November 2009 - Pharo-dev - lists.pharo.org

Re: [Pharo-project] Fwd: nice bug with regex

by Lukas Renggli

2009/11/5 Lukas Renggli <renggli(a)gmail.com>: > I don't know, I did not write the code. I trust it, because it passes > the test suite of Henry Spencer's regexp.c package. > > Lukas > > > 2009/11/4 Nicolas Cellier <nicolas.cellier.aka.nice(a)gmail.com>: >> This version does not seem to solve http://bugs.squeak.org/view.php?id=5391 >> >> { >> '15' matchesRegex: '[1-9]|1[0-9]'. >> '15' matchesRegex: '1[0-9]|[1-9]'. >> } #(false true) I am not saying that this bug is not worth too look at and to fix. I am just no expert in regular expressions and I don't know what the desired behavior is? And why? Still looking for a standard that describes the semantics of regular expressions. I think everybody agrees that the updated code is an improvement, even if it does not fix all problems. No? Lukas >> >> >> 2009/11/4 Lukas Renggli <renggli(a)gmail.com>: >>> I committed VB-Regex-lr.34 to the PharoInbox that fixes this issue. >>> >>> In fact, this is a completely new port of VB-Regex with numerous >>> bug-fixes and enhancements. The bug you found was fixed in version >>> 1.1c in December 2004. The version 1.1 we used up to now was from >>> October 1999. >>> >>> The detailed log of changes from Vassily you can see below. >>> >>> Cheers, >>> Lukas >>> >>> Name: VB-Regex-lr.34 >>> Author: lr >>> Time: 4 November 2009, 10:40:27 pm >>> UUID: 9254d5aa-3e22-4e3b-83aa-2d633ff05570 >>> Ancestors: VB-Regex-StephaneDucasse.33 >>> >>> VERSION 1.2.3 (November 2007) >>> >>> 1. Regexs with ^ or $ applied to copy empty strings caused infinite >>> loops, e.g. ('' copyWithRegex: '^.*$' matchesReplacedWith: 'foo'). >>> Applied a similar correction to that from version 1.1c, to >>> #copyStream:to:(replacingMatchesWith:|translatingMatchesUsing:). >>> 2. Extended RxParser testing to run each test for >>> #copy:translatingMatchesUsing: as well as #search:. >>> 3. Corrected #testSuite test that a dot does not match a null, which >>> was passing by luck with Smalltalk code in a literal array. >>> 4. Added test to end of test suite for fix 1 above. >>> >>> VERSION 1.2.2 (November 2006) >>> >>> There was no way to specify a backslash in a character set. Now [\\] >>> is accepted. >>> >>> VERSION 1.2.1 Â (August 2006) >>> >>> 1. Support for returning all ranges (startIndex to: stopIndex) >>> matching a regex - #allRangesOfRegexMatches:, #matchingRangesIn: >>> 2. Added hint to usage documentation on how to get more information >>> about matches when enumerating >>> 3. Syntax description of dot corrected: matches anything but NUL since 1.1a >>> >>> VERSION 1.2 Â Â (May 2006) >>> >>> Fixed case-insensitive search for character sets. >>> >>> VERSION 1.1c Â Â (December 2004) >>> >>> Fixed the issue with #matchesOnStream:do: which caused infinite loops >>> for matches >>> that matched empty strings. >>> >>> VERSION 1.1b Â Â (November 2001) >>> >>> Changes valueNowOrOnUnwindDo: to ensure:, plus incorporates some earlier fixes. >>> >>> VERSION 1.1a Â Â (May 2001) >>> >>> 1. Support for keeping track of multiple subexpressions. >>> 2. Dot (.) matches anything but NUL character, as it should per POSIX spec. >>> 3. Some bug fixes. >>> >>> >>> >>> 2009/11/4 StÃ©phane Ducasse <stephane.ducasse(a)inria.fr>: >>>>> l := '1Ã¨re partie\nmilieu\nfin'. >>>>> '.*' asRegex matchesIn l. >>>> >>>> -> infinite and freezing loop. >>>> >>>> Stef >>>> >>>> >>>> >>>> >>>> _______________________________________________ >>>> Pharo-project mailing list >>>> Pharo-project(a)lists.gforge.inria.fr >>>> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project >>>> >>> >>> >>> >>> -- >>> Lukas Renggli >>> http://www.lukas-renggli.ch >>> >>> _______________________________________________ >>> Pharo-project mailing list >>> Pharo-project(a)lists.gforge.inria.fr >>> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project >>> >> >> _______________________________________________ >> Pharo-project mailing list >> Pharo-project(a)lists.gforge.inria.fr >> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project >> > > > > -- > Lukas Renggli > http://www.lukas-renggli.ch > -- Lukas Renggli http://www.lukas-renggli.ch

Nov. 5, 2009

Re: [Pharo-project] Announcements question

by keith

On 4 Nov 2009, at 23:58, HernÃ¡n Morales Durand wrote: > Hi Keith, thanks for your reply. > I've tried your implementation a bit this way: Ok, the package is SeaWA-Core, First of all we add a double dispatch back to the Announcement, enabling the Announcement subclass to be able to specialize announcing policy. e.g. the default is to announce to all dependents, but you might want to announce to all dependents that meet a certain criteria. The class comment describes this as follows: ====== Unlike normal announcers which figure out what to announce to whom, as much responsibility is given to each Announcement as possible so that behaviour can be overriden completely for different scenarios. When handling an announcement we double back and send #announceTo: to the announcement itself to give it complete control, over what it does before or after informing dependants. Its default behaviour is #dependantsAnnounce . This allows an Announcement to replace the behaviour of Announcer-#dependantsAnnounce: altogether if it wants to. ======== A second enhancement is to return the announcement from the invocation of the trigger. Though I cant remember exactly why this is now. ======= Triggering example: (we always return the original announcement) theAnnouncement := self announcer announce: (SomeAnnouncement param: arg1). ======= Now here we come to the interesting bit. =============== Subscriptions example: self announcer on: SomeAnouncement do: [ :announcement | ... ]. self announcer on: SomeAnouncement send: #tellMe: to: anObject. Normally, the announcer is passed as the argument, however, each announcement class defines #args, so it can call a method that is not expecting an Announcement as the parameter. This enables you to wire up objects that were not designed with announcements in mind. example trigger: self announce: (HtmlUpdateAnnouncement on: html) self announcer on: HtmlUpdateAnouncement send: #renderOn: to: anObject. where: HtmlUpdateAnnouncement-#args ^ Array with: html ================= does that help? Keith

Nov. 5, 2009

Re: [Pharo-project] TestRunner ENH

by keith

> > > Why would I want to run my tests faster or slower ? I would always > choose fast ahahahha > > Of course, there is something I am missing. > > Kind regards, > > Mariano > In SUnit-improved TestRunner the time taken to run each test was recorded. Then there is an option to automatically categorize tests as (slow) , (long) or (net), where you can configure what time span is considered "slow", or "long" Then on subsequent runs you can organize to run the quick tests first, and you get something like 90%+ coverage in the first minute, with the slow and long tests to follow. Keith

Nov. 5, 2009

[Pharo-project] A couple more thoughts about Announcements

by Igor Stasenko

As far as i can see there is no 'capture' scenario, i.e. tell announcer to not broadcast the announcement to all subscribers, but only exclusively to a single one, who requested the capture. In AXAnnouncements there is an 'interception' , using which you can achieve the similar result, except that interceptor is added to each subscription, meaning that you will receive intercepted announcement + additional data as many times as the number of subscriptions registered. Such scenario, as capturing events used heavily in UI, but not only by UI of course. So, suppose we made a Sensor which using announcer to broadcast its events. And there are cases, when you need to get an exclusive capture of all (or specific kind of events) for a single subscriber, without letting them handled using normal flow, until some condition met, when you are releasing capture. Two examples: - mouse capture usually, when you start dragging the stuff, you start capturing mouse events, and release the capture once mouse button released. All widgets, over which you dragging a mouse, should not receive any mouse events, or receive a different event such as 'draggedOver' etc - keyboard focus pass all keyboard events directly to widget which currently has focus - some code examples, by using brute force establishing a loop which consuming all sensor events until mouse button released or key pressed. All of these cases can be supported by adding few things to announcer: - tell announcer to insert a subscriber at the starting of the subscribers list (instead at the end), so it will be first who receive the announcement - make announcer to stop broadcasting announcement down the list, if particular subscription returns false as result of #handle: - when you unsubscribe, things get back to normal without any extra effort and voila.. we can implement a mouse capture, keyboard focus, filtering and event transforming scenarios. -- Best regards, Igor Stasenko AKA sig.

Nov. 5, 2009

Re: [Pharo-project] Announcements question

by HernÃ¡n Morales Durand

Hi Keith, thanks for your reply. I've tried your implementation a bit this way: | testAnn | testAnn := WAAnnouncer with: { 1 . 2 }. testAnn subscribe: AnnouncementMockA send: #value: to: ExampleObject new. ... testAnn announce: AnnouncementMockA. but in ExampleObject>>value: anObject anObject happens to be an AnnouncementMockA and not { 1 . 2 } which is what I'm trying to achieve. I know I can do anything with the AnnouncementMockA at signalling time, but I have the collection { 1 . 2 } only at registering time. Do you have an example of this behavior with your implementation? Just thinking, I'm wondering why a subscriber must be a class : eventSource subscribe: AnnouncementMockA send: #value: to: anObject. but not an instance, it could be: eventSource subscribe: ( AnnouncementMockA named: 'a' with: argument ) send: #value: to: anObject. this way would prevent announcement class explosion too. HernÃ¡n 2009/11/4 keith <keith_hodges(a)yahoo.co.uk> > > On 4 Nov 2009, at 21:50, HernÃ¡n Morales Durand wrote: > > > Hi, I'm using AXAnnouncements but I suppose this question is valid > > for other implementations. In SASE (the "triggerEvent" event > > framework) one could subscribe an event for an object specifying an > > argument at registration time: > > > > eventSource > > when: #eventName send: #aSelector to: anObject with: argument > > > > I've looked at Announcements and it seems this kind of registration > > isn't supported without using blocks. Am I right? > > Before getting into details I just want to check. What people do to > > specify an argument at registration time with Announcements? > > > > Cheers, > > > > HernÃ¡n > > Your announcement is an object of your own design. You can put > parameters as instance variables in your announcement subclass. > > The default implementation does not support the ability to send > methods to objects in the manner that you describe. > My Wandering Announcements package does have an enhancement to support > this, > it is available in the squeaksource.com/Jetsam repository. > > Keith > _______________________________________________ > Pharo-project mailing list > Pharo-project(a)lists.gforge.inria.fr > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project >

Nov. 4, 2009

[Pharo-project] stupid but very intrigued question

by Esteban Lorenzano

Hi, I was downloading a new pharo-core image (to load some iphone development tools into it), and I see the pharo-core alpha for 1.1... and the image size diference is SO BIG, that I wonder why... What is droped to get an image so small? Cheers, Esteban

Nov. 4, 2009

Re: [Pharo-project] TestRunner ENH

by Mariano Martinez Peck

On Wed, Nov 4, 2009 at 12:57 PM, StÃ©phane Ducasse <stephane.ducasse(a)inria.fr > wrote: > I would like to have some feedback on > http://code.google.com/p/pharo/issues/detail?id=1392 > > Issue 1392: Adapted SUnit TestRunner can not be integrated easily in > environment > > I've build some extension on SUnit: > > http://www.squeaksource.com/SUnitsSlowNFast/ > . It automagically can run fast and slow tests (based on yesterdays weather). > > Besides the problem you ask, about being able to integrate this to the core image, I don't understand the features or goal of your TestRunner. Why would I want to run my tests faster or slower ? I would always choose fast ahahahha Of course, there is something I am missing. Kind regards, Mariano > However if i want to integrate this testrunner instead of the standard > testrunner, there is no way 2 do it without changing existing code. I > want > something that i can configure my testrunner instead of the default one. > > If i look at the > StandardToolSet>>class>>openTestRunner > Smalltalk at: #TestRunner ifPresent: [ :class | ^ class open ]. > > You notice that TestRunner is hardcoded chosen. > > My suggestion is to adapt the same solution as in the > > StandardToolSet>>class>>openClassBrowser > SystemBrowser default open > > message: another layer of indirection solves every problem and > introduce a > TestRunnerBrowser class, which is a subclass of AppRegistry, and there > the > default TestRunner is configured (which should be initialized in > TestRunner>>class>>initialize using self). So that if i load my new > TestRunner (which is a subclass of the TestRunner class) included in my > SUnitsSlowNFast package, mine is automagically configured in the > environment as the default one. > > In attachments i've created a solution. > > > > _______________________________________________ > Pharo-project mailing list > Pharo-project(a)lists.gforge.inria.fr > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project >

Nov. 4, 2009

Re: [Pharo-project] Fwd: nice bug with regex

by Lukas Renggli

I don't know, I did not write the code. I trust it, because it passes the test suite of Henry Spencer's regexp.c package. Lukas 2009/11/4 Nicolas Cellier <nicolas.cellier.aka.nice(a)gmail.com>: > This version does not seem to solve http://bugs.squeak.org/view.php?id=5391 > > { > '15' matchesRegex: '[1-9]|1[0-9]'. > '15' matchesRegex: '1[0-9]|[1-9]'. > } #(false true) > > > 2009/11/4 Lukas Renggli <renggli(a)gmail.com>: >> I committed VB-Regex-lr.34 to the PharoInbox that fixes this issue. >> >> In fact, this is a completely new port of VB-Regex with numerous >> bug-fixes and enhancements. The bug you found was fixed in version >> 1.1c in December 2004. The version 1.1 we used up to now was from >> October 1999. >> >> The detailed log of changes from Vassily you can see below. >> >> Cheers, >> Lukas >> >> Name: VB-Regex-lr.34 >> Author: lr >> Time: 4 November 2009, 10:40:27 pm >> UUID: 9254d5aa-3e22-4e3b-83aa-2d633ff05570 >> Ancestors: VB-Regex-StephaneDucasse.33 >> >> VERSION 1.2.3 (November 2007) >> >> 1. Regexs with ^ or $ applied to copy empty strings caused infinite >> loops, e.g. ('' copyWithRegex: '^.*$' matchesReplacedWith: 'foo'). >> Applied a similar correction to that from version 1.1c, to >> #copyStream:to:(replacingMatchesWith:|translatingMatchesUsing:). >> 2. Extended RxParser testing to run each test for >> #copy:translatingMatchesUsing: as well as #search:. >> 3. Corrected #testSuite test that a dot does not match a null, which >> was passing by luck with Smalltalk code in a literal array. >> 4. Added test to end of test suite for fix 1 above. >> >> VERSION 1.2.2 (November 2006) >> >> There was no way to specify a backslash in a character set. Now [\\] >> is accepted. >> >> VERSION 1.2.1 Â (August 2006) >> >> 1. Support for returning all ranges (startIndex to: stopIndex) >> matching a regex - #allRangesOfRegexMatches:, #matchingRangesIn: >> 2. Added hint to usage documentation on how to get more information >> about matches when enumerating >> 3. Syntax description of dot corrected: matches anything but NUL since 1.1a >> >> VERSION 1.2 Â Â (May 2006) >> >> Fixed case-insensitive search for character sets. >> >> VERSION 1.1c Â Â (December 2004) >> >> Fixed the issue with #matchesOnStream:do: which caused infinite loops >> for matches >> that matched empty strings. >> >> VERSION 1.1b Â Â (November 2001) >> >> Changes valueNowOrOnUnwindDo: to ensure:, plus incorporates some earlier fixes. >> >> VERSION 1.1a Â Â (May 2001) >> >> 1. Support for keeping track of multiple subexpressions. >> 2. Dot (.) matches anything but NUL character, as it should per POSIX spec. >> 3. Some bug fixes. >> >> >> >> 2009/11/4 StÃ©phane Ducasse <stephane.ducasse(a)inria.fr>: >>>> l := '1Ã¨re partie\nmilieu\nfin'. >>>> '.*' asRegex matchesIn l. >>> >>> -> infinite and freezing loop. >>> >>> Stef >>> >>> >>> >>> >>> _______________________________________________ >>> Pharo-project mailing list >>> Pharo-project(a)lists.gforge.inria.fr >>> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project >>> >> >> >> >> -- >> Lukas Renggli >> http://www.lukas-renggli.ch >> >> _______________________________________________ >> Pharo-project mailing list >> Pharo-project(a)lists.gforge.inria.fr >> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project >> > > _______________________________________________ > Pharo-project mailing list > Pharo-project(a)lists.gforge.inria.fr > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project > -- Lukas Renggli http://www.lukas-renggli.ch

Nov. 4, 2009

Re: [Pharo-project] Fwd: nice bug with regex

by Nicolas Cellier

This version does not seem to solve http://bugs.squeak.org/view.php?id=5391 { '15' matchesRegex: '[1-9]|1[0-9]'. '15' matchesRegex: '1[0-9]|[1-9]'. } #(false true) 2009/11/4 Lukas Renggli <renggli(a)gmail.com>: > I committed VB-Regex-lr.34 to the PharoInbox that fixes this issue. > > In fact, this is a completely new port of VB-Regex with numerous > bug-fixes and enhancements. The bug you found was fixed in version > 1.1c in December 2004. The version 1.1 we used up to now was from > October 1999. > > The detailed log of changes from Vassily you can see below. > > Cheers, > Lukas > > Name: VB-Regex-lr.34 > Author: lr > Time: 4 November 2009, 10:40:27 pm > UUID: 9254d5aa-3e22-4e3b-83aa-2d633ff05570 > Ancestors: VB-Regex-StephaneDucasse.33 > > VERSION 1.2.3 (November 2007) > > 1. Regexs with ^ or $ applied to copy empty strings caused infinite > loops, e.g. ('' copyWithRegex: '^.*$' matchesReplacedWith: 'foo'). > Applied a similar correction to that from version 1.1c, to > #copyStream:to:(replacingMatchesWith:|translatingMatchesUsing:). > 2. Extended RxParser testing to run each test for > #copy:translatingMatchesUsing: as well as #search:. > 3. Corrected #testSuite test that a dot does not match a null, which > was passing by luck with Smalltalk code in a literal array. > 4. Added test to end of test suite for fix 1 above. > > VERSION 1.2.2 (November 2006) > > There was no way to specify a backslash in a character set. Now [\\] > is accepted. > > VERSION 1.2.1 Â (August 2006) > > 1. Support for returning all ranges (startIndex to: stopIndex) > matching a regex - #allRangesOfRegexMatches:, #matchingRangesIn: > 2. Added hint to usage documentation on how to get more information > about matches when enumerating > 3. Syntax description of dot corrected: matches anything but NUL since 1.1a > > VERSION 1.2 Â Â (May 2006) > > Fixed case-insensitive search for character sets. > > VERSION 1.1c Â Â (December 2004) > > Fixed the issue with #matchesOnStream:do: which caused infinite loops > for matches > that matched empty strings. > > VERSION 1.1b Â Â (November 2001) > > Changes valueNowOrOnUnwindDo: to ensure:, plus incorporates some earlier fixes. > > VERSION 1.1a Â Â (May 2001) > > 1. Support for keeping track of multiple subexpressions. > 2. Dot (.) matches anything but NUL character, as it should per POSIX spec. > 3. Some bug fixes. > > > > 2009/11/4 StÃ©phane Ducasse <stephane.ducasse(a)inria.fr>: >>> l := '1Ã¨re partie\nmilieu\nfin'. >>> '.*' asRegex matchesIn l. >> >> -> infinite and freezing loop. >> >> Stef >> >> >> >> >> _______________________________________________ >> Pharo-project mailing list >> Pharo-project(a)lists.gforge.inria.fr >> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project >> > > > > -- > Lukas Renggli > http://www.lukas-renggli.ch > > _______________________________________________ > Pharo-project mailing list > Pharo-project(a)lists.gforge.inria.fr > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project >

Nov. 4, 2009

Re: [Pharo-project] Sockets, events, semaphores, etc.

by John M McIntosh

Well as the one responsible for the initial vm implementation, let me say: (a) There is fall over code for one semaphore versus three because the windows rewrite took a couple of years to complete, obviously some legacy code could disappear? I doubt anyone has tested the Socket smalltalk code against a 10 year old VM for example, so someone could drop the test and primitive method, etc. (b) The concept came about, via pacing on my boathouse deck and talking to John Maloney at the time was to have three semaphores, one for read, one for write, and one for control. The is/was a state diagram that showed how the three semaphores get signalled depending on the state of the read/write & control (open, closed, accept) transition based on unix BSD socket logic. The VM for unix/linux/os-x based machines technically polls the state of the file descriptors at least on the order of 1 to 3 ms and if the image isn't busy will sleep until the next known wakeup time to service a Delay or if a pending UI event or socket signal occurs.. I'd hunt in the squeak mailing list archives for it. (c) I think I wrote 2 Socket implementations, then it's evolved into what you have today. One of the challenges facing the implementer up in the smalltalk layer is to decide, when: a) Will a listen result in an accept? b) When can I write to a Socket? c) When is the Socket write completed? d) When can I read from a Socket? e) Is there more to read from a Socket? f) If I close a socket, when will the other party close there side? g) When the other party closes the socket when do I know that? In many of the cases waiting forever isn't a choice the developer wants to make, improper choices are subtle issues. For example you open a socket, write 64k of data, close the socket. How long do you wait? Oh yes on a dialup connection running at 32,000 baud why that wait time is much longer than in the lab with 10mb ethernet everywhere... Can I wait forever? Well no since we do at some point want to cleanup and let the other party deal with whatever happened. Can the socket change state between the point we get the first state change and then attempt to service it? Well of course... I believe the listen loop polling is there to deal with issues on some *cough* platforms where the socket might be busted because the listen/ accept handshake might not complete correctly and the socket goes weird, so the timeout/reset dance cleans up the mess, but no-one really determined what the issue was and likely it was platform version dependent anyway. I do recall some effort to implement some complicated schemes where pending queued connection requests could be queued up in the VM to work around problems with mechanics at the BSD socket level with the listen/accept protocols. Also see Craig's work on Flow On 2009-11-04, at 11:35 AM, Schwab,Wilhelm K wrote: > Hello all, > > Am I falling into a bad area of the image, or does Pharo poll for > socket connections? As an example, look at > ConnectionQueue>>getConnectionOrNil. Even the Socket waiting > methods are all time-bombed. I would expect these methods to wait > on a semaphore signalled by the vm on a socket event. It is one > thing to offer time-bombed options, but I see nothing event driven > at present :( Am I missing something? > > Bill > > > _______________________________________________ > Pharo-project mailing list > Pharo-project(a)lists.gforge.inria.fr > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project -- = = = ======================================================================== John M. McIntosh <johnmci(a)smalltalkconsulting.com> Twitter: squeaker68882 Corporate Smalltalk Consulting Ltd. http://www.smalltalkconsulting.com = = = ========================================================================

Nov. 4, 2009