January 2011 - Pharo-users - lists.pharo.org

Re: [Pharo-users] Left recursion in PetitParser

by Lukas Renggli

PPParser>>#cycleSet is contained in the package PetitAnalyzer. You can get it from the PetitParser repository at http://source.lukas-renggli.ch/petit. The package contains all kinds of reflective facilities to query and transform grammars. Lukas On 31 January 2011 16:35, Andrew Black <t-ablack(a)microsoft.com> wrote: > > > -----Original Message----- >> From: pharo-users-bounces(a)lists.gforge.inria.fr [mailto:pharo-users-bounces@lists.gforge.inria.fr] On Behalf Of Lukas Renggli >> Sent: 21 January 2011 14:39 >> To: A friendly place where any question about pharo is welcome >> Subject: Re: [Pharo-users] Left recursion in PetitParser > >> > I know how to left-factor the grammar by hand.Â But should I have to? >> > How powerful is thy magic, Lukas? > >> There is Â no magic. > >> You can fix the problem by wrapping one of the parsers in the cycle with #memoize. That could be done automatically (#cycleSet finds the cycles), but I would rather refactor the grammar as it leads to extremely inefficient parsers. > >> If you want to build complicated expression grammars (where typically left-recursion occurs) it might be useful to use the PPExpressionParser, see the class comment for an example. > > I got my parser working by refactoring the grammar; Â Using not to do a little look-ahead and direct things down the correct path helped. > > I saw the tests for memoize, and get the idea of how to use it. Â But I don't find a cycleSet method anywhere ... > > Â Â Andrew > > > -- Lukas Renggli www.lukas-renggli.ch

Jan. 31, 2011

Re: [Pharo-users] Left recursion in PetitParser

by Andrew Black

-----Original Message----- > From: pharo-users-bounces(a)lists.gforge.inria.fr [mailto:pharo-users-bounces@lists.gforge.inria.fr] On Behalf Of Lukas Renggli > Sent: 21 January 2011 14:39 > To: A friendly place where any question about pharo is welcome > Subject: Re: [Pharo-users] Left recursion in PetitParser > > I know how to left-factor the grammar by hand.Â But should I have to?Â > > How powerful is thy magic, Lukas? > There is no magic. > You can fix the problem by wrapping one of the parsers in the cycle with #memoize. That could be done automatically (#cycleSet finds the cycles), but I would rather refactor the grammar as it leads to extremely inefficient parsers. > If you want to build complicated expression grammars (where typically left-recursion occurs) it might be useful to use the PPExpressionParser, see the class comment for an example. I got my parser working by refactoring the grammar; Using not to do a little look-ahead and direct things down the correct path helped. I saw the tests for memoize, and get the idea of how to use it. But I don't find a cycleSet method anywhere ... Andrew

Jan. 31, 2011

Re: [Pharo-users] Scrolling in a Morphic game

by Igor Stasenko

On 29 January 2011 20:39, Aidan Gauland <aidalgol(a)no8wireless.co.nz> wrote: > On Wed, Jan 26, 2011 at 10:51:29AM +0100, Tudor Girba wrote: >> Perhaps you want to take a look at the ScrollPane or GeneralScrollPane implementation. > > I found TransformMorph when looking at ScrollPane. Â I think I'll use > that. >\ Yep. For 'moving' parts of scene add the transform morph and put all other morphs to it, then you can control their relative placement using single object. > I'll post to this list when my game engine is usable, since there is > some interest in the scrolling part. > > Regards, > Aidan > > -----BEGIN PGP SIGNATURE----- > Version: GnuPG v1.4.10 (GNU/Linux) > > iEYEARECAAYFAk1EbOIACgkQjyzTRvYJmaVCoACdEVfyuP27WuuuhQ5S4IBoHdvd > 0QMAn1pLqqst8KrxM75WFAXBbnQ+ZsUn > =A6eJ > -----END PGP SIGNATURE----- > > -- Best regards, Igor Stasenko AKA sig.

Jan. 29, 2011

Re: [Pharo-users] Scrolling in a Morphic game

by Aidan Gauland

On Wed, Jan 26, 2011 at 10:51:29AM +0100, Tudor Girba wrote: > Perhaps you want to take a look at the ScrollPane or GeneralScrollPane implementation. I found TransformMorph when looking at ScrollPane. I think I'll use that. I'll post to this list when my game engine is usable, since there is some interest in the scrolling part. Regards, Aidan

Jan. 29, 2011

Re: [Pharo-users] Screencast proposition: Watch someone read code.

by Igor Stasenko

On 29 January 2011 14:23, Helene Bilbo <thereluctantprogrammer(a)fastmail.fm> wrote: > > > StÃ©phane Ducasse wrote: >> >> Hi helene >> >> Where are you located? >> Because we could do some pair programming. >> > Berlin, Germany > > > StÃ©phane Ducasse wrote: >> >> I did that with olivier yesterday and I understand your point. >> What I could do is to record when I'm programming my card collection >> application and like that the >> world would see me failing and retrying :) >> Stef >> > > Yes of course, i would like to watch a live-programming screencast. > But to me, the difficulty is READING existing code much more than writing > new code. I do the âwrite in the debuggerâ thing, that was mentioned on > pharocast and in some tutorials and it works really nice. > Yeah. Understanding code is difficult. Even your own code could make no sense to you after half of the year :) What i can say.. there is no universal cure.. different people learn things differently. Definitely, you should master the browser and usual shortcuts, so it will help you finding things quicker and follow the code logic in order to understand it. > Helene. > > -- > View this message in context: http://forum.world.st/Screencast-proposition-Watch-someone-read-code-tp3244… > Sent from the Pharo Smalltalk Users mailing list archive at Nabble.com. > > -- Best regards, Igor Stasenko AKA sig.

Jan. 29, 2011

Re: [Pharo-users] Screencast proposition: Watch someone read code.

by laurent laffont

On Sat, Jan 29, 2011 at 11:42 AM, StÃ©phane Ducasse < stephane.ducasse(a)inria.fr> wrote: > Hi helene > > Where are you located? > Because we could do some pair programming. > I did that with olivier yesterday and I understand your point. > What I could do is to record when I'm programming my card collection > application and like that the > world would see me failing and retrying :) > May be someone can record like http://jamesshore.com/Blog/Lets-Play/Episode-37.html . Laurent > > Stef > > On Jan 28, 2011, at 5:04 PM, Helene Bilbo wrote: > > > > > Hi, > > i have a proposition for a Pharo-Screencast: > > > > I would love to look over someoneâs shoulder while she/he is *reading* > code. > > A possible plot would be to download something from squeaksource and then > > trying to make sense out of it - and talking while doing so. (It would > also > > be good to see all the relevant functions of the browser in action). > > > > Background: While i have read some books now (blue book, pharo by > example, > > seaside by example, best practice patterns) i have never actually seen > > people programming. So i absolutely miss the craftsmanship side of > things. > > > > Anyway thank you for Pharocasts. > > Best regards, > > Helene. > > -- > > View this message in context: > http://forum.world.st/Screencast-proposition-Watch-someone-read-code-tp3244… > > Sent from the Pharo Smalltalk Users mailing list archive at Nabble.com. > > > > >

Jan. 29, 2011

Making a .st file from the source of a method

by Sean P. DeNigris

I have a class that I use to set up my images. A subclass has a method: setupScript (Smalltalk at: #UpdateStreamer) new updateFromServer. "Do other stuff..." I want to save the statements of the method above in a .st file to pass as an argument to the vm. That's all working, but in order to retrieve the two statements above, I do: setupScriptContents | scriptMethod sourceString sourceStringWithBracketsStripped | scriptMethod := PharoTrunkImage compiledMethodAt: #setupScript. sourceString := scriptMethod methodNode body asString. sourceStringWithBracketsStripped := sourceString copyFrom: 3 to: sourceString size - 2. ^ sourceStringWithBracketsStripped. That works, but I don't like it. What's the best way to do this (or is this it)? Thanks. Sean -- View this message in context: http://forum.world.st/Making-a-st-file-from-the-source-of-a-method-tp324624… Sent from the Pharo Smalltalk Users mailing list archive at Nabble.com.

Jan. 29, 2011

Re: [Pharo-users] Screencast proposition: Watch someone read code.

by Helene Bilbo

"StÃ©phane Ducasseâ wrote: > > > Now I'm also thinking that we should have a converging approach by also > fixing Morphic so I will start reading both code. > > Stef > Hi StÃ©phane, i quoted this from the Pharo Smalltalk list -> this is the thing you should comment while doing it and record as a screencast. It would be wonderful :) Best regards, Helene. -- View this message in context: http://forum.world.st/Screencast-proposition-Watch-someone-read-code-tp3244… Sent from the Pharo Smalltalk Users mailing list archive at Nabble.com.

Jan. 29, 2011

Re: [Pharo-users] Screencast proposition: Watch someone read code.

by Helene Bilbo

StÃ©phane Ducasse wrote: > > Hi helene > > Where are you located? > Because we could do some pair programming. > Berlin, Germany StÃ©phane Ducasse wrote: > > I did that with olivier yesterday and I understand your point. > What I could do is to record when I'm programming my card collection > application and like that the > world would see me failing and retrying :) > Stef > Yes of course, i would like to watch a live-programming screencast. But to me, the difficulty is READING existing code much more than writing new code. I do the âwrite in the debuggerâ thing, that was mentioned on pharocast and in some tutorials and it works really nice. Helene. -- View this message in context: http://forum.world.st/Screencast-proposition-Watch-someone-read-code-tp3244… Sent from the Pharo Smalltalk Users mailing list archive at Nabble.com.

Jan. 29, 2011

Re: [Pharo-users] Screencast proposition: Watch someone read code.

by Helene Bilbo

âsimondenierâ wrote: > > Read the comments if there are some, especially with snippets of code for > do-it. > Read the tests, try to find how one instantiate a particular object, then > put a self halt in a test or create a new object in a workspace and open > an inspector on it; interact with the object to send it messages and see > how it reacts (with the class browser by the side of course). > > Learn the strengths and weaknesses of the various browsers, and especially > the shortcuts to quickly launch them (select some text in the source > then): > alt+b -> browse class > alt+m -> browse implementors (method definitions) > alt+n -> senders (messages calls) > alt+shit+r -> references to class > etc. > > So basically yes. Understanding a Smalltalk program is somehow a heuristic > process (like all program understanding): browse code, browse implementors > or senders until you can not get further or lost track, make hypothesis > about how it works and check by browsing again and querying live > instances. > > In Smalltalk the language and its syntax are simple, so you get more work > to learn the whole API and the tools, especially since they are a bit > unconventional. So I can understand that you feel lost. But the best way > would be to learn by doing, with somebody while pair-programming. > Thank you for the hints and encouragement. My problems are probably not Smalltalk/Pharo specific - i might reveal that i have no formal training in programming or computer science. I did little things in Objective-C/XCode and Java/Eclipse and i really love the Smalltalk/Pharo,Cuis,Squeak experience compared to that. As i was not trained in other things, Pharo does not at all feel âunconventionalâ to me. Helene. -- View this message in context: http://forum.world.st/Screencast-proposition-Watch-someone-read-code-tp3244… Sent from the Pharo Smalltalk Users mailing list archive at Nabble.com.

Jan. 29, 2011