August 2012 - Pharo-users - lists.pharo.org

Pharo usage poll

by Esteban Lorenzano

Hi guys, I created a small poll to know which versions of pharo are you using and with which purpose... can you be so kind and fill the form? is really small and can provide some insights about further directions :) https://docs.google.com/spreadsheet/viewform?fromEmail=true&formkey=dFZKeXU… Thanks, Esteban

Aug. 21, 2012

Re: [Pharo-users] NeoJSON to ignore root level

by Milan Mimica

That you very much for your input and a great library. I will use a variation of the second solution. On 20 August 2012 15:34, Sven Van Caekenberghe <sven(a)beta9.be> wrote: > Hi Milan, > > On 20 Aug 2012, at 14:02, Milan Mimica <milan.mimica(a)gmail.com> wrote: > > > Hello > > > > I have obtained a json from a HTTP response which looks like: > > > {"login":{"verified":true,"authCookie":"2eb62b41-5156-4056-9b5d-9252b52c93a2"}} > > > > I want to map this into a Smalltalk class LoginResponse which has two > instance variables: verified and authCookie. So I need the mapper to ignore > the root property name "login". I think that is a quite common use case. > How can it be done? > > > > > > -- > > Milan Mimica > > http://sparklet.sf.net > > I understand your question, but currently there is no out of the box > support for class/type mapping using maps like in your example (there are > other variation too, like class/type properties, nested or not). The reason > is that this involves schema checking which I didn't want to do due to the > added complexity and lack of standardization (this is intentionally outside > the JSON spec proper). > > You have 2 options: > > - either use generic parsing and do the conversions and schema checking > after the parsing (which is what most JSON parser make you do anyway) > > NeoJSONReader fromString: > '{"login":{"verified":true,"authCookie":"2eb62b41-5156-4056-9b5d-9252b52c93a2"}}'. > > - write a custom mapping that looks a bit like a hack > > (NeoJSONReader on: > '{"login":{"verified":true,"authCookie":"2eb62b41-5156-4056-9b5d-9252b52c93a2"}}' > readStream) > for: #LoginResponseWrapper customDo: [ :mapping | > mapping reader: [ :jsonReader | | result | > result := nil. > jsonReader parseMapKeysDo: [ :key | > key = #login > ifTrue: [ result := jsonReader > next ] ]. > result ] ]; > nextAs: #LoginResponseWrapper > > the above example will go into the login element if it is present and > return a dictionary. > > To make it return your class, this should do the trick > > (NeoJSONReader on: > '{"login":{"verified":true,"authCookie":"2eb62b41-5156-4056-9b5d-9252b52c93a2"}}' > readStream) > mapInstVarsFor: LoginResponse; > for: #LoginResponseWrapper customDo: [ :mapping | > mapping reader: [ :jsonReader | | result | > result := nil. > jsonReader parseMapKeysDo: [ :key | > key = #login > ifTrue: [ result := jsonReader > nextAs: LoginResponse ] ]. > result ] ]; > nextAs: #LoginResponseWrapper > > I can image a more general purpose LoginResponseWrapper that maps multiple > types at once. > > HTH, > > Sven > > > > -- > Sven Van Caekenberghe > http://stfx.eu > Smalltalk is the Red Pill > > > > > > -- Milan Mimica http://sparklet.sf.net

Aug. 20, 2012

Re: [Pharo-users] NeoJSON to ignore root level

by Sven Van Caekenberghe

Hi Milan, On 20 Aug 2012, at 14:02, Milan Mimica <milan.mimica(a)gmail.com> wrote: > Hello > > I have obtained a json from a HTTP response which looks like: > {"login":{"verified":true,"authCookie":"2eb62b41-5156-4056-9b5d-9252b52c93a2"}} > > I want to map this into a Smalltalk class LoginResponse which has two instance variables: verified and authCookie. So I need the mapper to ignore the root property name "login". I think that is a quite common use case. How can it be done? > > > -- > Milan Mimica > http://sparklet.sf.net I understand your question, but currently there is no out of the box support for class/type mapping using maps like in your example (there are other variation too, like class/type properties, nested or not). The reason is that this involves schema checking which I didn't want to do due to the added complexity and lack of standardization (this is intentionally outside the JSON spec proper). You have 2 options: - either use generic parsing and do the conversions and schema checking after the parsing (which is what most JSON parser make you do anyway) NeoJSONReader fromString: '{"login":{"verified":true,"authCookie":"2eb62b41-5156-4056-9b5d-9252b52c93a2"}}'. - write a custom mapping that looks a bit like a hack (NeoJSONReader on: '{"login":{"verified":true,"authCookie":"2eb62b41-5156-4056-9b5d-9252b52c93a2"}}' readStream) for: #LoginResponseWrapper customDo: [ :mapping | mapping reader: [ :jsonReader | | result | result := nil. jsonReader parseMapKeysDo: [ :key | key = #login ifTrue: [ result := jsonReader next ] ]. result ] ]; nextAs: #LoginResponseWrapper the above example will go into the login element if it is present and return a dictionary. To make it return your class, this should do the trick (NeoJSONReader on: '{"login":{"verified":true,"authCookie":"2eb62b41-5156-4056-9b5d-9252b52c93a2"}}' readStream) mapInstVarsFor: LoginResponse; for: #LoginResponseWrapper customDo: [ :mapping | mapping reader: [ :jsonReader | | result | result := nil. jsonReader parseMapKeysDo: [ :key | key = #login ifTrue: [ result := jsonReader nextAs: LoginResponse ] ]. result ] ]; nextAs: #LoginResponseWrapper I can image a more general purpose LoginResponseWrapper that maps multiple types at once. HTH, Sven -- Sven Van Caekenberghe http://stfx.eu Smalltalk is the Red Pill

Aug. 20, 2012

NeoJSON to ignore root level

by Milan Mimica

Hello I have obtained a json from a HTTP response which looks like: {"login":{"verified":true,"authCookie":"2eb62b41-5156-4056-9b5d-9252b52c93a2"}} I want to map this into a Smalltalk class LoginResponse which has two instance variables: verified and authCookie. So I need the mapper to ignore the root property name "login". I think that is a quite common use case. How can it be done? -- Milan Mimica http://sparklet.sf.net

Aug. 20, 2012

Re: [Pharo-users] Downloading a .tar.gz file with Zinc in Pharo-1.4(Summer)

by Sven Van Caekenberghe

On 18 Aug 2012, at 00:25, Sven Van Caekenberghe <sven(a)beta9.be> wrote: > Hmm, interesting ! > > I have to study this further and read the specs. After a lot of reading and experimenting, these are my conclusions. If anyone can contribute to this issue, please do. Doing the following works correctly as far as I understand it ZnClient new url: 'http://www.iana.org/time-zones/repository/tzdata-latest.tar.gz'; downloadTo: FileDirectory default pathName; yourself. ZnClient new get: 'http://www.iana.org/time-zones/repository/tzdata-latest.tar.gz'; yourself. The second expression is more interesting because it allows you to inspect the complete response object not using a streaming download. This reveals that the content-type is application/x-tar which is weird because the client requests a file whose primary type is x-gzip. It also reveals that the content-encoding is gzip, which Zn correctly handles by uncompressing the transferred bytes. From this last fact, a content-type of application/x-tar is then again understandable. Given the fact that the server is Apache/2.2.3 (CentOS) this has to be the result of some explicit configuration. My standard apache works differently ZnClient new get: 'http://stfx.eu/test.tar.gz'; yourself. Here, no content-encoding is done and the requested file is served as application/x-gzip, hence Zn will not do any uncompression. What confuses me is that both curl and browsers including safari and firefox seem to things differently: they do not do any uncompression on the first URL which results in the compressed file being downloaded and saved, effectively ignoring the gzip content-encoding. This must be a feature, but I can't find any references describing it. In my understanding of the specs, the following would be a way to request not to use gzip content-encoding, but it does not work as expected ZnClient new url: 'http://www.iana.org/time-zones/repository/tzdata-latest.tar.gz'; headerAt: 'Accept-Encoding' put: 'identity'; downloadTo: FileDirectory default pathName; yourself. Sven

Aug. 20, 2012

Re: [Pharo-users] HowTo make a Morph that accepts dropped files and folders and produces a new FileSystem FileReference in Pharo 1.4

by Helene Bilbo

StÃ©phane Ducasse wrote > > once you got was is not working please report to us the solution :) > Ha, it's something totally different than i thought :))) The only thing Ricardo (Thank you for sharing your image, Ricardo!!) and i did differently was, that he made a real big morph to try it, while i tried with the standard size produced by Morph new. So the thing to find out now is, when Pharo considers a Cursor/Hand to be above a certain Morph. The arrow beeing completely inside the Morph does not seem to be enough. I will write when i found out more. Or i will be happy if someone already knows about it and can tell me.. helene. -- View this message in context: http://forum.world.st/HowTo-make-a-Morph-that-accepts-dropped-files-and-fol… Sent from the Pharo Smalltalk Users mailing list archive at Nabble.com.

Aug. 18, 2012

Re: [Pharo-users] uploaded something into MetacelloRepository by mistake.

by Mariano Martinez Peck

On Fri, Aug 17, 2012 at 11:41 PM, Paul DeBruicker <pdebruic(a)gmail.com>wrote: > Hi - > > I just uploaded ChronosTimeZoneUtilities-**PaulDeBruicker.4.mcz into the > squeaksource MetacelloRepository by mistake. Could someone please delete > it from there? I'll put it in the proper place right now. > > Done. -- Mariano http://marianopeck.wordpress.com

Aug. 18, 2012

Re: [Pharo-users] uploaded something into MetacelloRepository by mistake.

by StÃ©phane Ducasse

> Hi - > > I just uploaded ChronosTimeZoneUtilities-PaulDeBruicker.4.mcz into the squeaksource MetacelloRepository by mistake. No problem paul > Could someone please delete it from there? I'll put it in the proper place right now. We will try :) But there is no stress. Typically it would be good to have a kind of validation that check that only ConfigurationOfXXXXX can be published Stef > > > Thanks. >

Aug. 18, 2012

Re: [Pharo-users] HowTo make a Morph that accepts dropped files and folders and produces a new FileSystem FileReference in Pharo 1.4

by StÃ©phane Ducasse

helene once you got was is not working please report to us the solution :) Stef On Aug 17, 2012, at 4:34 PM, Helene Bilbo wrote: > > Ricardo Moran wrote >> >> Hi, maybe my version of Pharo 1.4 is too old (Pharo1.4a Latest update: >> #14406) or maybe this only works in Windows but I made a morph that >> handles >> dropped files by just overriding these two methods: >> >> wantsDropFiles: anEvent >> ^ true >> >> dropFiles: evt >> "Do what you want here" >> > > That's the first thing i tried. I coudn't find #14406 but i retried in other > old builds also in 1.3 but i always have the same problem: Non on the > methods in this morph get's ever called, because it is the World > (PasteUpMorph) that handles everything, before my morph even get's a chance. > > Might it have to do something with the OS? I run on MacOSX. > > Or can you share an image where this is working? > > > > -- > View this message in context: http://forum.world.st/HowTo-make-a-Morph-that-accepts-dropped-files-and-fol… > Sent from the Pharo Smalltalk Users mailing list archive at Nabble.com. >

Aug. 18, 2012

Re: [Pharo-users] best practices for model/view dependencies

by Johann Hibschman

Thanks, that (along with the other comments here) was enough to get me started. I managed to get a basic proof-of-concept up and running. Oddly enough, I hadn't thought to check the Help browser until I saw the Announcements-Help package. Once I read it, I found it very helpful, so if anyone else comes looking for pointers, that's a good place to direct them. Regards, Johann On Fri, Aug 17, 2012 at 1:32 PM, StÃ©phane Ducasse <stephane.ducasse(a)inria.fr> wrote: > > On Aug 17, 2012, at 3:46 PM, Johann Hibschman wrote: > >> I'm still relatively new to Pharo and to Smalltalk in general, > > welcome :) > >> and >> I've hit a situation where I'd like a model that notifies dependent >> views when it changes. >> >> Classically, it looks like this was done through #addDependent:, >> #update, #changed, #release and friends. I don't need more than a >> symbol message and an associated object, so I could live with that >> API, but I don't quite understand #release. > > Do not use changed:/update because > - it broadcasts to all the dependents > - you have to build case statement in the update > >> It sounds like originally >> Object class held (non-weak) references to all objects that had >> dependents, so unless the dependents were explicitly released, the >> objects would never be released. I get the impression that at some >> point this was switched to using weak references, so I wouldn't have >> to worry about closed parent/dependent loops, but I can't find this >> explicitly spelled out anywhere. >> >> - Question 1: Does the basic addDependent/etc. system on Pharo use >> weak references? > > You can use on:send:to: on Morph or when: anEventSelector send: aMessaageSelector to: anObject on object > You can also use Announcement. > >> - Question 2: Am I correct in understanding that the main point of the >> Model class was to avoid having an entry in the Object class variable, >> so a closed parent/dependent loop could be garbage collected? > > This table in the Object is an old idea and we should remove it. > People should use ValueHolder > >> - Question 3: If Pharo uses weak references, is there any point to the >> Model class? >> >> I've also seen some references to the Announcements framework, but I >> can't find as much documentation on it as I can on the "basic" >> Smalltalk dependency mechanism. It seems a bit more complex, with >> actual event classes, but I've not understood it yet. > > In fact you get an object and not only a symbol so you can pass more logic > >> - Question 4: Should I prefer using Announcements for even mostly trivial cases? > when: anEventSelector > send: aMessaageSelector > to: anObject > > which is should unify with the on:send:to: of Morph > >> - Question 5: What benefits would Announcements give me? > > that you can pass a complex object around. > Read the blog of Vassili on Announcement. > >> >> Any overall pointers would be appreciated as well. >> >> Thanks, >> Johann >> > >

Aug. 18, 2012