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

[Pharo-project] Seaside image

by Schwab,Wilhelm K

Hello all, I have started to resurrect some of my Seaside code (originally written for 2.8) in the Seaside downstream image for Pharo 1.3 (Seaside 3, I think??). I can't do anything with request contexts in the image. I can try loading Seaside myself in a new image, but I was wondering if anyone has seen this. Bill ----------------- I don't understand what you have been doing wrong. "WACurrenrRequestContext value" will always be available if you are inside a callback or rendering - virtually anywhere in Seaside. Wherever you have a WAComponent you can also send #session to it. On 15 March 2012 18:34, Schwab,Wilhelm K <bschwab at anest.ufl.edu<http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/seaside>> wrote: > I am trying to create anchors that will "dynamically" serve pdf > documents. The archive in question is way too big for a FileLibrary and a > separate web server is probably overkill since I intend to do all of this > locally and would rather not have to run Apache on the machines in question > (security paranoia or laziness, take your pick<g>) . So, I was hoping to > create mime documents on the fly. > > I found suggestions such as: > > component session requestContext respond:[ :response | > response > document:self getPdfBytes > mimeType: 'application/pdf' > fileName: 'seaside-test.pdf'; > doNotCache > > I don't see #getPdfBytes, but suspect I can wing it given the file names > in anchors. > > The problem is that everything I ask for a requestContext feigns > ignorance. I can't even get #session because of the same error. I am > using Seaside 3 from the Pharo 1.3 image and Seaside downstream project. > > Any suggestions? > > Bill

March 16, 2012

[Pharo-project] ArrayedCollection methods depending on Bitmap/BitBlt

by Guillermo Polito

- ArrayedCollection>>restoreEndianness Do we want that in an arrayed collection? It's only senders seem to expect a Bitmap, not any collection. - ArrayedCollection>>swapHalves does not have senders... :/ Does any one use them?? http://code.google.com/p/pharo/issues/detail?id=5492&thanks=5492&ts=1331911… Guille

March 16, 2012

Re: [Pharo-project] Extending SystemBrowser

by Ezequiel Lamonica

Ok, I just started with OB. Thanks! On 15 March 2012 18:46, StÃ©phane Ducasse <stephane.ducasse(a)inria.fr> wrote: > > On Mar 13, 2012, at 9:21 PM, Ezequiel Lamonica wrote: > > > Hi everybody, > > > > I have some problems to extend the SystemBrowser. > > I want to add a panel at the left of package panel to list products. > > Product is a new concept that I created to manage a group of packages. > > > > For example, Pharo is a product and Seaside is another that depends on > Pharo. > > > > This is the path that I want to follow: > > Products -> Packages -> Classes -> Protocols -> Methods > > > > I know in Pharo 1.4 the systemBrowser will be changed but I have a proof > of concept early and I have to show something working integrated in Pharo > 1.3 image. > > Use Glamour or Spec or OB but not systemBrowser. > Run away from SystemBrowser > > > > What could I do to start? > > Any advice? > > > > > > > > >

March 16, 2012

[Pharo-project] Try with Coral

by Alexandre Bergel

Hi! I just gave a try of Coral. Apparently I am not able to run the examples: /users/alexandrebergel/Downloads/Coral> ./coral.sh scriptHappyFace.cst ok, bye... /users/alexandrebergel/Downloads/Coral> ./coral.sh scriptHappyFace.cst ok, bye... /users/alexandrebergel/Downloads/Coral> ./coral.sh scriptCarpet.cst ok, bye... /users/alexandrebergel/Downloads/Coral> In addition: /users/alexandrebergel/Downloads/Coral> ./coral.sh -h Usage: coral.sh [-di] [script-file.coral] coral.sh [-h] -d debug mode: run headful and do not automatically quit -i interactive mode: handle input on a character-wise basis instead of line-wise -h display this help message /users/alexandrebergel/Downloads/Coral> Scripts should ends with .coral or .cst? Cheers, Alexandre -- _,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;: Alexandre Bergel http://www.bergel.eu ^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;._,.;:~^~:;.

March 16, 2012

Re: [Pharo-project] Socket listenOn: 0 & accept

by drush66

Sven Van Caekenberghe wrote > > Most socket API's allow for the creation of a server socket on the next > available port, often by specifying 0 instead of a port. When the socket > is bound, one can retrieve the local port and let the client(s) know. I > tried to do that in Pharo today, and these steps seem to work, by > accepting an incoming connection gives a primitive failed. > > socket listenOn: 0. > I would say that: socket listen. would be nicer interface than giving special meaning to number 0 in this context. Davorin Rusevljan ----- http://www.cloud208.com/ -- View this message in context: http://forum.world.st/Socket-listenOn-0-accept-tp4459813p4477991.html Sent from the Pharo Smalltalk mailing list archive at Nabble.com.

March 16, 2012

Re: [Pharo-project] Does anyone have tried/succeeded to use PetitParser for parsing dates?

by Fabrizio Perin

2012/3/16 StÃ©phane Ducasse <stephane.ducasse(a)inria.fr> > > how did you handle things like > > 1/1/11 (or didn't you?) > > or worse > > 1/1/1 -> 01.01.2001? > This is not related to parsing but to the interpretation of the data :) > > I agree with Stef, your seems to be more an interpretation problem. If you are looking for a parser that can parse any kind of datas you can write something like: #digit asParser plus flatten separatedBy: ('/' asParser / '.' asParser) But to know what you reading you should have some convention to follow. Cheers, Fabrizio There is no magic no? > BTW this is the birthday of one of my sons :) > > > > > how have you treated dates like > > 1/12/2012 -> 12.01.2012 or 1.12.2012? > > > > or even more "problematic" > > 12/12/2012 > > Again why a computer looking at only one sample would be prescient? > > > and the like there are so many way to parse the dates.... > > > > Regards > > Friedrich > > > > -- > > Q-Software Solutions GmbH; Sitz: Bruchsal; Registergericht: Mannheim > > Registriernummer: HRB232138; Geschaeftsfuehrer: Friedrich Dominicus > > > > >

March 16, 2012

Re: [Pharo-project] Unicode depending on Zinc

by StÃ©phane Ducasse

> Sven the problem is that it just hides the problem. >> If Zinc is not loaded then this code is broken code. > > I know ;-) > > My last answer was to the question of how to avoid 'hardcoding a reference'. > > My first answer was: add a method that can generate another method with constant data (since the spec doesn't change often anyway, but it is cool to show in code that you derived the constants from a spec). Yes Stef >

March 16, 2012

Re: [Pharo-project] Unicode depending on Zinc

by Sven Van Caekenberghe

On 16 Mar 2012, at 12:15, StÃ©phane Ducasse wrote: > Sven the problem is that it just hides the problem. > If Zinc is not loaded then this code is broken code. I know ;-) My last answer was to the question of how to avoid 'hardcoding a reference'. My first answer was: add a method that can generate another method with constant data (since the spec doesn't change often anyway, but it is cool to show in code that you derived the constants from a spec).

March 16, 2012

Re: [Pharo-project] Unicode depending on Zinc

by StÃ©phane Ducasse

On Mar 16, 2012, at 9:26 AM, Pavel Krivanek wrote: > - the expression "self environment" is not good because the message > #environment is defined only in classes and for nil, not in Object. environment is for classes not object. = namespace. Marcus your solution is interesting (I would like to know how much it costs) Now for our question it just avoids the the hardcoding but it the code it no present â¦. in VW they have {Core.Zinc} for: [:each | â¦] (I do not have the exact message) now we could have the same. Still it means that using a lot such constructs will put a lot of optionally deadcode in the system. Stef

March 16, 2012

Re: [Pharo-project] Unicode depending on Zinc

by StÃ©phane Ducasse

On Mar 16, 2012, at 11:49 AM, Sven Van Caekenberghe wrote: > If you want to avoid hardcoding a class reference to Zn, you could do a simple download like this: > > 'http://zn.stfx.eu/zn/numbers.txt' asZnUrl retrieveContents Sven the problem is that it just hides the problem. If Zinc is not loaded then this code is broken code. Stef > > It avoids an entry to Undeclared or the ugly test, and it since it is implemented as > > retrieveContents > "Download and return the resource that I refer to. > This will typically return a String or a ByteArray (see ZnClient>>#get:). > If something goes wrong, an exception will be signalled." > > " 'http://zn.stfx.eu/zn/numbers.txt' asZnUrl retrieveContents. " > > ^ ZnClient new > beOneShot; > enforceHttpSuccess: true; > get: self > > it is easy to use for simple scripting. > > Sven > > On 16 Mar 2012, at 00:24, Igor Stasenko wrote: > >> you can always make dependency less hardcoded: >> >> Smalltalk at: #ZnClient ifAbsent: [ self error: 'sorry dude' ] >> ifPresent: [:client | >> ... download stuff , whatever ] >> >> >> On 16 March 2012 00:18, Guillermo Polito <guillermopolito(a)gmail.com> wrote: >>> That's why I generated a method from the file, and then put the link in the >>> comment... >>> >>> >>> On Thu, Mar 15, 2012 at 8:17 PM, Guillermo Polito >>> <guillermopolito(a)gmail.com> wrote: >>>> >>>> But what if I want to have an image without Zinc? Ok, I will not use that >>>> method, but I'll get an Undeclared entry and therefore a dirty image... :/ >>>> >>>> >>>> On Thu, Mar 15, 2012 at 12:37 PM, Sven Van Caekenberghe <sven(a)beta9.be> >>>> wrote: >>>>> >>>>> Hi Guillermo, >>>>> >>>>> On 15 Mar 2012, at 16:25, Guillermo Polito wrote: >>>>> >>>>>> Of course with a nice comment saying something like: >>>>>> >>>>>> "Initialize the mappings from unicode lower to upper case and vice >>>>>> versa. sourceCodes and destinyCodes contains the unicode values for the >>>>>> mappings, and they match one by one. >>>>>> The source of this mapping can be found at: >>>>>> http://www.unicode.org/Public/UNIDATA/CaseFolding.txt' " >>>>> >>>>> Have a look at what I did with >>>>> http://code.google.com/p/pharo/issues/detail?id=4187 >>>>> >>>>> The spec tables are constants, like in your latest example, but I added a >>>>> method that can generate these from an internet accessible resource (using >>>>> Zn as HTTP client). That way, there is some kind of dependency, but not >>>>> really at run time. I think it is cool to literally use these public spec >>>>> documents. >>>>> >>>>> Sven >>>>> >>>> >>> >> >> >> >

March 16, 2012