Take latest pharo1.0-10470-BETAdev09.10.2 (must be a dev image). - Select (say) AST-Nodes category (or any other category) - refactoring scope>super-sends - Yields: Error: Unknown character in literal array The same is observed for self-sends ... Also using the package browser. According to David Roethlisberger, this should be a problem in RB. It encounters a nil character while parsing all the methods, which happens when an unknown character is encountered. The problem appears to be an unknown character somewhere within ClosureCompilerTest>>#testInjectIntoDecompilations Deleting the body of this method "fixes" the problem :-( - on _______________________________________________ Pharo-project mailing list [hidden email] http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project |
I'll look into this. Look like a bug in the RBScanner.
Cheers, Lukas 2009/10/17 Oscar Nierstrasz <[hidden email]>: > > Take latest pharo1.0-10470-BETAdev09.10.2 (must be a dev image). > > - Select (say) AST-Nodes category (or any other category) > - refactoring scope>super-sends > - Yields: Error: Unknown character in literal array > > The same is observed for self-sends ... Also using the package browser. > According to David Roethlisberger, this should be a problem in RB. It > encounters a nil character while parsing > all the methods, which happens when an unknown character is encountered. > > The problem appears to be an unknown character somewhere within > ClosureCompilerTest>>#testInjectIntoDecompilations > > Deleting the body of this method "fixes" the problem :-( > > - on > > > _______________________________________________ > Pharo-project mailing list > [hidden email] > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project > -- Lukas Renggli http://www.lukas-renggli.ch _______________________________________________ Pharo-project mailing list [hidden email] http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project |
excellent!
Noce to see that you are online.. Stef On Oct 17, 2009, at 11:14 AM, Lukas Renggli wrote: > I'll look into this. Look like a bug in the RBScanner. > > Cheers, > Lukas > > 2009/10/17 Oscar Nierstrasz <[hidden email]>: >> >> Take latest pharo1.0-10470-BETAdev09.10.2 (must be a dev image). >> >> - Select (say) AST-Nodes category (or any other category) >> - refactoring scope>super-sends >> - Yields: Error: Unknown character in literal array >> >> The same is observed for self-sends ... Also using the package >> browser. >> According to David Roethlisberger, this should be a problem in RB. >> It >> encounters a nil character while parsing >> all the methods, which happens when an unknown character is >> encountered. >> >> The problem appears to be an unknown character somewhere within >> ClosureCompilerTest>>#testInjectIntoDecompilations >> >> Deleting the body of this method "fixes" the problem :-( >> >> - on >> >> >> _______________________________________________ >> Pharo-project mailing list >> [hidden email] >> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project >> > > > > -- > Lukas Renggli > http://www.lukas-renggli.ch > > _______________________________________________ > Pharo-project mailing list > [hidden email] > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project _______________________________________________ Pharo-project mailing list [hidden email] http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project |
The question is what should the expression
#( . ) should parse as? Neither the ST-80 book, nor the ANSI standard define anything. Another nice example where the Smalltalk language is underspecified ;-) Squeak and VisualWorks parse the dot as a the symbol #'.', which is a good choice, I guess. Unfortunately not that easy to fix in RBScanner ... Lukas -- Lukas Renggli http://www.lukas-renggli.ch _______________________________________________ Pharo-project mailing list [hidden email] http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project |
I fixed this and several other issues in RBScanner. It should again be
able to parse the complete image. Name: AST-lr.176 Author: lr Time: 17 October 2009, 1:24:59 pm UUID: b80f7c84-c776-4f37-a86d-a092d07dd527 Ancestors: AST-DamienCassou.175 - fixed the RBScanner to parse literals like #( . ; ^ [ ] { } ) - removed all conditionals of the RBScanner for other platforms (VisualWorks, IBM) for additional speed - removed unused code from RBScanner - moved RBParser test from Refactoring-Tests to this package, because it tests functionality in this package Cheers, Lukas 2009/10/17 Lukas Renggli <[hidden email]>: > The question is what should the expression > > #( . ) > > should parse as? Neither the ST-80 book, nor the ANSI standard define > anything. Another nice example where the Smalltalk language is > underspecified ;-) > > Squeak and VisualWorks parse the dot as a the symbol #'.', which is a > good choice, I guess. Unfortunately not that easy to fix in RBScanner > ... > > Lukas > > -- > Lukas Renggli > http://www.lukas-renggli.ch > -- Lukas Renggli http://www.lukas-renggli.ch _______________________________________________ Pharo-project mailing list [hidden email] http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project |
In reply to this post by Lukas Renggli
Lukas,
Apologies if this is a waste of time, but my first thought is of something that I have seen thanks to defect 1135 (now closed). When the fonts stopped working, I would frequently see a column of boxes (representing some type of non-printable character??) down the left margin of my code in the browsers. The code in question could have come into Pharo via SIF, which while greatly appreciated for what it is doing for me, is not perfect. The code was exported from Dolphin on Windows; I was probably still doing some of my Pharo work on Windows at that point. So, the offending code crossed dialects and perhaps platforms. Bottom line: I am wondering whether Shout is suppressing non-displayable characters that are confusing the RB?? Bill -----Original Message----- From: [hidden email] [mailto:[hidden email]] On Behalf Of Lukas Renggli Sent: Saturday, October 17, 2009 4:14 AM To: [hidden email] Subject: Re: [Pharo-project] Issue 1311: Refactoring scope is broken I'll look into this. Look like a bug in the RBScanner. Cheers, Lukas 2009/10/17 Oscar Nierstrasz <[hidden email]>: > > Take latest pharo1.0-10470-BETAdev09.10.2 (must be a dev image). > > - Select (say) AST-Nodes category (or any other category) > - refactoring scope>super-sends > - Yields: Error: Unknown character in literal array > > The same is observed for self-sends ... Also using the package browser. > According to David Roethlisberger, this should be a problem in RB. It > encounters a nil character while parsing all the methods, which > happens when an unknown character is encountered. > > The problem appears to be an unknown character somewhere within > ClosureCompilerTest>>#testInjectIntoDecompilations > > Deleting the body of this method "fixes" the problem :-( > > - on > > > _______________________________________________ > Pharo-project mailing list > [hidden email] > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project > -- Lukas Renggli http://www.lukas-renggli.ch _______________________________________________ Pharo-project mailing list [hidden email] http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project _______________________________________________ Pharo-project mailing list [hidden email] http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project |
> Bottom line: I am wondering whether Shout is suppressing non-displayable characters that are confusing the RB??
From what I saw the issues in RB had nothing to do with unprintable characters, but from the question how to parse expressions like #( . ; ^ [ ] { } ). I've changed RB to parse them the same as the Squeak/VW parser does. Lukas -- Lukas Renggli http://www.lukas-renggli.ch _______________________________________________ Pharo-project mailing list [hidden email] http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project |
Free forum by Nabble | Edit this page |