April 2012 - Pharo-dev - lists.pharo.org

Re: [Pharo-project] STON (was Re: SimpleLiteralArray Spec)

by Dale Henrichs

Very cool! Let's see ... export the pharo implementation of STON in Cypress package format (using FileTree), then port STON to other dialects and voila:) the next generation of Cypress can use STON in place of JSON! Dale ----- Original Message ----- | From: "Sven Van Caekenberghe" <sven(a)beta9.be> | To: Pharo-project(a)lists.gforge.inria.fr | Sent: Tuesday, April 24, 2012 1:14:08 PM | Subject: Re: [Pharo-project] STON (was Re: SimpleLiteralArray Spec) | | | On 24 Apr 2012, at 21:40, Sven Van Caekenberghe wrote: | | > Apart from the fact that I used single quotes for strings in STON, | > it already works (and I could enhance the parser to accept both I | > guess). | | --- | Name: Ston-Core-SvenVanCaekenberghe.27 | Author: SvenVanCaekenberghe | Time: 24 April 2012, 10:13:07 pm | UUID: dd7c16b2-950c-4246-bf60-130de447cddf | Ancestors: Ston-Core-SvenVanCaekenberghe.26 | | extended the parser to allow either single or double quotes as String | delimiters which gives us JSON backward compatibilty ! (thanks Dale | Henrichs for this great idea). | --- | | | |

April 24, 2012

[Pharo-project] RE : Re: STON (was Re: SimpleLiteralArray Spec)

by Philippe Back

Wow this is great. How to ship binary around? Use case: forms EnvoyÃ© depuis un mobile Samsung Sven Van Caekenberghe <sven(a)beta9.be> a Ã©critÂ : On 24 Apr 2012, at 21:40, Sven Van Caekenberghe wrote: > Apart from the fact that I used single quotes for strings in STON, it already works (and I could enhance the parser to accept both I guess). --- Name: Ston-Core-SvenVanCaekenberghe.27 Author: SvenVanCaekenberghe Time: 24 April 2012, 10:13:07 pm UUID: dd7c16b2-950c-4246-bf60-130de447cddf Ancestors: Ston-Core-SvenVanCaekenberghe.26 extended the parser to allow either single or double quotes as String delimiters which gives us JSON backward compatibilty ! (thanks Dale Henrichs for this great idea). ---

April 24, 2012

Re: [Pharo-project] STON (was Re: SimpleLiteralArray Spec)

by Sven Van Caekenberghe

On 24 Apr 2012, at 21:40, Sven Van Caekenberghe wrote: > Apart from the fact that I used single quotes for strings in STON, it already works (and I could enhance the parser to accept both I guess). --- Name: Ston-Core-SvenVanCaekenberghe.27 Author: SvenVanCaekenberghe Time: 24 April 2012, 10:13:07 pm UUID: dd7c16b2-950c-4246-bf60-130de447cddf Ancestors: Ston-Core-SvenVanCaekenberghe.26 extended the parser to allow either single or double quotes as String delimiters which gives us JSON backward compatibilty ! (thanks Dale Henrichs for this great idea). ---

April 24, 2012

Re: [Pharo-project] STON (was Re: SimpleLiteralArray Spec)

by Sven Van Caekenberghe

On 24 Apr 2012, at 21:05, Dale Henrichs wrote: > Sven, > > Just an off-the-wall thought... If JSON were a subset of STON ... or possibly STON would recognize JSON and parse despite the different syntax...we'd have a nice multipurpose parser/writer that could speak JSON or STON interchangeably... > > Dale That is a great idea, Dale, I never thought of that ! Apart from the fact that I used single quotes for strings in STON, it already works (and I could enhance the parser to accept both I guess). Here is an example using a JSON REST call to convert IP addresses to countries: *nix Shell: $ curl http://easy.t3-platform.net/rest/geo-ip?address=205.140.197.101 {"latitude":38.0,"address":"205.140.197.101","country":"US","longitude":-97.0} Smalltalk: ZnClient new systemPolicy; beOneShot; url: 'http://easy.t3-platform.net/rest/geo-ip'; queryAt: 'address' put: '205.140.197.101'; accept: ZnMimeType applicationJson; contentReader: [ :entity | STON fromString: (entity contents replaceAll: $" with: $') ]; get. a Dictionary('address'->'205.140.197.101' 'country'->'US' 'latitude'->38.0 'longitude'->-97.0 ) Thx, Sven

April 24, 2012

Re: [Pharo-project] STON (was Re: SimpleLiteralArray Spec)

by Dale Henrichs

Sven, Just an off-the-wall thought... If JSON were a subset of STON ... or possibly STON would recognize JSON and parse despite the different syntax...we'd have a nice multipurpose parser/writer that could speak JSON or STON interchangeably... Dale ----- Original Message ----- | From: "Sven Van Caekenberghe" <sven(a)beta9.be> | To: Pharo-project(a)lists.gforge.inria.fr | Sent: Tuesday, April 24, 2012 10:35:49 AM | Subject: Re: [Pharo-project] STON (was Re: SimpleLiteralArray Spec) | | | On 24 Apr 2012, at 19:22, Dale Henrichs wrote: | | > Sven, | > | > I looked at your sample document in the pdf: | > | > TestDomainObject { | > #created : DateAndTime [ '2012-02-14T16:40:15+01:00' ], | > #modified : DateAndTime [ '2012-02-14T16:40:18+01:00' ], | > #integer : 39581, | > #float : 73.84789359463944, | > #description : 'This is a test', | > #color : #green, | > #tags : [ | > #two, | > #beta, | > #medium | > ], | > #bytes : ByteArray [ 'afabfdf61d030f43eb67960c0ae9f39f' ], | > #boolean : false | > } | > | > And this is very readable (to me)... not as many '[]' as I inferred | > from reading the tests:). | > | > I take back my YAML comparison ... this has the nice flavor of | > Smalltalk ... | > | > Dale | | Thanks! | | The main 'hack' is to extend JSON's list/array [] and map/dictionary | {} with a very simple way to instanciate any class, by preceding | them with a class name (or alias) and giving the control to | instanciate to the class or instance. Furthermore, for Array and | Dictionary these are optional. For the concrete representations | choosen for say DateAndTime or ByteArray, human readability was the | deciding factor. But any of these could be changed without hurting | the concept. | | Note that there are also references and implicit object counting and | graph tracing with a defined visit order. In a discussion with | Mariano I kind of decided to no longer treat Strings as | shareable/referenceable objects. This is not yet reflected in the | document but it is in the code. | | Sven |

April 24, 2012

Re: [Pharo-project] SimpleLiteralArray Spec

by Dale Henrichs

Now I can take a crack at your questions:) ----- Original Message ----- | From: "Sven Van Caekenberghe" <sven(a)beta9.be> | To: "Pharo Development" <Pharo-project(a)lists.gforge.inria.fr> | Sent: Tuesday, April 24, 2012 4:26:12 AM | Subject: [Pharo-project] SimpleLiteralArray Spec | ...snip | | Questions remain: | | - is the escaping OK ? | - what about symbols without a leading # ? If you can get rid of extra syntax, that would be good ... | - what should be allowed in unquoted symbols ? | - what should be allowed for classnames ? For Cypress we decided that the name of the directory (className.class) would not necessarily be the real name of the class since the classname rules might vary across different platforms, so something along the same lines might make sense by encoding the real name of the class in a field of the object while preserving the external name ... that way you the external class name could be 'classname' as defined, while the real name of the class could be specified in an optional classname field when the real name can't be mapped to 'classnamed' | - what about namespaces ? For Cypress namespaces are handled by special class attributes , so perhaps there would be an optional class_properties field where additional meta information about the class name resolution can be stored | - do we allow infinite precision integers ? | - maybe the toplevel should always be an array ? I like the idea that the top level can be any legal construct ... | | | Now: shoot! | | | Sven | | | -- | Sven Van Caekenberghe | http://stfx.eu | Smalltalk is the Red Pill | | | | | |

April 24, 2012

[Pharo-project] Editor focus & hotkeys

by blake

So, I'm entering code into a browser (Nautilus browser, though I don't think this behavior is specific to Nautilus), press Ctrl+S to save and... ...the code editor loses focus. Looks like it goes to the method pane, which seems reasonable. But now, to enter my next method, I have to take my hands off the keyboard and click the mouse. Nautilus seems insensate to the Tab key...how do I get from the method pane back to the code editor sans mouse? ===Blake===

April 24, 2012

Re: [Pharo-project] STON (was Re: SimpleLiteralArray Spec)

by Sven Van Caekenberghe

On 24 Apr 2012, at 19:22, Dale Henrichs wrote: > Sven, > > I looked at your sample document in the pdf: > > TestDomainObject { > #created : DateAndTime [ '2012-02-14T16:40:15+01:00' ], > #modified : DateAndTime [ '2012-02-14T16:40:18+01:00' ], > #integer : 39581, > #float : 73.84789359463944, > #description : 'This is a test', > #color : #green, > #tags : [ > #two, > #beta, > #medium > ], > #bytes : ByteArray [ 'afabfdf61d030f43eb67960c0ae9f39f' ], > #boolean : false > } > > And this is very readable (to me)... not as many '[]' as I inferred from reading the tests:). > > I take back my YAML comparison ... this has the nice flavor of Smalltalk ... > > Dale Thanks! The main 'hack' is to extend JSON's list/array [] and map/dictionary {} with a very simple way to instanciate any class, by preceding them with a class name (or alias) and giving the control to instanciate to the class or instance. Furthermore, for Array and Dictionary these are optional. For the concrete representations choosen for say DateAndTime or ByteArray, human readability was the deciding factor. But any of these could be changed without hurting the concept. Note that there are also references and implicit object counting and graph tracing with a defined visit order. In a discussion with Mariano I kind of decided to no longer treat Strings as shareable/referenceable objects. This is not yet reflected in the document but it is in the code. Sven

April 24, 2012

Re: [Pharo-project] STON (was Re: SimpleLiteralArray Spec)

by Dale Henrichs

Sven, For tODE I am very interested having something with this kind of potential ... shipping "real" ST objects back and forth between Amber and Pharo or GemStone has some real appeal... So Goran, the STON format looks somewhat similar to Tirade ... what would Tirade look like for the same data set? Dale ----- Original Message ----- | From: "Dale Henrichs" <dhenrich(a)vmware.com> | To: Pharo-project(a)lists.gforge.inria.fr | Sent: Tuesday, April 24, 2012 10:22:36 AM | Subject: Re: [Pharo-project] STON (was Re: SimpleLiteralArray Spec) | | Sven, | | I looked at your sample document in the pdf: | | TestDomainObject { | #created : DateAndTime [ '2012-02-14T16:40:15+01:00' ], | #modified : DateAndTime [ '2012-02-14T16:40:18+01:00' ], | #integer : 39581, | #float : 73.84789359463944, | #description : 'This is a test', | #color : #green, | #tags : [ | #two, | #beta, | #medium | ], | #bytes : ByteArray [ 'afabfdf61d030f43eb67960c0ae9f39f' ], | #boolean : false | } | | And this is very readable (to me)... not as many '[]' as I inferred | from reading the tests:). | | I take back my YAML comparison ... this has the nice flavor of | Smalltalk ... | | Dale | | ----- Original Message ----- | | From: "Sven Van Caekenberghe" <sven(a)beta9.be> | | To: Pharo-project(a)lists.gforge.inria.fr | | Sent: Tuesday, April 24, 2012 10:06:22 AM | | Subject: [Pharo-project] STON (was Re: SimpleLiteralArray Spec) | | | | Norbert, Dale, et al, | | | | | | On 24 Apr 2012, at 18:43, Norbert Hartl wrote: | | | | > What is STON and STON-object syntax? | | | | Arhh, it is not yet finished, I did it for myself, not for trying | | to | | convince the world, since that it too hard anyway. | | | | This is the work-in-progress paper: | | | | | | | | | | The code lives here: http://ss3.gemstone.com/ss/STON | | | | It is a slight variation of JSON, adding some Smalltalk concepts, | | as | | well as a solution to the class problem. | | | | But again, I am not trying to convince anyone to use this, I did it | | as a proof of concept for myself, and I kind of like it. | | | | On the other hand, JSON is so ubiquitous that I wonder a bit, why | | bother ? | | | | Sven | | | | | | -- | | Sven Van Caekenberghe | | http://stfx.eu | | Smalltalk is the Red Pill | | | | | | | | | | |

April 24, 2012

Re: [Pharo-project] STON (was Re: SimpleLiteralArray Spec)

by Dale Henrichs

Sven, I looked at your sample document in the pdf: TestDomainObject { #created : DateAndTime [ '2012-02-14T16:40:15+01:00' ], #modified : DateAndTime [ '2012-02-14T16:40:18+01:00' ], #integer : 39581, #float : 73.84789359463944, #description : 'This is a test', #color : #green, #tags : [ #two, #beta, #medium ], #bytes : ByteArray [ 'afabfdf61d030f43eb67960c0ae9f39f' ], #boolean : false } And this is very readable (to me)... not as many '[]' as I inferred from reading the tests:). I take back my YAML comparison ... this has the nice flavor of Smalltalk ... Dale ----- Original Message ----- | From: "Sven Van Caekenberghe" <sven(a)beta9.be> | To: Pharo-project(a)lists.gforge.inria.fr | Sent: Tuesday, April 24, 2012 10:06:22 AM | Subject: [Pharo-project] STON (was Re: SimpleLiteralArray Spec) | | Norbert, Dale, et al, | | | On 24 Apr 2012, at 18:43, Norbert Hartl wrote: | | > What is STON and STON-object syntax? | | Arhh, it is not yet finished, I did it for myself, not for trying to | convince the world, since that it too hard anyway. | | This is the work-in-progress paper: | | | | | The code lives here: http://ss3.gemstone.com/ss/STON | | It is a slight variation of JSON, adding some Smalltalk concepts, as | well as a solution to the class problem. | | But again, I am not trying to convince anyone to use this, I did it | as a proof of concept for myself, and I kind of like it. | | On the other hand, JSON is so ubiquitous that I wonder a bit, why | bother ? | | Sven | | | -- | Sven Van Caekenberghe | http://stfx.eu | Smalltalk is the Red Pill | | | | |

April 24, 2012