes, that would be nice. Unfortunately a bit more involved than this >> 1-liner… >> >> Von: Esug-list <esug-list-bounces(a)lists.esug.org> Im Auftrag von >> Christophe Dony >> Gesendet: Samstag, 28. März 2020 15:17 >> An: esug-list(a)lists.esug.org >> Betreff: Re: [Esug-list] destructDo: >> >> Hi. >> I like tupleDo: >> >> or matchDo: >> >> and what about installing a full pattern matching, as in many >> languages e.g. Scheme : >> https://docs.racket-lang.org/reference/match.html >> >> #(1 #(2 3 6 7) 4) matchDo: [:a (:b :@c) :d] | …] here c would be #(3 >> 6 7) >> >> may be it already exist? >> if not, would really be a + >> >> Christophe >> >> >> >> Le 28/03/2020 à 14:15, Mark Plas a écrit : >> Hello Christian, >> >> " I don’t really like the name," >> >> Maybe you could call it #tupleDo:? >> >> #(1 2 3) tupleDo: [:a :b :c | …] >> >> Mark >> >> >> >> Mark Plas >> >> Software Engineer >> T +32 2 467 34 30 >> mark.plas(a)mediagenix.tv >> >> >> >> <image001.png> >> >> >> Nieuwe Gentsesteenweg 21/1 >> >> 1702 Groot-Bijgaarden - Belgium >> >> >> <image002.png> <image003.png><image004.png>www.mediagenix.tv >> >> >> Directions to MEDIAGENIX >> >> >> >> This e-mail and any files attached to it are confidential and intended only for the use of the individuals or entity to whom they are addressed. If you have received this e-mail in error, please notify the sender immediately. >> >> From: Esug-list <esug-list-bounces(a)lists.esug.org> On Behalf Of >> Christian Haider >> Sent: zaterdag 28 maart 2020 13:43 >> To: vwnc(a)cs.uiuc.edu; esug-list(a)lists.esug.org; >> amber-lang(a)googlegroups.com >> Subject: [Esug-list] destructDo: >> >> Hi, >> >> I am using a nice little method #destructDo: for a while now and it feels good. >> In VisualWorks: >> SequenceableCollection>>destructDo: aBlock >> "Evaluate aBlock with the receiver's elements as parameters. >> aBlock takes its arguments from the receiver. >> 'ok' >> #(1 2 3) destructDo: [:a :b :c | a + b + c] >> #(1 2 3) destructDo: [:a :b | a + b] >> #(1 2 3) destructDo: [:a | a] >> #(1 2 3) destructDo: [42] >> 'not ok' >> #(1 2 3) destructDo: [:a :b :c :d | a + b + c + d] >> " >> ^aBlock cullWithArguments: self >> asArray >> >> In Amber: >> SequenceableCollection>>destructDo: aBlock >> ^aBlock valueWithPossibleArguments: >> self >> >> In Pharo and other dialects, I don’t know, but should be as easy. >> >> For example you can do >> (('2020-03-28' tokensBasedOn: $-) collect: #asNumber) destructDo: [:year :month :day | >> Date newDay: day monthNumber: month >> year: year] >> >> I like that the block is not the receiver (like with #valueWithArguments or #cullWithArguments), but the last argument. >> >> Now the questions: >> • I am sure that others came up with this. Anybody knows? >> • What are you using for this pattern? >> • I don’t really like the name, but haven’t found anything better yet. Maybe #destructedDo: or just #destructed: or: #withPartsDo:… maybe something shorter? Ideas? >> • What are you thinking about this? >> >> Happy hacking, >> Christian >> >> >> >> >> _______________________________________________ >> Esug-list mailing list >> Esug-list(a)lists.esug.org >> http://lists.esug.org/mailman/listinfo/esug-list_lists.esug.org >> >> >> >> >> _______________________________________________ >> Esug-list mailing list >> >> Esug-list(a)lists.esug.org >> http://lists.esug.org/mailman/listinfo/esug-list_lists.esug.org

March 28, 2020

Re: [Esug-list] destructDo:

by Bernhard Pieber

Hi everyone, I find Martin's arguments regarding "destruct" and "do" rather convincing. Which brings me to: #(1 2 3) destructured: [:a :b :c | a + b + c] Christian, have you ever used this method in a case in which the number of block arguments was smaller than the receiver collection's size? Bernhard > Am 28.03.2020 um 16:13 schrieb Martin McClure <martin.mcclure(a)gemtalksystems.com>: > > Hi Christian and all, > > #destructDo: is probably not the best name. "Destruct" can be interpreted as either short for "destructured" (which is what you want) or more like "destroy" as in "self-destruct." I initially took "destruct" in the "destroy" sense and guessed that #destructDo: might remove each element from the collection as it iterated. You could lengthen it to #destructured: or #destructuredDo:, but if there's a good short name that would be better. > > #tupleDo: feels better. > > But both #tupleDo: and #destructDo: contain "do" which implies iteration. Unless I'm missing something, your #destructDo: does not iterate. I'd really expect a #tupleDo: to iterate -- I'd expect with > > #(1 2 3 4) tupleDo: [:a :b: | a + b] > > that the block would be evaluated twice, once with 1 and 2 and once with 3 and 4. This would be a nice pattern; you could also have a tupleCollect: a tupleDetect: and so on. > > #mapTo: is quite similar to #map: in PetitParser, except IIRC PetitParser's #map: expects an exact match in argument count. But maybe #map: is a good name for consistency. > > Hope this helps. > > Regards, > -Martin > > On 3/28/20 7:24 AM, Christian Haider wrote: >> Yes, that would be nice. Unfortunately a bit more involved than this 1-liner… >> >> Von: Esug-list <esug-list-bounces(a)lists.esug.org> Im Auftrag von Christophe Dony >> Gesendet: Samstag, 28. März 2020 15:17 >> An: esug-list(a)lists.esug.org >> Betreff: Re: [Esug-list] destructDo: >> >> Hi. >> I like tupleDo: >> >> or matchDo: >> >> and what about installing a full pattern matching, as in many languages e.g. Scheme : https://docs.racket-lang.org/reference/match.html >> >> #(1 #(2 3 6 7) 4) matchDo: [:a (:b :@c) :d] | …] >> here c would be #(3 6 7) >> >> may be it already exist? >> if not, would really be a + >> >> Christophe >> >> >> >> Le 28/03/2020 à 14:15, Mark Plas a écrit : >> Hello Christian, >> >> " I don’t really like the name," >> >> Maybe you could call it #tupleDo:? >> >> #(1 2 3) tupleDo: [:a :b :c | …] >> >> Mark >> >> >> >> Mark Plas >> >> Software Engineer >> T +32 2 467 34 30 >> mark.plas(a)mediagenix.tv >> >> >> >> <image001.png> >> >> >> Nieuwe Gentsesteenweg 21/1 >> >> 1702 Groot-Bijgaarden - Belgium >> >> >> <image002.png> <image003.png><image004.png>www.mediagenix.tv >> >> >> Directions to MEDIAGENIX >> >> >> >> This e-mail and any files attached to it are confidential and intended only for the use of the individuals or entity to whom they are addressed. If you have received this e-mail in error, please notify the sender immediately. >> >> From: Esug-list <esug-list-bounces(a)lists.esug.org> On Behalf Of Christian Haider >> Sent: zaterdag 28 maart 2020 13:43 >> To: vwnc(a)cs.uiuc.edu; esug-list(a)lists.esug.org; amber-lang(a)googlegroups.com >> Subject: [Esug-list] destructDo: >> >> Hi, >> >> I am using a nice little method #destructDo: for a while now and it feels good. >> In VisualWorks: >> SequenceableCollection>>destructDo: aBlock >> "Evaluate aBlock with the receiver's elements as parameters. >> aBlock takes its arguments from the receiver. >> 'ok' >> #(1 2 3) destructDo: [:a :b :c | a + b + c] >> #(1 2 3) destructDo: [:a :b | a + b] >> #(1 2 3) destructDo: [:a | a] >> #(1 2 3) destructDo: [42] >> 'not ok' >> #(1 2 3) destructDo: [:a :b :c :d | a + b + c + d] >> " >> ^aBlock cullWithArguments: self asArray >> >> In Amber: >> SequenceableCollection>>destructDo: aBlock >> ^aBlock valueWithPossibleArguments: self >> >> In Pharo and other dialects, I don’t know, but should be as easy. >> >> For example you can do >> (('2020-03-28' tokensBasedOn: $-) collect: #asNumber) destructDo: [:year :month :day | >> Date newDay: day monthNumber: month year: year] >> >> I like that the block is not the receiver (like with #valueWithArguments or #cullWithArguments), but the last argument. >> >> Now the questions: >> • I am sure that others came up with this. Anybody knows? >> • What are you using for this pattern? >> • I don’t really like the name, but haven’t found anything better yet. Maybe #destructedDo: or just #destructed: or: #withPartsDo:… maybe something shorter? Ideas? >> • What are you thinking about this? >> >> Happy hacking, >> Christian >> >> >> >> >> _______________________________________________ >> Esug-list mailing list >> Esug-list(a)lists.esug.org >> http://lists.esug.org/mailman/listinfo/esug-list_lists.esug.org >> >> >> >> >> _______________________________________________ >> Esug-list mailing list >> >> Esug-list(a)lists.esug.org >> http://lists.esug.org/mailman/listinfo/esug-list_lists.esug.org

March 28, 2020

Re: [Esug-list] destructDo:

by Pape, Tobias

> On 28.03.2020, at 17:04, Fabio Niephaus <lists(a)fniephaus.com> wrote: > > > > On Sat, 28 Mar 2020 at 5:00 pm, Norm Green <norm.green(a)gemtalksystems.com> wrote: > I agree with Martin. The default connotation of 'destruct' in English is to destroy. I would suggest either #destructuredDo: or #tuppleDo: to avoid confusion. > > I agree. How about this: > >> #(1 2 3 4) feedInto: [:a :b: | a + b] > Or `#(1 2 3 4) partsIn: [:a :b | a + b]` Best regards -Tobias > Cheers, > Fabio > > > > Norm > > > On 3/28/2020 8:13 AM, Martin McClure wrote: >> Hi Christian and all, >> >> #destructDo: is probably not the best name. "Destruct" can be interpreted as either short for "destructured" (which is what you want) or more like "destroy" as in "self-destruct." I initially took "destruct" in the "destroy" sense and guessed that #destructDo: might remove each element from the collection as it iterated. You could lengthen it to #destructured: or #destructuredDo:, but if there's a good short name that would be better. >> >> #tupleDo: feels better. >> >> But both #tupleDo: and #destructDo: contain "do" which implies iteration. Unless I'm missing something, your #destructDo: does not iterate. I'd really expect a #tupleDo: to iterate -- I'd expect with >> >> #(1 2 3 4) tupleDo: [:a :b: | a + b] >> >> that the block would be evaluated twice, once with 1 and 2 and once with 3 and 4. This would be a nice pattern; you could also have a tupleCollect: a tupleDetect: and so on. >> >> #mapTo: is quite similar to #map: in PetitParser, except IIRC PetitParser's #map: expects an exact match in argument count. But maybe #map: is a good name for consistency. >> >> Hope this helps. >> >> Regards, >> -Martin >> >> On 3/28/20 7:24 AM, Christian Haider wrote: >>> Yes, that would be nice. Unfortunately a bit more involved than this 1-liner… >>> >>> >>> >>> Von: Esug-list <esug-list-bounces(a)lists.esug.org> Im Auftrag von Christophe Dony >>> Gesendet: Samstag, 28. März 2020 15:17 >>> An: esug-list(a)lists.esug.org >>> Betreff: Re: [Esug-list] destructDo: >>> >>> >>> >>> Hi. >>> I like tupleDo: >>> >>> or matchDo: >>> >>> and what about installing a full pattern matching, as in many languages e.g. Scheme : https://docs.racket-lang.org/reference/match.html >>> >>> #(1 #(2 3 6 7) 4) matchDo: [:a (:b :@c) :d] | …] >>> here c would be #(3 6 7) >>> >>> may be it already exist? >>> if not, would really be a + >>> >>> Christophe >>> >>> >>> >>> Le 28/03/2020 à 14:15, Mark Plas a écrit : >>> >>> Hello Christian, >>> >>> >>> >>> " I don’t really like the name," >>> >>> >>> >>> Maybe you could call it #tupleDo:? >>> >>> >>> >>> #(1 2 3) tupleDo: [:a :b :c | …] >>> >>> >>> >>> Mark >>> >>> >>> >>> >>> >>> Mark Plas >>> >>> Software Engineer >>> >>> T +32 2 467 34 30 >>> >>> mark.plas(a)mediagenix.tv >>> >>> >>> >>> >>> >>> >>> >>> Nieuwe Gentsesteenweg 21/1 >>> >>> >>> 1702 Groot-Bijgaarden - Belgium >>> >>> >>> www.mediagenix.tv >>> >>> >>> Directions to MEDIAGENIX >>> >>> >>> >>> >>> This e-mail and any files attached to it are confidential and intended only for the use of the individuals or entity to whom they are addressed. If you have received this e-mail in error, please notify the sender immediately. >>> >>> From: Esug-list <esug-list-bounces(a)lists.esug.org> On Behalf Of Christian Haider >>> Sent: zaterdag 28 maart 2020 13:43 >>> To: vwnc(a)cs.uiuc.edu; esug-list(a)lists.esug.org; amber-lang(a)googlegroups.com >>> Subject: [Esug-list] destructDo: >>> >>> >>> >>> Hi, >>> >>> >>> >>> I am using a nice little method #destructDo: for a while now and it feels good. >>> >>> In VisualWorks: >>> >>> SequenceableCollection>>destructDo: aBlock >>> >>> "Evaluate aBlock with the receiver's elements as parameters. >>> >>> aBlock takes its arguments from the receiver. >>> >>> 'ok' >>> >>> #(1 2 3) destructDo: [:a :b :c | a + b + c] >>> >>> #(1 2 3) destructDo: [:a :b | a + b] >>> >>> #(1 2 3) destructDo: [:a | a] >>> >>> #(1 2 3) destructDo: [42] >>> >>> 'not ok' >>> >>> #(1 2 3) destructDo: [:a :b :c :d | a + b + c + d] >>> >>> " >>> >>> ^aBlock cullWithArguments: self asArray >>> >>> >>> >>> In Amber: >>> >>> SequenceableCollection>>destructDo: aBlock >>> >>> ^aBlock valueWithPossibleArguments: self >>> >>> >>> >>> In Pharo and other dialects, I don’t know, but should be as easy. >>> >>> >>> >>> For example you can do >>> >>> (('2020-03-28' tokensBasedOn: $-) collect: #asNumber) destructDo: [:year :month :day | >>> >>> Date newDay: day monthNumber: month year: year] >>> >>> >>> >>> I like that the block is not the receiver (like with #valueWithArguments or #cullWithArguments), but the last argument. >>> >>> >>> >>> Now the questions: >>> >>> • I am sure that others came up with this. Anybody knows? >>> • What are you using for this pattern? >>> • I don’t really like the name, but haven’t found anything better yet. Maybe #destructedDo: or just #destructed: or: #withPartsDo:… maybe something shorter? Ideas? >>> • What are you thinking about this? >>> >>> >>> Happy hacking, >>> >>> Christian >>> >>> >>> >>> >>> >>> >>> >>> _______________________________________________ >>> Esug-list mailing list >>> Esug-list(a)lists.esug.org >>> http://lists.esug.org/mailman/listinfo/esug-list_lists.esug.org >>> >>> >>> >>> >>> _______________________________________________ >>> Esug-list mailing list >>> >>> Esug-list(a)lists.esug.org >>> http://lists.esug.org/mailman/listinfo/esug-list_lists.esug.org >> >> >> >> _______________________________________________ >> Esug-list mailing list >> >> Esug-list(a)lists.esug.org >> http://lists.esug.org/mailman/listinfo/esug-list_lists.esug.org > > _______________________________________________ > Esug-list mailing list > Esug-list(a)lists.esug.org > http://lists.esug.org/mailman/listinfo/esug-list_lists.esug.org > _______________________________________________ > Esug-list mailing list > Esug-list(a)lists.esug.org > http://lists.esug.org/mailman/listinfo/esug-list_lists.esug.org -- Tobias Pape Software Architecture Group | http://www.hpi.uni-potsdam.de/swa/ Future SOC Lab | https://hpi.de/future-soc-lab Hasso-Plattner-Institut für Digital Engineering gGmbH | Universität Potsdam Prof.-Dr.-Helmert-Str. 2-3, D-14482 Potsdam, Germany Amtsgericht Potsdam, HRB 12184 | Geschäftsführung: Prof. Dr. Christoph Meinel

March 28, 2020

Re: [Esug-list] destructDo:

by Christian Haider

(somehow, I didn’t get Martins mail) Yes, Martin, I agree with the misleading xDo: connotation. Thanks for your nice deconstruction of what I wanted :). Cyril suggests #bind:, which does not feel fully right. How about #boundTo: ? #(1 2 3) boundTo: [:a :b :c | …] Happy hacking, Christian Von: Esug-list <esug-list-bounces(a)lists.esug.org> Im Auftrag von Fabio Niephaus Gesendet: Samstag, 28. März 2020 17:04 An: esug-list(a)lists.esug.org Betreff: Re: [Esug-list] destructDo: On Sat, 28 Mar 2020 at 5:00 pm, Norm Green <norm.green(a)gemtalksystems.com <mailto:norm.green@gemtalksystems.com> > wrote: I agree with Martin. The default connotation of 'destruct' in English is to destroy. I would suggest either #destructuredDo: or #tuppleDo: to avoid confusion. I agree. How about this: #(1 2 3 4) feedInto: [:a :b: | a + b] Cheers, Fabio Norm On 3/28/2020 8:13 AM, Martin McClure wrote: Hi Christian and all, #destructDo: is probably not the best name. "Destruct" can be interpreted as either short for "destructured" (which is what you want) or more like "destroy" as in "self-destruct." I initially took "destruct" in the "destroy" sense and guessed that #destructDo: might remove each element from the collection as it iterated. You could lengthen it to #destructured: or #destructuredDo:, but if there's a good short name that would be better. #tupleDo: feels better. But both #tupleDo: and #destructDo: contain "do" which implies iteration. Unless I'm missing something, your #destructDo: does not iterate. I'd really expect a #tupleDo: to iterate -- I'd expect with #(1 2 3 4) tupleDo: [:a :b: | a + b] that the block would be evaluated twice, once with 1 and 2 and once with 3 and 4. This would be a nice pattern; you could also have a tupleCollect: a tupleDetect: and so on. #mapTo: is quite similar to #map: in PetitParser, except IIRC PetitParser's #map: expects an exact match in argument count. But maybe #map: is a good name for consistency. Hope this helps. Regards, -Martin On 3/28/20 7:24 AM, Christian Haider wrote: Yes, that would be nice. Unfortunately a bit more involved than this 1-liner… Von: Esug-list <mailto:esug-list-bounces@lists.esug.org> <esug-list-bounces(a)lists.esug.org> Im Auftrag von Christophe Dony Gesendet: Samstag, 28. März 2020 15:17 An: esug-list(a)lists.esug.org <mailto:esug-list@lists.esug.org> Betreff: Re: [Esug-list] destructDo: Hi. I like tupleDo: or matchDo: and what about installing a full pattern matching, as in many languages e.g. Scheme : https://docs.racket-lang.org/reference/match.html #(1 #(2 3 6 7) 4) matchDo: [:a (:b :@c) :d] | …] here c would be #(3 6 7) may be it already exist? if not, would really be a + Christophe Le 28/03/2020 à 14:15, Mark Plas a écrit : Hello Christian, " I don’t really like the name," Maybe you could call it #tupleDo:? #(1 2 3) tupleDo: [:a :b :c | …] Mark Mark Plas Software Engineer T +32 2 467 34 30 mark.plas(a)mediagenix.tv <mailto:mark.plas@mediagenix.tv> <https://www.google.com/maps/dir/?api=1&destination=50.872900,4.286429> Nieuwe Gentsesteenweg 21/1 <https://www.google.com/maps/dir/?api=1&destination=50.872900,4.286429> 1702 Groot-Bijgaarden - Belgium <https://nl.linkedin.com/company/mediagenix-ng> <https://twitter.com/mediagenix_tv> <https://www.facebook.com/MEDIAGENIX/> <http://www.mediagenix.tv> www.mediagenix.tv <http://www.mediagenix.tv/wp-content/uploads/2017/08/Directions.pdf> Directions to MEDIAGENIX This e-mail and any files attached to it are confidential and intended only for the use of the individuals or entity to whom they are addressed. If you have received this e-mail in error, please notify the sender immediately. From: Esug-list <mailto:esug-list-bounces@lists.esug.org> <esug-list-bounces(a)lists.esug.org> On Behalf Of Christian Haider Sent: zaterdag 28 maart 2020 13:43 To: vwnc(a)cs.uiuc.edu <mailto:vwnc@cs.uiuc.edu> ; esug-list(a)lists.esug.org <mailto:esug-list@lists.esug.org> ; amber-lang(a)googlegroups.com <mailto:amber-lang@googlegroups.com> Subject: [Esug-list] destructDo: Hi, I am using a nice little method #destructDo: for a while now and it feels good. In VisualWorks: SequenceableCollection>>destructDo: aBlock "Evaluate aBlock with the receiver's elements as parameters. aBlock takes its arguments from the receiver. 'ok' #(1 2 3) destructDo: [:a :b :c | a + b + c] #(1 2 3) destructDo: [:a :b | a + b] #(1 2 3) destructDo: [:a | a] #(1 2 3) destructDo: [42] 'not ok' #(1 2 3) destructDo: [:a :b :c :d | a + b + c + d] " ^aBlock cullWithArguments: self asArray In Amber: SequenceableCollection>>destructDo: aBlock ^aBlock valueWithPossibleArguments: self In Pharo and other dialects, I don’t know, but should be as easy. For example you can do (('2020-03-28' tokensBasedOn: $-) collect: #asNumber) destructDo: [:year :month :day | Date newDay: day monthNumber: month year: year] I like that the block is not the receiver (like with #valueWithArguments or #cullWithArguments), but the last argument. Now the questions: * I am sure that others came up with this. Anybody knows? * What are you using for this pattern? * I don’t really like the name, but haven’t found anything better yet. Maybe #destructedDo: or just #destructed: or: #withPartsDo:… maybe something shorter? Ideas? * What are you thinking about this? Happy hacking, Christian _______________________________________________ Esug-list mailing list Esug-list(a)lists.esug.org <mailto:Esug-list@lists.esug.org> http://lists.esug.org/mailman/listinfo/esug-list_lists.esug.org _______________________________________________ Esug-list mailing list Esug-list(a)lists.esug.org <mailto:Esug-list@lists.esug.org> http://lists.esug.org/mailman/listinfo/esug-list_lists.esug.org _______________________________________________ Esug-list mailing list Esug-list(a)lists.esug.org <mailto:Esug-list@lists.esug.org> http://lists.esug.org/mailman/listinfo/esug-list_lists.esug.org _______________________________________________ Esug-list mailing list Esug-list(a)lists.esug.org <mailto:Esug-list@lists.esug.org> http://lists.esug.org/mailman/listinfo/esug-list_lists.esug.org

March 28, 2020

Re: [Esug-list] destructDo:

by Christian Haider

Yes, exactly the same idea. Cool! -----Ursprüngliche Nachricht----- Von: Esug-list <esug-list-bounces(a)lists.esug.org> Im Auftrag von Cyril Ferlicot D. Gesendet: Samstag, 28. März 2020 17:02 An: esug-list(a)lists.esug.org Betreff: Re: [Esug-list] destructDo: Le 28/03/2020 à 13:42, Christian Haider a écrit : > Hi, > > > > I am using a nice little method #destructDo: for a while now and it > feels good. > > In VisualWorks: > > SequenceableCollection>>destructDo: aBlock > > "Evaluate aBlock with the receiver's > elements as parameters. > > aBlock takes its arguments from the > receiver. > > 'ok' > > #(1 2 3) destructDo: [:a :b :c | a + b > + c] > > #(1 2 3) destructDo: [:a :b | a + b] > > #(1 2 3) destructDo: [:a | a] > > #(1 2 3) destructDo: [42] > > 'not ok' > > #(1 2 3) destructDo: [:a :b :c :d | a > + b + c + d] > > " > > ^aBlock cullWithArguments: self > asArray > > > > In Amber: > > SequenceableCollection>>destructDo: aBlock > > ^aBlock valueWithPossibleArguments: > self > > > > In Pharo and other dialects, I dont know, but should be as easy. > > > > For example you can do > > (('2020-03-28' tokensBasedOn: $-) collect: #asNumber) > destructDo: [:year :month :day | > > Date newDay: day monthNumber: month year: > year] > > > > I like that the block is not the receiver (like with > #valueWithArguments or #cullWithArguments), but the last argument. > > > > Now the questions: > > * I am sure that others came up with this. Anybody knows? > * What are you using for this pattern? > * I dont really like the name, but havent found anything better yet. > Maybe #destructedDo: or just #destructed: or: #withPartsDo: maybe > something shorter? Ideas? > * What are you thinking about this? > > > > Happy hacking, > > Christian > > Hi Christian, In Pharo 8 was introduced a #bind method that seems similar to what you do: SequenceableCollection>>bind: aBlock "Extract items from the receiver and use them as argumeents of aBlock. Signal an error (ArgumentsCountMismatch) when the arguments count of aBlock does not match the receiver size." "(#(1 2 3) bind: [ :first :second :third | first + second + third ]) >>> 6" ^ aBlock valueWithArguments: self The difference is that arguments are not optionals but I think it would be cool to make them optional in this #bind: method. > > > _______________________________________________ > Esug-list mailing list > Esug-list(a)lists.esug.org > http://lists.esug.org/mailman/listinfo/esug-list_lists.esug.org > -- Cyril Ferlicot https://ferlicot.fr

March 28, 2020

Re: [Esug-list] destructDo:

by Fabio Niephaus

On Sat, 28 Mar 2020 at 5:00 pm, Norm Green <norm.green(a)gemtalksystems.com> wrote: > I agree with Martin. The default connotation of 'destruct' in English is > to destroy. I would suggest either #destructuredDo: or #tuppleDo: to avoid > confusion. > I agree. How about this: #(1 2 3 4) feedInto: [:a :b: | a + b] Cheers, Fabio > > Norm > > > On 3/28/2020 8:13 AM, Martin McClure wrote: > > Hi Christian and all, > > #destructDo: is probably not the best name. "Destruct" can be interpreted > as either short for "destructured" (which is what you want) or more like > "destroy" as in "self-destruct." I initially took "destruct" in the > "destroy" sense and guessed that #destructDo: might remove each element > from the collection as it iterated. You could lengthen it to #destructured: > or #destructuredDo:, but if there's a good short name that would be better. > > #tupleDo: feels better. > > But both #tupleDo: and #destructDo: contain "do" which implies iteration. > Unless I'm missing something, your #destructDo: does not iterate. I'd > really expect a #tupleDo: to iterate -- I'd expect with > > #(1 2 3 4) tupleDo: [:a :b: | a + b] > > that the block would be evaluated twice, once with 1 and 2 and once with 3 > and 4. This would be a nice pattern; you could also have a tupleCollect: a > tupleDetect: and so on. > > #mapTo: is quite similar to #map: in PetitParser, except IIRC > PetitParser's #map: expects an exact match in argument count. But maybe > #map: is a good name for consistency. > > Hope this helps. > > Regards, > -Martin > > On 3/28/20 7:24 AM, Christian Haider wrote: > > Yes, that would be nice. Unfortunately a bit more involved than this > 1-liner… > > > > *Von:* Esug-list <esug-list-bounces(a)lists.esug.org> > <esug-list-bounces(a)lists.esug.org> *Im Auftrag von *Christophe Dony > *Gesendet:* Samstag, 28. März 2020 15:17 > *An:* esug-list(a)lists.esug.org > *Betreff:* Re: [Esug-list] destructDo: > > > > Hi. > I like tupleDo: > > or matchDo: > > and what about installing a full pattern matching, as in many languages > e.g. Scheme : https://docs.racket-lang.org/reference/match.html > > #(1 #(2 3 6 7) 4) matchDo: [:a (:b :@c) :d] | …] > here c would be #(3 6 7) > > may be it already exist? > if not, would really be a + > > Christophe > > > > Le 28/03/2020 à 14:15, Mark Plas a écrit : > > Hello Christian, > > > > " I don’t really like the name," > > > > Maybe you could call it #tupleDo:? > > > > #(1 2 3) tupleDo: [:a :b :c | …] > > > > Mark > > > > > > *Mark Plas* > > Software Engineer > > T +32 2 467 34 30 > > mark.plas(a)mediagenix.tv > > > > > > Nieuwe Gentsesteenweg 21/1 > <https://www.google.com/maps/dir/?api=1&destination=50.872900,4.286429> > > 1702 Groot-Bijgaarden - Belgium > <https://www.google.com/maps/dir/?api=1&destination=50.872900,4.286429> > > <https://nl.linkedin.com/company/mediagenix-ng> > <https://twitter.com/mediagenix_tv> <https://www.facebook.com/MEDIAGENIX/> > www.mediagenix.tv > > Directions to MEDIAGENIX > <http://www.mediagenix.tv/wp-content/uploads/2017/08/Directions.pdf> > > > > *This e-mail and any files attached to it are confidential and intended > only for the use of the individuals or entity to whom they are addressed. > If you have received this e-mail in error, please notify the sender > immediately.* > > *From:* Esug-list <esug-list-bounces(a)lists.esug.org> > <esug-list-bounces(a)lists.esug.org> *On Behalf Of *Christian Haider > *Sent:* zaterdag 28 maart 2020 13:43 > *To:* vwnc(a)cs.uiuc.edu; esug-list(a)lists.esug.org; > amber-lang(a)googlegroups.com > *Subject:* [Esug-list] destructDo: > > > > Hi, > > > > I am using a nice little method #destructDo: for a while now and it feels > good. > > In VisualWorks: > > SequenceableCollection>>destructDo: aBlock > > "Evaluate aBlock with the receiver's > elements as parameters. > > aBlock takes its arguments from the > receiver. > > 'ok' > > #(1 2 3) destructDo: [:a :b :c | a + b + c] > > #(1 2 3) destructDo: [:a :b | a + b] > > #(1 2 3) destructDo: [:a | a] > > #(1 2 3) destructDo: [42] > > 'not ok' > > #(1 2 3) destructDo: [:a :b :c :d | a + b > + c + d] > > " > > ^aBlock cullWithArguments: self asArray > > > > In Amber: > > SequenceableCollection>>destructDo: aBlock > > ^aBlock valueWithPossibleArguments: self > > > > In Pharo and other dialects, I don’t know, but should be as easy. > > > > For example you can do > > (('2020-03-28' tokensBasedOn: $-) collect: #asNumber) > destructDo: [:year :month :day | > > Date newDay: day monthNumber: month year: > year] > > > > I like that the block is not the receiver (like with #valueWithArguments > or #cullWithArguments), but the last argument. > > > > Now the questions: > > - I am sure that others came up with this. Anybody knows? > - What are you using for this pattern? > - I don’t really like the name, but haven’t found anything better yet. > Maybe #destructedDo: or just #destructed: or: #withPartsDo:… maybe > something shorter? Ideas? > - What are you thinking about this? > > > > Happy hacking, > > Christian > > > > > > > _______________________________________________ > > Esug-list mailing list > > Esug-list(a)lists.esug.org > > http://lists.esug.org/mailman/listinfo/esug-list_lists.esug.org > > > > _______________________________________________ > Esug-list mailing listEsug-list@lists.esug.orghttp://lists.esug.org/mailman/listinfo/esug-list_lists.esug.org > > > > _______________________________________________ > Esug-list mailing listEsug-list@lists.esug.orghttp://lists.esug.org/mailman/listinfo/esug-list_lists.esug.org > > > _______________________________________________ > Esug-list mailing list > Esug-list(a)lists.esug.org > http://lists.esug.org/mailman/listinfo/esug-list_lists.esug.org >

March 28, 2020

Re: [Esug-list] destructDo:

by Cyril Ferlicot D.

Le 28/03/2020 à 13:42, Christian Haider a écrit : > Hi, > > > > I am using a nice little method #destructDo: for a while now and it > feels good. > > In VisualWorks: > > SequenceableCollection>>destructDo: aBlock > > "Evaluate aBlock with the receiver's > elements as parameters. > > aBlock takes its arguments from the > receiver. > > 'ok' > > #(1 2 3) destructDo: [:a :b :c | a + b + c] > > #(1 2 3) destructDo: [:a :b | a + b] > > #(1 2 3) destructDo: [:a | a] > > #(1 2 3) destructDo: [42] > > 'not ok' > > #(1 2 3) destructDo: [:a :b :c :d | a + > b + c + d] > > " > > ^aBlock cullWithArguments: self asArray > > > > In Amber: > > SequenceableCollection>>destructDo: aBlock > > ^aBlock valueWithPossibleArguments: self > > > > In Pharo and other dialects, I don’t know, but should be as easy. > > > > For example you can do > > (('2020-03-28' tokensBasedOn: $-) collect: #asNumber) > destructDo: [:year :month :day | > > Date newDay: day monthNumber: month year: > year] > > > > I like that the block is not the receiver (like with #valueWithArguments > or #cullWithArguments), but the last argument. > > > > Now the questions: > > * I am sure that others came up with this. Anybody knows? > * What are you using for this pattern? > * I don’t really like the name, but haven’t found anything better yet. > Maybe #destructedDo: or just #destructed: or: #withPartsDo:… maybe > something shorter? Ideas? > * What are you thinking about this? > > > > Happy hacking, > > Christian > > Hi Christian, In Pharo 8 was introduced a #bind method that seems similar to what you do: SequenceableCollection>>bind: aBlock "Extract items from the receiver and use them as argumeents of aBlock. Signal an error (ArgumentsCountMismatch) when the arguments count of aBlock does not match the receiver size." "(#(1 2 3) bind: [ :first :second :third | first + second + third ]) >>> 6" ^ aBlock valueWithArguments: self The difference is that arguments are not optionals but I think it would be cool to make them optional in this #bind: method. > > > _______________________________________________ > Esug-list mailing list > Esug-list(a)lists.esug.org > http://lists.esug.org/mailman/listinfo/esug-list_lists.esug.org > -- Cyril Ferlicot https://ferlicot.fr

March 28, 2020

Re: [Esug-list] destructDo:

by Norm Green

I agree with Martin. The default connotation of 'destruct' in English is to destroy. I would suggest either #destructuredDo: or #tuppleDo: to avoid confusion. Norm On 3/28/2020 8:13 AM, Martin McClure wrote: > Hi Christian and all, > > #destructDo: is probably not the best name. "Destruct" can be > interpreted as either short for "destructured" (which is what you > want) or more like "destroy" as in "self-destruct." I initially took > "destruct" in the "destroy" sense and guessed that #destructDo: might > remove each element from the collection as it iterated. You could > lengthen it to #destructured: or #destructuredDo:, but if there's a > good short name that would be better. > > #tupleDo: feels better. > > But both #tupleDo: and #destructDo: contain "do" which implies > iteration. Unless I'm missing something, your #destructDo: does not > iterate. I'd really expect a #tupleDo: to iterate -- I'd expect with > > #(1 2 3 4) tupleDo: [:a :b: | a + b] > > that the block would be evaluated twice, once with 1 and 2 and once > with 3 and 4. This would be a nice pattern; you could also have a > tupleCollect: a tupleDetect: and so on. > > #mapTo: is quite similar to #map: in PetitParser, except IIRC > PetitParser's #map: expects an exact match in argument count. But > maybe #map: is a good name for consistency. > > Hope this helps. > > Regards, > -Martin > > On 3/28/20 7:24 AM, Christian Haider wrote: >> >> Yes, that would be nice. Unfortunately a bit more involved than this >> 1-liner… >> >> *Von:*Esug-list <esug-list-bounces(a)lists.esug.org> *Im Auftrag von >> *Christophe Dony >> *Gesendet:* Samstag, 28. März 2020 15:17 >> *An:* esug-list(a)lists.esug.org >> *Betreff:* Re: [Esug-list] destructDo: >> >> Hi. >> I like tupleDo: >> >> or matchDo: >> >> and what about installing a full pattern matching, as in many >> languages e.g. Scheme : https://docs.racket-lang.org/reference/match.html >> >> #(1 #(2 3 6 7) 4) matchDo: [:a (:b :@c) :d] | …] >> here c would be #(3 6 7) >> >> may be it already exist? >> if not, would really be a + >> >> Christophe >> >> >> >> Le 28/03/2020 à 14:15, Mark Plas a écrit : >> >> Hello Christian, >> >> "I don’t really like the name," >> >> Maybe you could call it #tupleDo:? >> >> #(1 2 3) tupleDo: [:a :b :c | …] >> >> Mark >> >> *Mark Plas* >> >> Software Engineer >> >> T +32 2 467 34 30 >> >> mark.plas(a)mediagenix.tv <mailto:mark.plas@mediagenix.tv> >> >> >> >> Nieuwe Gentsesteenweg 21/1 >> <https://www.google.com/maps/dir/?api=1&destination=50.872900,4.286429> >> >> >> >> 1702 Groot-Bijgaarden - Belgium >> <https://www.google.com/maps/dir/?api=1&destination=50.872900,4.286429> >> >> >> >> <https://nl.linkedin.com/company/mediagenix-ng> >> <https://twitter.com/mediagenix_tv><https://www.facebook.com/MEDIAGENIX/>www.mediagenix.tv >> <http://www.mediagenix.tv> >> >> >> >> Directions to MEDIAGENIX >> <http://www.mediagenix.tv/wp-content/uploads/2017/08/Directions.pdf> >> >> >> >> /This e-mail and any files attached to it are confidential and >> intended only for the use of the individuals or entity to whom >> they are addressed. If you have received this e-mail in error, >> please notify the sender immediately./ >> >> *From:*Esug-list <esug-list-bounces(a)lists.esug.org> >> <mailto:esug-list-bounces@lists.esug.org> *On Behalf Of >> *Christian Haider >> *Sent:* zaterdag 28 maart 2020 13:43 >> *To:* vwnc(a)cs.uiuc.edu <mailto:vwnc@cs.uiuc.edu>; >> esug-list(a)lists.esug.org <mailto:esug-list@lists.esug.org>; >> amber-lang(a)googlegroups.com <mailto:amber-lang@googlegroups.com> >> *Subject:* [Esug-list] destructDo: >> >> Hi, >> >> I am using a nice little method #destructDo: for a while now and >> it feels good. >> >> In VisualWorks: >> >> SequenceableCollection>>destructDo: aBlock >> >> "Evaluate aBlock with the receiver's elements as >> parameters. >> >> aBlock takes its arguments from the receiver. >> >> 'ok' >> >> #(1 2 3) destructDo: [:a :b :c | a + b + c] >> >> #(1 2 3) destructDo: [:a :b | a + b] >> >> #(1 2 3) destructDo: [:a | a] >> >> #(1 2 3) destructDo: [42] >> >> 'not ok' >> >> #(1 2 3) destructDo: [:a :b :c :d | a + b + c + d] >> >> " >> >> ^aBlock cullWithArguments: self >> asArray >> >> In Amber: >> >> SequenceableCollection>>destructDo: aBlock >> >> ^aBlock >> valueWithPossibleArguments: self >> >> In Pharo and other dialects, I don’t know, but should be as easy. >> >> For example you can do >> >> (('2020-03-28' tokensBasedOn: $-) collect: #asNumber) destructDo: >> [:year :month :day | >> >> Date newDay: day monthNumber: month year: year] >> >> I like that the block is not the receiver (like with >> #valueWithArguments or #cullWithArguments), but the last argument. >> >> Now the questions: >> >> * I am sure that others came up with this. Anybody knows? >> * What are you using for this pattern? >> * I don’t really like the name, but haven’t found anything >> better yet. Maybe #destructedDo: or just #destructed: or: >> #withPartsDo:… maybe something shorter? Ideas? >> * What are you thinking about this? >> >> Happy hacking, >> >> Christian >> >> >> >> >> _______________________________________________ >> >> Esug-list mailing list >> >> Esug-list(a)lists.esug.org <mailto:Esug-list@lists.esug.org> >> >> http://lists.esug.org/mailman/listinfo/esug-list_lists.esug.org >> >> >> _______________________________________________ >> Esug-list mailing list >> Esug-list(a)lists.esug.org >> http://lists.esug.org/mailman/listinfo/esug-list_lists.esug.org > > > _______________________________________________ > Esug-list mailing list > Esug-list(a)lists.esug.org > http://lists.esug.org/mailman/listinfo/esug-list_lists.esug.org

March 28, 2020