- Pharo-users - lists.pharo.org

Re: [Pharo-users] Pharo 6.0 Generating Ruby-on-Rails Applications

by H. Hirzel

Maybe I should note that a Magritte is surely useful for ideas. But a solution not using the Magritte modeling approach might be better for the task of actually generating applications for different systems / language environments. Magritte is foremost about modelling domain objects to be used _within_ the image. It is about descriptions (declarative style) enriching the definitions of domain models. A DSL following an imperative style (issuing "build" commands) is very fine as well. The DSL code within the methods might be queried and model information extracted in case of need if it is not explicit. HH. On 9/8/17, phil(a)highoctane.be <phil(a)highoctane.be> wrote: > Nice. I'll definitely have a look because I need to prototype something > quite domain heavy and it may help a lot. > > Phil > > On Fri, Sep 8, 2017 at 5:09 PM, Peter Fisk <peter.fisk(a)gmail.com> wrote: > >> Hi Phil, >> >> I already have PHP support for CodeIgniter and Laravel. >> >> Adding support for Zend Framework might take a day or so to implement. >> >> Any framework which supports MVC and AJAX should be fairly easy to >> integrate >> (eg Python/Django, NodeJS/Adonis, etc). The main difference amongst the >> frameworks lies in the database object mappings. >> >> The specific language support is handled by streams. >> >> There are streams for HTML, CSS, JavaScript, PHP, Ruby, Python and >> sub-streams for Ruby/Rails, JavaScript/Qooxdoo, and so on. >> >> So, the framework is not at all Ruby or Ruby-on-Rails specific. >> >> Cheers, >> >> -- Peter >> >> >> >> >> -- >> Sent from: http://forum.world.st/Pharo-Smalltalk-Users-f1310670.html >> >> >> >

Sept. 9, 2017

Re: [Pharo-users] Standalone HTML Model

by H. Hirzel

Sean, the question is of course - is it the HTML only model you are interested in or is it rather a 'web document model' (i.e. something which is displayed in a browser) thus including HTML and CSS. --HH. On 9/9/17, H. Hirzel <hannes.hirzel(a)gmail.com> wrote: > On 9/9/17, Sean P. DeNigris <sean(a)clipperadams.com> wrote: >> I'd like to create HTML via a DSL, like Seaside's canvas builder, but >> without >> loading a whole web framework. Any ideas? >> > > Hello Sean > > > 1. > One idea is _not_ to go for a HTML builder DSL but for a document > model plus a generator. > The generator is a visitor for the document object model. > > There is the Pillar document model and a visitor already in place. > > The current challenge - actually it is not a challenge - but just > needs some time - is to extract the Pillar document model (easy) and > the tests (needs more time) so that it can be loaded as a stand-alone > package. > > See a long discussion in a thread in August about including Pillar in the > image. > > The same then applies for the HTML generator. That is just a file-out > from an image which has the full Pillar loaded. > > So IMHO it is not so much about inventing a new HTML builder DSL but > to "mold" exisiting infrastructure into shape. > > This has the advantage that you can generate other representations easily. > > > 2. > Another idea is to look at HttpView2 done 10 years ago and come up > with a revised version. > http://wiki.squeak.org/squeak/182 > > "No Files. No HTML. Just Smalltalk code!" > > A builder: http://wiki.squeak.org/squeak/637 > The element hierarchy: http://wiki.squeak.org/squeak/840 > > > Discussion > --------------- > > Comparison of 1 and 2 > > In 1 you use a document model (DOM) which is independent from the > representation in HTML code. > > Than you have a generator generating the HTML code. The DOM is > actually quite close to the HTML DOM. > > The Pillar DOM is nicely supported in Pharo with the inspector. > > Neither 1 nor 2 actually take CSS into account. An issue which is > important when you generate web pages. > > So from the Smalltalk data model point of view (web document modeling) > you have a Smalltalk object which is then rendered on two output > streams > > - HTML > - CSS > > Things get interesting in particular if you want to generate > responsive web sites. > The you need to do more on the CSS side and with generating pictures as > well. > > Regards > Hannes > >> >> ----- >> Cheers, >> Sean >> -- >> Sent from: http://forum.world.st/Pharo-Smalltalk-Users-f1310670.html >> >> >

Sept. 9, 2017

Re: [Pharo-users] Standalone HTML Model

by H. Hirzel

On 9/9/17, Sean P. DeNigris <sean(a)clipperadams.com> wrote: > I'd like to create HTML via a DSL, like Seaside's canvas builder, but > without > loading a whole web framework. Any ideas? > Hello Sean 1. One idea is _not_ to go for a HTML builder DSL but for a document model plus a generator. The generator is a visitor for the document object model. There is the Pillar document model and a visitor already in place. The current challenge - actually it is not a challenge - but just needs some time - is to extract the Pillar document model (easy) and the tests (needs more time) so that it can be loaded as a stand-alone package. See a long discussion in a thread in August about including Pillar in the image. The same then applies for the HTML generator. That is just a file-out from an image which has the full Pillar loaded. So IMHO it is not so much about inventing a new HTML builder DSL but to "mold" exisiting infrastructure into shape. This has the advantage that you can generate other representations easily. 2. Another idea is to look at HttpView2 done 10 years ago and come up with a revised version. http://wiki.squeak.org/squeak/182 "No Files. No HTML. Just Smalltalk code!" A builder: http://wiki.squeak.org/squeak/637 The element hierarchy: http://wiki.squeak.org/squeak/840 Discussion --------------- Comparison of 1 and 2 In 1 you use a document model (DOM) which is independent from the representation in HTML code. Than you have a generator generating the HTML code. The DOM is actually quite close to the HTML DOM. The Pillar DOM is nicely supported in Pharo with the inspector. Neither 1 nor 2 actually take CSS into account. An issue which is important when you generate web pages. So from the Smalltalk data model point of view (web document modeling) you have a Smalltalk object which is then rendered on two output streams - HTML - CSS Things get interesting in particular if you want to generate responsive web sites. The you need to do more on the CSS side and with generating pictures as well. Regards Hannes > > ----- > Cheers, > Sean > -- > Sent from: http://forum.world.st/Pharo-Smalltalk-Users-f1310670.html > >

Sept. 9, 2017

Re: [Pharo-users] Let's prepare for Google Code-In 2017 with Pharo Consortium

by Stephane Ducasse

Hi Excellent initiative. Now could we have a github account to store the idea of small tasks? Stef On Wed, Sep 6, 2017 at 2:46 PM, Jigyasa Grover <grover.jigyasa1(a)gmail.com> wrote: > Hello Pharo-ers ! > > On behalf of the community, I would like to thank each one of you for their > significant contribution in the recently concluded *Google Summer of Code > 2017* with *Pharo Consortium*. We aspire to take-off on a long flight after > this successful stint. In this spirit, I proposed at ESUG 2017 that we take > part in the upcoming *Google Code-In 2017*. > > GCI is a contest to introduce pre-university students (ages 13-17) to open > source software development. Because Google Code-in is often the first > experience many students have with open source, the contest is designed to > make it easy for students to jump right in with /bite-sized tasks/ (3 to 5 > days) having categorised as: Code, Documentation/Training, > Outreach/Research, Quality Assurance, or User Interface. > > Students search for a task from the list created by us that interests them > and claim it. They have at least three days to complete it. The student > works on the task, getting guidance from mentors, if needed. Once the > student is ready, they submit their work for review through the contest > website. Mentors from the organisation evaluate the work, provide feedback, > and if it's complete, accept it so that student can claim a new task. > Otherwise, specific feedback is provided, and the student goes back to > improving it. > > Participants who complete at least 1 task get a digital certificate. > Participants who complete 3 or more tasks receive a t-shirt too. At the end > of the contest, each organisation chooses five finalists to receive limited > edition Google Code-in hoodies. Two finalists from each organisation get to > be the grand prize winners and go on a trip to Google Headquarters in > California ! > > The contest runs for about 6-8 weeks in December & January generally. With > organisation applications beginning in October/November, we need to be well > prepared for this seasonâs application with a hope to get selected. > > As an open source evangelist and a Pharo developer, I would like to invite > all the experienced members to be a part of the "GCI with Pharo Consortium" > Team and mentor students. To get added to the âprivate repositoryâ (we need > to keep prospective students away from the task list before the program > begins ð ) where you can put in your tasks/suggestions kindly EMail me or > any of the previously added mentor. > > Looking forward to an appreciable representation from the Pharo community. > > Best regards > Jigyasa Grover > Pharo Consortium Org Admin, Google Summer of Code 2017 > grover.jigyasa1(a)gmail.com > > Link to know more about GCI: https://codein.withgoogle.com/archive/ > <https://codein.withgoogle.com/archive/ > > Link to sample tasks: > https://developers.google.com/open-source/gci/resources/example-tasks > <https://developers.google.com/open-source/gci/resources/example-tasks> > Link to FAQs: https://developers.google.com/open-source/gci/faq > <https://developers.google.com/open-source/gci/faq > > Link to contest rules: > https://developers.google.com/open-source/gci/resources/contest-rules > <https://developers.google.com/open-source/gci/resources/contest-rules> > Link to GCI presentation for outreach: > https://developers.google.com/open-source/gci/resources/downloads/Google-Co… > <https://developers.google.com/open-source/gci/resources/downloads/Google-Co…> > > > > -- > Sent from: http://forum.world.st/Pharo-Smalltalk-Users-f1310670.html >

Sept. 9, 2017

Re: [Pharo-users] Pharo 6 FileSystem>>changeDirectory: is missing

by Stephane Ducasse

Alistair did I miss some pending PR to be integrated for FS? I'm getting back to reality after a totally fun and intense ESUG conference. Stef On Thu, Sep 7, 2017 at 8:26 AM, Alistair Grant <akgrant0710(a)gmail.com> wrote: > Hi Andreas, > > On Wed, Sep 06, 2017 at 04:53:27PM -0700, Andreas Sunardi wrote: >> Hi Alistair, >> >> I found fogbugz #19717 where this was discussed. >> >> https://pharo.fogbugz.com/f/cases/19717/ >> FileSystem-workingDirectory-wrong-after-image-moved-to-a-new-folder >> >> The issue seems to be the current working directory is saved in instance >> variable workingDirectory. When the image is moved to another directory, this >> workingDirectory becomes incorrect. Instance variable 'store', through its # >> defaultWorkingDirectory, always gives the right answer (however, this answer is >> the image directory, not directory where the command is invoked, which is the >> issue in #05723 you mentioned). >> >> Here's a snippet from fogbugz #19717 original problem on how to reproduce it: >> QUOTE >> 1/ open an image. evaluate './pharo-local' asFileReference and keep the >> inspector. >> 2/ save and quit >> 3/ move the image to another directory >> 4/ open the image >> 5/ evaluate self fullName on the file reference => it will give a wrong path >> with a reference to the working directory used at file reference creation. >> ENDQUOTE > > Thanks for finding this! > > In #20164 I reintroduced caching the working directory, but update it > when the session is stopped and started, so the above test produces the > expected results. As you can see in the comments for the associated PR, > caching the value was a significant peformance improvement for at least > one person. > https://pharo.fogbugz.com/f/cases/20164/Caching-DiskStore-defaultWorkingDir… > > >> I don't know if it's still debatable, which './pharo-local' the file reference >> should refer to, but I think I agree with fogbugz #19717 that it should stay as >> relative path. >> >> But this is a separate issue than losing the ability to change working >> directory, which seems to me an unintentional side-effect in the solution. I'll >> see if I can jump onto $05723 to request support for changing working >> directory. >> >> On a side note, is current working directory the same as image directory in >> Pharo? I wonder why we have these issues if they are separate things. I haven't >> dived into this, so I can't say much. > > Yes, which was the original driver for #05723. After it is integrated > the working directory will be the process working directory. The image > directory can of course still be accessed with FileLocator > imageDirectory. > > Cheers, > Alistair > > >> -- >> Andreas >> >> On Wed, Sep 6, 2017 at 12:42 PM, Alistair Grant <akgrant0710(a)gmail.com> wrote: >> >> On Wed, Sep 06, 2017 at 10:50:05AM -0700, Andreas Sunardi wrote: >> > It isn't only #changeDirectory: method that is missing. Method # >> > workingDirectoryPath: and instance variable workingDirectory are also >> missing. >> > >> > FileSystem >> changeDirectory: aPath >> > self workingDirectoryPath: (self resolve: aPath) >> > >> > FileSystem >> workingDirectoryPath: aPath >> > aPath isAbsolute >> > ifFalse: [ self error: 'Cannot set the working directory to a >> relative >> > path' ]. >> > workingDirectory := aPath >> > >> > >> > To solve my problem, I implemented those and also fixed # >> initializeWithStore: >> > and changed #workingDirectoryPath >> > >> > FileSystem >> initializeWithStore: aStore >> > store := aStore. >> > workingDirectory := store defaultWorkingDirectory >> > >> > FileSystem >> workingDirectoryPath >> > ^ workingDirectory >> > >> > >> > Those are from Pharo 5. However, this breaks Monticello (when opening >> > Monticello Browser). A FileSystem instance has instance variable >> > workingDirectory set to nil. This should be impossible. I set 'self halt' >> in # >> > initializeWithStore:. It doesn't get triggered, which tells me the >> instance >> > isn't created by normal way. I don't know what's going on there. >> > >> > >> > To fix it, I changed #workingDirectoryPath to >> > >> > FileSystem >> workingDirectoryPath >> > workingDirectory ifNil: [ >> > workingDirectory := store defaultWorkingDirectory ]. >> > ^ workingDirectory >> > >> > >> > That solves my problem, but this is a specific tool. I don't know what >> other >> > problems those changes will cause. This #ifNil: guard is not in Pharo 5, >> so >> > that makes me worry. In general, the change in FileSystem gives >> impression that >> > the it is intentional. But I haven't found the new Pharo 6.1 way to >> change >> > working directory, if there's any. >> >> I couldn't find a fogbugz issue relating to this - if anyone knows the >> issue, please post it here as it would be good to understand the >> rationale for the change. >> >> My interpretation of the changes are that the decision was made to hard >> code the working directory to the image directory. There are quite a >> few people who don't agree with this. :-) >> >> There is already a proposed patch to change the working directory to the >> process working directory, see fogbugz #05723. >> https://pharo.fogbugz.com/f/cases/5723/Default-Working-Directory >> >> The patch is waiting on a fix to include UFFI in the kernel. >> >> It currently doesn't allow the working directory to be changed, >> but if I remember correctly, Rajula developed the code to change the >> working directory. >> >> I think this approach has the advantage that it will automtically work >> with forked processes, e.g. using OSProcess / OSSubprocess. >> >> It would be worthwhile adding a comment to the issue asking Rajula to >> add the change directory functionality to the patch. >> >> Cheers, >> Alistair >> >> >> >> > -- >> > Andreas >> > >> > On Tue, Sep 5, 2017 at 12:46 AM, Stephane Ducasse < >> stepharo.self(a)gmail.com> >> > wrote: >> > >> > Thanks for reporting. >> > I do not remember an action around me for this change. >> > Do you have the definition in Pharo 50 at hand? >> > >> > Stef >> > >> > On Wed, Aug 30, 2017 at 11:09 PM, Andreas Sunardi < >> a.sunardi(a)gmail.com> >> > wrote: >> > > I found FileSystem class has changed from Pharo 5 to Pharo 6. I've >> been >> > > using FileSystem>>changeDirectory to make my program (Pharo 5) runs >> in >> > the >> > > current working directory (thus able to find local files). >> > > >> > > This is now broken because #changeDirectory doesn't exist anymore. >> The >> > > change seems intentional. >> > > >> > > Question: Is there a different way in Pharo 6 do set working >> directory or >> > is >> > > this a bug? >> > > >> > > -- >> > > Andreas >

Sept. 9, 2017

Re: [Pharo-users] Spec "bindings"

by Stephane Ducasse

Hi Rob I discussed with peter and I hope that he has more and younger braincells than me. I'm *****super***** busy with lectures preparation, workshop preparation and business issues. So I would love to get some time to get concentrated but not before beginning of October at the minimum. May be when I am at Prague we can discuss with Peter. Stef On Sat, Sep 9, 2017 at 4:43 AM, Rob Rothwell <r.j.rothwell(a)gmail.com> wrote: > > > On Fri, Sep 8, 2017 at 1:24 AM, Stephane Ducasse <stepharo.self(a)gmail.com> > wrote: >> >> I like the idea about your extension. >> We should really get it so that we understand it. > > > I published some experiments on smalltalkhub in a project called > "Spec-Bindings" (owner RobRothwell) that loads into a clean 6.1-64 image. > > It just has some LabelModel and TextModel capabilities with a few tests and > very simple examples. I've never really used smalltalkhub before, though, > so please let me know if I didn't get it set up right. > > I changed and added methods only at this time, but there ComposableModel > probably needs an instance variable for the subject/domain model. I didn't > want to start a new branch of Spec-Core, though, so I created a Presenter > subclass to create new UIs with, but the widgets all have a ValueHolder > already (with different names per widget). > >> >> >> >> > countLabel initializeSubject: subject countHolder >> >> > >> >> > >> > You're welcome. Thank you very much for your time--it's helpful to know >> > there probably isn't some sophisticated wizardly way to do this that I >> > am >> > not even able to see or comprehend! >> >> :) >> >> > >> > >> >> >> >> >> >> > Then, instead of adding a LabelModel to a ComposableModel, you could >> >> > add >> >> > something like a NumberPresenter and specify the spec you want to use >> >> > (which >> >> > would in turn display the desired widget). >> >> > I didn't go that far yet, and created a NumberLabelModel for my >> >> > example >> >> > above. It worked, but seems like the wrong way to go about it. Or >> >> > maybe >> >> > you >> >> > need both...I'm not sure! >> >> > >> >> > I also want to take a look at self-updating complex widgets as well >> >> > (lists, >> >> > tables, etc...) so that telling a list presenter it's list means that >> >> > adding >> >> > a value to that list could automatically update the widget and so on. >> >> > I had some crude prototyping success with that in VW before deciding >> >> > (hopefully for the last time) that "native" widgets aren't as >> >> > important >> >> > to >> >> > me as a solid and simple way to connect a model to a UI presentation. >> >> >> >> >> >> > This seems much easier to achieve in Pharo, and after years of >> >> > intermittent >> >> > attempts I am hopeful that my knowledge level has finally become >> >> > sufficient >> >> > to contribute something to this heroic effort. >> >> >> >> Superb. >> >> >> >> We are discussing with peter (but EsUG is so intense and I should >> >> finish the lectures for real now that I do not have the braincells) >> >> about a pragma driven to declare menus because this is a real lack. >> > >> > >> > Good luck! >> > >> >> >> >> >> >> >> >> What I suggest is >> >> - Please please continue your experiment. >> >> - It would be great if we can compare some typical examples. >> >> - After that we can take a decision and improve/modify Spec >> >> and update the documentation. >> > >> > >> > That's a really good idea since I am no Spec expert! I can probably >> > make a >> > few things work the way I'd like them to, though, in order to produce >> > some >> > typical examples like master-detail lists and so on. >> >> Yes excellent. >> >> >> > >> >> >> >> I want to write a small interface for a small app. >> >> I checked my code >> >> and I thought that it was strange that we cannot initialize presenter >> >> once the model is done >> > >> > >> > Right...you can see in Spec that each type of widget creates a static >> > ValueHolder ( such as '' asValueHolder ) for each type of widget and >> > does >> > not currently offer a method to swap it out with a new one. >> >> yes this is not really good. > > > Hopefully my examples show that it shouldn't be too hard to modify this. > >> >> >> >> > >> > Also, there are probably more available Morphic widgets that could be >> > adapted (like a progress bar as an number presenter, for example). >> > >> >> >> >> >> >> GameListModel new >> >> on: GameCollector smallCollection; >> >> openWithSpec >> >> >> >> >> >> "protocol: #initialization" >> >> >> >> initializeWidgets >> >> listModel := self newList. >> >> listModel items: (collector collectionNamed: #owned) >> >> >> >> and then back then I realized that I was wrong. >> >> To fit Spec (but Spec is certainly too limited) > > > Like you said, it just needs some continuous improvement! > >> >> >> > >> > >> > It would be interesting to hear the arguments against lists and trees >> > updating themselves when their underlying objects change. It looks like >> > Spec, VW, and Pharo all require you to monitor and replace the UI list >> > when >> > something is updated. >> >> mistake :) >> Lack of analysis. >> We are learning while doing it so let us continue. > > > I'll try to recreate my VW experiments, which needed a sort of "complex > object value holder" so you could have something like a Person object in > your list that that displayed some representative string like a full name. > Then, if any of the components of Person (or any of their their > sub-components, like aPerson personName firstName) changed, these changes > propagated to the to top and let the list know that one of it's items had > changed so it knew to update it in the list. > > Take care, > > Rob > >> >> > >> >> >> >> >> >> Stef >> >> >> >> thanks for this discussion. >> > >> > >> > You're very welcome. >> > >> > Take care, >> > >> > Rob >> > >> >> >> >> >> >> > >> >> > Take care, >> >> > >> >> > Rob >> >> > >> >> > On Tue, Sep 5, 2017 at 6:08 PM, Stephane Ducasse >> >> > <stepharo.self(a)gmail.com> >> >> > wrote: >> >> >> >> >> >> Rob >> >> >> >> >> >> Spec deserves another pass. >> >> >> I see your point with showOn: now I do not see how you could avoid >> >> >> the >> >> >> presenter building. But may be I'm not enough into it. >> >> >> So I'm interested in your feedback. >> >> >> >> >> >> Stef >> >> >> >> >> >> Stef >> >> >> >> >> >> On Tue, Sep 5, 2017 at 8:49 PM, Rob Rothwell >> >> >> <r.j.rothwell(a)gmail.com> >> >> >> wrote: >> >> >> > Hello, >> >> >> > >> >> >> > I was wondering what more experienced users than myself thought of >> >> >> > the >> >> >> > idea >> >> >> > of an explicit Spec "connection point" to a domain model object >> >> >> > similar >> >> >> > to >> >> >> > Dolphin's "showOn:" method, like: >> >> >> > >> >> >> > CounterApp showOn: counter. >> >> >> > >> >> >> > This would perhaps trigger something like Dolphin's >> >> >> > Presenter>>model: >> >> >> > message (although I've always found the use of "model" confusing >> >> >> > when >> >> >> > there >> >> >> > are so many "models" involved.) after the widgets had been >> >> >> > created: >> >> >> > >> >> >> > ComposableModel>>initializeBindings: anObject >> >> >> > >> >> >> > In many cases, if your domain model uses ValueHolders as well >> >> >> > (like >> >> >> > Spec >> >> >> > does), making a connection could just mean replacing the Spec >> >> >> > ValueHolder >> >> >> > with your domain ValueHolder so changes to the domain model would >> >> >> > automatically propagate to the UI presentation without providing >> >> >> > explicit >> >> >> > code in ComposableModel>>initializePresenter. >> >> >> > >> >> >> > However, since I am still trying to understand and learn Spec, >> >> >> > it's >> >> >> > quite >> >> >> > possible I am missing some key point and there are reasons not to >> >> >> > explore >> >> >> > this line of thinking! >> >> >> > >> >> >> > Thank you, >> >> >> > >> >> >> > Rob >> >> >> > >> >> >> >> >> > >> >> >> > >> >

Sept. 9, 2017

Re: [Pharo-users] Standalone HTML Model

by Stephane Ducasse

Hi sean I want the same :). Now I have to finish my lectures for within two weeks and I should not read this mailing-list :) I do not remember what we did in Pillar. Now if you extract the canvas/builder from Seaside and rename it to avoid conflict I'm all ears. Stef On Sat, Sep 9, 2017 at 12:15 AM, Sean P. DeNigris <sean(a)clipperadams.com> wrote: > I'd like to create HTML via a DSL, like Seaside's canvas builder, but without > loading a whole web framework. Any ideas? > > > > ----- > Cheers, > Sean > -- > Sent from: http://forum.world.st/Pharo-Smalltalk-Users-f1310670.html >

Sept. 9, 2017

Re: [Pharo-users] STON: Skip Inst Var

by Sven Van Caekenberghe

> On 9 Sep 2017, at 05:44, Sean P. DeNigris <sean(a)clipperadams.com> wrote: > > I was migrating some data from Pharo 4 to Pharo 6.1 and fell over the fact > that FileSystem has a #workingDirectory instVar in the source, but not in > the target. I ended up monkey patching STON until the objects were loaded > and then reverting, but is there a way to skip an inst var via the proper > STON API? STON does not work like NeoJSON where you can define your own mapping. For STON, there is one, system-wide way to serialize/deserialize a specific class. You would have to change the way the offending object get handled. > ----- > Cheers, > Sean > -- > Sent from: http://forum.world.st/Pharo-Smalltalk-Users-f1310670.html >

Sept. 9, 2017

Re: [Pharo-users] #parseMethod:onError: isn't as obvious as expected - is it me?

by Marcus Denker

there is #parseFaultyMethod:, too. We use RBParser for syntax highlighting. We do already have âsendersâ and âimplementors working in the AST when using Suggestions. See SugsMenuBuilder>>#findBestNodeFor: in the case of non-accepted code, it uses the compiler chain with #optionParseErrors (which uses parseFaultyMethod: to parse). context hasUnacceptedEdits ifTrue: [ root := Smalltalk compiler source: context code; options: #(+ optionParseErrors); parse. context selectedClass ifNotNil: [ :theClass | root methodNode methodClass: theClass ]. root doSemanticAnalysis ] This works well but fails when you actually are in an error node.. for the same code the âdumbâ way we use with standard âsenders ofâ just takes line, looks for spaces and feeds that word into the âsenders ofâ.. So to make this work better we need to combine both approaches and deal better with the Error Node content. (e.g. trying to fix it and re-parse or just use the dumb âget the work under the cursorâ way as a first step). Marcus > On 8 Sep 2017, at 03:43, Tim Mackinnon <tim(a)testit.works> wrote: > > Damn you smalltalkâ¦ you suck me in to little side problemsâ¦. > > RBParser is actually a lot more flexible than I had realised - and not so difficult to understand for a mere mortalâ¦ > > It naively seems that (what I still consider a confusingly named method) is quite simple to extend to handle at least medium sized errors just by doing something like: > > ast := [RBParser parseMethod: source onError: [ :msg :pos :parser | |seqNode errorNode| > > errorNode := RBParseErrorNode > errorMessage: msg value: parser currentToken value printString at: pos. > > seqNode := parser sequenceNodeClass new. > parser > step; > parseStatements: true into: seqNode. > > seqNode > addNodeFirst: errorNode; > yourself > > ]] onDNU: #body: do: [ nil âHandle non method source fragments"]. > > > NOTE: I did have to add >>#currentToken as a method to RBParser as it doesnât expose very much for the onError blocks, which I think it should to make them a bit more useful. > > Tim > >> On 7 Sep 2017, at 14:39, Tim Mackinnon <tim(a)testit.works <mailto:tim@testit.works>> wrote: >> >> Thanks Thierry - Iâve been noodling in the background and next I hit the question of âcommentsâ - which I found RB does handleâ¦ and then as per your point - how to skip past an error and try further down (which it seems that the scanner might have the info to let me skip simple things). You seem to get sucked in more and more when trying to to this properly - which does make it sound like a very good internship or GSoc project. >> >> I think I may have enough to get something polished enough to release to the wider world. I certainly love having better keyboard support when Iâm coding - as we type a lot (even though we have simple syntax) - and having efficient tools takes away the friction. >> >> Next on my hit list, it add the same support to the refactoring tools - which while sophisticated are so clunky in use when compared to IntelliJ. >> >> Tim >> >>> On 7 Sep 2017, at 08:02, Thierry Goubier <thierry.goubier(a)gmail.com <mailto:thierry.goubier@gmail.com>> wrote: >>> >>> Hi Tim, >>> >>> 2017-09-01 13:39 GMT+02:00 Tim Mackinnon <tim(a)testit.works <mailto:tim@testit.works>>: >>> Thanks Thierry - this is proving an interesting problem domain ... >>> >>> I too am using #bestNodeFor: although I don't find it always gives the best node (it does if you are clearly in the target range, but if you are on the outer boundary like the next position after a selector or next to a "." or ";" it favours the parent and not the closest node. So in usage I find I need to tweak it a bit. >>> >>> Yes, when you are on a boundary, then finding the right node may be difficult. When developping the ancestor to smart suggestion, I considered #bestNodeFor: as good enough, but I considered looking into "up" and "down" ast navigation at a time, and I wasn't alone (I think it was active in the Pharo editor at a point). >>> >>> >>> I'll look at smacc though - also there is that experimental setting to allow parse errors, I don't know if it helps in any way. >>> >>> There is a general question there, which is how you try to parse as much as possible while jumping over the error, making the error area as small as possible (that would be really great for syntax highlighting, by the way). >>> >>> The problem is that in hand-written parsers, you need to hard-code the error management in the parser. With a table-driven parser (as is SmaCC), then you can explore in a systematic way what are the possible states that would allow the parsing to continue correctly after making the error area as small as possible. >>> >>> This would make for a very nice internship subject... >>> >>> >>> All this said, I think I have a workable suggestion that is not much code that might be open to scrutiny from those wiser than me. >>> >>> Looking at the keyboard handling in the system - it's quite sprawling and tricky to follow. I think there is lots of room for refactoring. >>> >>> It is very easy in Pharo to implement a nice and clean keyboard handling; all the necessary components have been in place for years, and we already have implementations (using the KM API). >>> >>> Now, most Pharo developpers just jump into hardcoding keyboard handling instead. >>> >>> Thierry >>> >>> >>> I'm also not sure if I like the pragma usage either - I find it quite difficult to follow what's calling what. >>> >>> Tim >>> >>> Sent from my iPhone >>> >>> On 1 Sep 2017, at 09:18, Thierry Goubier <thierry.goubier(a)gmail.com <mailto:thierry.goubier@gmail.com>> wrote: >>> >>>> Hi Tim, >>>> >>>> The RB ast has a specific method for finding the right node: in AltBrowser, I use ast bestNodeFor: aTarget selectionInterval. >>>> >>>> For the second case (parse what is selected), you may want to look into SmaCC: it has specific entry points (used for refactoring) to try all possible starts in a parser for a given piece of text and returning all the possible correct asts. As a hand-writen parser, the RBParser can't do that unless you add the necessary entry points by hand and hack around the error handling, because most of the parse attempts end on an error (as they should). >>>> >>>> Regards, >>>> >>>> Thierry >>>> >>>> 2017-09-01 8:50 GMT+02:00 Tim Mackinnon <tim(a)testit.works <mailto:tim@testit.works>>: >>>> Marcus - I'd be interested in how you thought to solve this. >>>> >>>> My solution seems to work - but it's not the most elegant. Retrying lots of different ways like I show, smacks a bit of desperation (this said, it is quite compact). >>>> >>>> Apart from that - I was actually interested in thoughts on the parse method - the onError naming is a bit misleading compared to other methods in the image. I wonder if it should be called something like "onErrorMerge:" to signal that it's going to reuse the results? >>>> >>>> Stef - with regards to broken source, my proposed solution reuses what is already there - thus if you highlight specific code, it reverts to the original strategy and tries to compile just that code. This bugs me however - so if you highlight nothing, it tries to compile the source multiple ways, I drop down to chopping the source up to where your cursor is - which seems like a decent cheap strategy. >>>> >>>> I don't know if our parser is able to recover from simple errors and just have error nodes in the ast, and whether the #bestNodeFor method then ignored error nodes... this is what I think Dolphin used to do. >>>> >>>> Tim >>>> >>>> Sent from my iPhone >>>> >>>> > On 31 Aug 2017, at 19:11, Marcus Denker <marcus.denker(a)inria.fr <mailto:marcus.denker@inria.fr>> wrote: >>>> > >>>> > >>>> >> On 31 Aug 2017, at 19:07, Stephane Ducasse <stepharo.self(a)gmail.com <mailto:stepharo.self@gmail.com>> wrote: >>>> >> >>>> >> On Wed, Aug 30, 2017 at 9:50 PM, Tim Mackinnon <tim(a)testit.works <mailto:tim@testit.works>> wrote: >>>> >>> Iâm looking for some feedback on the usage of RBParser>>#parseMethod:onError:. >>>> >>> >>>> >>> I am looking at ways of improving the way we edit code - and making things work (as least for me - but maybe for everyone) a bit more like IntelliJ where they have some great code operations and very good keyboard support. We are certainly close, but at times feel a bit clumsy - which is a shame as we have far better infrastructure to support equivalent features. So I thought I would have a go. >>>> >> >>>> >> Excellent!!! >>>> >> >>>> >> >>>> >>> Anyway - step one (which I think Iâve come close on), was to teach senders/implementors to look at the AST so you donât have to highlight exactly what you want to search on - instead it can infer from your cursorâ¦ however my next step was to improve how we can select code to ease refactoring, bracketing etcâ¦ >>>> >> >>>> >> Yes but we have to handle broken code then. >>>> >> Did you check how smart suggestions did it for code that compiled? >>>> >> >>>> > >>>> > I looked some weeks ago and made some notes what to doâ¦ but I fear this has to wait for next week, >>>> > there is no way that I can do anything till ESUGâ¦ >>>> > >>>> > Marcus >>>> >>>> >>>> >>> >> >

Sept. 9, 2017

Re: [Pharo-users] Standalone HTML Model

by Sven Van Caekenberghe

There is ZnHtmlOutputStream, not really a DSL, but it might help. > On 9 Sep 2017, at 00:15, Sean P. DeNigris <sean(a)clipperadams.com> wrote: > > I'd like to create HTML via a DSL, like Seaside's canvas builder, but without > loading a whole web framework. Any ideas? > > > > ----- > Cheers, > Sean > -- > Sent from: http://forum.world.st/Pharo-Smalltalk-Users-f1310670.html >

Sept. 9, 2017