March 2022 - Pharo-users - lists.pharo.org

Re: Illegal Leading Byte

by Sven Van Caekenberghe

Hi Craig, > On 3 Mar 2022, at 23:15, craig <craig(a)hivemind.net> wrote: > > Hi Guys, > > > I'm reading a text file which is supposed to be ASCII encoded. This file contains a list of filepaths and was created by a Python program. > > > Well, it turns-out that file names on Windows can contain illegal UTF-8 characters. This causes ZnUTF8Encoder to signal 'Illegal leading byte for utf-8 encoding' and crash the program. > > > I would like to handle this situation more elegantly, is there a more appropriate code-page to use for the Windows filesystem? > > <4b5aa143.png> > > > > Craig We support more than 80 different character encoders. Of course, you should first know what encoder is being used, after that, it is easy to use a different encoder. Consider: '/tmp/foo.txt' asFileReference readStreamEncoded: #utf8 do: [ :in | in upToEnd ]. '/tmp/foo.txt' asFileReference readStreamEncoded: #windows1252 do: [ :in | in upToEnd ]. '/tmp/foo.txt' asFileReference readStreamEncoded: #latin1 do: [ :in | in upToEnd ]. ZnCharacterEncoder knownEncodingIdentifiers. #windows1252 asZnCharacterEncoder. If you could post a small example of your file, I could try to help. It will probably be #windows1252 or #latin1. HTH, Sven

March 4, 2022

Illegal Leading Byte

by craig

Hi Guys, I'm reading a text file which is supposed to be ASCII encoded. This file contains a list of filepaths and was created by a Python program. Well, it turns-out that file names on Windows can contain illegal UTF-8 characters. This causes ZnUTF8Encoder to signal 'Illegal leading byte for utf-8 encoding' and crash the program. I would like to handle this situation more elegantly, is there a more appropriate code-page to use for the Windows filesystem? Craig

March 3, 2022

Re: The orange 'content modified' indicator

by Esteban Lorenzano

Re-hi :) Not because something was used before is good :) As I said, spec does not handles the concept of "dirty". Also, it does not has the concept of "accept" (accept what? what is dirty? within what context?). Knowing that is something that depends on your application. Editors, for example tend to put an asterisk in the name (and maybe a "mod" string in the status bar)... this is for example more friendly with colorblind people. I do not doubt the triangle was useful for you, but this is a punctual solution for a punctual problem, not a generic mechanism. Still as I said, you can easily implement something that would match your needs within your application. We do not have an "add triangle" style, but you can add/remove a style changing the border, you can hide/show an icon, whatever you want. There was someone in the community doing his own version of "controls with model" with spec (I do not remember who). I guess you need something like that, where you can keep the status of your model and know if something is dirty and how to react to that. Not hard to implement in a reusable way. Esteban On Mar 3 2022, at 4:59 pm, Robert Briggs via Pharo-users <pharo-users(a)lists.pharo.org> wrote: > Hi Esteban > > > Thankyou for responding. > > You say itâs a very bad way to show something is âdirtyâ yet Pharo continues to use it in the System Browser and elsewhere. Also from my researching It was used in Spec but has been deprecated in Spec2. > > I think It is also used in VisualWorks and earlier SmallTalk systems. > > Personally I find it very useful when editing code across a number of different objects as it reminds me that Iâve not saved the edits (which has caught me out when subsequently running a test for example. > > What is considered to be a better way to show that something is âdirtyâ, and is anything planned for Spec2 in the future in this respect? > > Regards > R > From: Esteban Lorenzano <estebanlm(a)netc.eu> > Reply to: Any question about pharo is welcome <pharo-users(a)lists.pharo.org> > Date: Thursday, 3 March 2022 at 15:16 > To: Any question about pharo is welcome <pharo-users(a)lists.pharo.org> > Cc: Robert Briggs <rbriggs2009(a)me.com> > Subject: [Pharo-users] Re: The orange 'content modified' indicator > > > > > Hi, > > > > No, this is not possible. > Main reason is because that little triangle is a very bad way to show something is "dirty", but also because Spec itself does not has the concept of "dirty". If you want such mechanism you need to implement it yourself (you can listen when content changes and add/remove a style, for example). > > > > Esteban > > > On Mar 3 2022, at 3:28 pm, Robert Briggs via Pharo-users <pharo-users(a)lists.pharo.org> wrote: > > Hi > > > > > > > > > > Pharo browserâs show a little triangular orange triangle in the top right corner when code is edited by the user but not yet Accepted. > > > > > > > > Is there a mechanism for doing the same, e.g. in an editable SpTextPresenter? > > > > > > > > Iâve looked around related Spec2 Classes but unable to find a method to do this. > > > > > > > > Advice please. > > > > > > > > Regards > > > > > > R > > > >

March 3, 2022

Re: The orange 'content modified' indicator

by Kasper Ãsterbye

Agrees with Robert > Den 3. mar. 2022 kl. 16.59 skrev Robert Briggs via Pharo-users <pharo-users(a)lists.pharo.org>: > > ï»¿ > Hi Esteban > > Thankyou for responding. > > You say itâs a very bad way to show something is âdirtyâ yet Pharo continues to use it in the System Browser and elsewhere. Also from my researching It was used in Spec but has been deprecated in Spec2. > > I think It is also used in VisualWorks and earlier SmallTalk systems. > > Personally I find it very useful when editing code across a number of different objects as it reminds me that Iâve not saved the edits (which has caught me out when subsequently running a test for example. > > What is considered to be a better way to show that something is âdirtyâ, and is anything planned for Spec2 in the future in this respect? > > Regards > R > From: Esteban Lorenzano <estebanlm(a)netc.eu> > Reply to: Any question about pharo is welcome <pharo-users(a)lists.pharo.org> > Date: Thursday, 3 March 2022 at 15:16 > To: Any question about pharo is welcome <pharo-users(a)lists.pharo.org> > Cc: Robert Briggs <rbriggs2009(a)me.com> > Subject: [Pharo-users] Re: The orange 'content modified' indicator > > Hi, > > No, this is not possible. > Main reason is because that little triangle is a very bad way to show something is "dirty", but also because Spec itself does not has the concept of "dirty". If you want such mechanism you need to implement it yourself (you can listen when content changes and add/remove a style, for example). > > Esteban > > On Mar 3 2022, at 3:28 pm, Robert Briggs via Pharo-users <pharo-users(a)lists.pharo.org> wrote: > Hi > > > > Pharo browserâs show a little triangular orange triangle in the top right corner when code is edited by the user but not yet Accepted. > > > > Is there a mechanism for doing the same, e.g. in an editable SpTextPresenter? > > > > Iâve looked around related Spec2 Classes but unable to find a method to do this. > > > > Advice please. > > > > Regards > > R > >

March 3, 2022

Re: The orange 'content modified' indicator

by Robert Briggs

Hi Esteban Thankyou for responding. You say itâs a very bad way to show something is âdirtyâ yet Pharo continues to use it in the System Browser and elsewhere.Â Also from my researching It was used in Spec but has been deprecated in Spec2. I think It is also used in VisualWorks and earlier SmallTalk systems. Personally I find it very useful when editing code across a number of different objects as it reminds me that Iâve not saved the edits (which has caught me out when subsequently running a test for example.Â What is considered to be a better way to show that something is âdirtyâ, and is anything planned for Spec2 in the future in this respect? Regards R From: Esteban Lorenzano <estebanlm(a)netc.eu> Reply to: Any question about pharo is welcome <pharo-users(a)lists.pharo.org> Date: Thursday, 3 March 2022 at 15:16 To: Any question about pharo is welcome <pharo-users(a)lists.pharo.org> Cc: Robert Briggs <rbriggs2009(a)me.com> Subject: [Pharo-users] Re: The orange 'content modified' indicator Hi, No, this is not possible. Main reason is because that little triangle is a very bad way to show something is "dirty", but also because Spec itself does not has the concept of "dirty". If you want such mechanism you need to implement it yourself (you can listen when content changes and add/remove a style, for example). Esteban On Mar 3 2022, at 3:28 pm, Robert Briggs via Pharo-users <pharo-users(a)lists.pharo.org> wrote: Hi Pharo browserâs show a little triangular orange triangle in the top right corner when code is edited by the user but not yet Accepted. Is there a mechanism for doing the same, e.g. in an editable SpTextPresenter? Iâve looked around related Spec2 Classes but unable to find a method to do this. Advice please. Regards R

March 3, 2022

Re: The orange 'content modified' indicator

by Esteban Lorenzano

Hi, No, this is not possible. Main reason is because that little triangle is a very bad way to show something is "dirty", but also because Spec itself does not has the concept of "dirty". If you want such mechanism you need to implement it yourself (you can listen when content changes and add/remove a style, for example). Esteban On Mar 3 2022, at 3:28 pm, Robert Briggs via Pharo-users <pharo-users(a)lists.pharo.org> wrote: > Hi > > > Pharo browserâs show a little triangular orange triangle in the top right corner when code is edited by the user but not yet Accepted. > > Is there a mechanism for doing the same, e.g. in an editable SpTextPresenter? > > Iâve looked around related Spec2 Classes but unable to find a method to do this. > > Advice please. > > Regards > R

March 3, 2022

The orange 'content modified' indicator

by Robert Briggs

Hi Pharo browserâs show a little triangular orange triangle in the top right corner when code is edited by the user but not yet Accepted. Is there a mechanism for doing the same, e.g. in an editable SpTextPresenter? Iâve looked around related Spec2 Classes but unable to find a method to do this. Advice please. Regards R

March 3, 2022

Dragging objects with restrictions

by Eduardo Ochs

Hi people, I am a beginner with SmallTalks. At the moment I understand the basics about the language but very little about the graphical interface... Can you recommend me docs and tutorials on how to do something like this in Pharo? http://angg.twu.net/eev-videos/2022dragABC.mp4 The link above is to a video that is 20 seconds long. In the first 15 seconds (suggestion: ignore them!) I use weird Emacs tricks to draw three objects in a Tcl/Tk canvas - the letters A, B, and C - and to bind "mousebutton1 motion" to different actions in each of them... ...and in the last 5 seconds of the video I draw the letters A, B and C with the mouse to show what are these actions. In short: the letter A moves freely, the letter B moves in the same horizontal line as A, and when I drag C around it moves B horizontally to force B to be in the same horizontal line as C. Thanks in advance! =) Eduardo Ochs http://angg.twu.net/#eev

March 2, 2022

Re: ByteString asDate

by Robert Briggs

Thanks Bernardo. Seems your suggestion creates a Protected Block which I understand.Â I will read the document some more a necessary.Â Thank you for taking time to respond. From: Bernardo Ezequiel Contreras <vonbecmann(a)gmail.com> Reply to: Any question about pharo is welcome <pharo-users(a)lists.pharo.org> Date: Wednesday, 2 March 2022 at 14:19 To: Any question about pharo is welcome <pharo-users(a)lists.pharo.org> Subject: [Pharo-users] Re: ByteString asDate see chapter 13 in http://files.pharo.org/books-pdfs/deep-into-pharo/2013-DeepIntoPharo-EN.pdf it's a chapter dedicated to exceptions. the #on:do: message is in section 13.4. regarding the asDate message, i really don't know. you need a kind of validator. sorry, if i'm not much help. On Wed, Mar 2, 2022 at 7:35 AM Robert Briggs via Pharo-users <pharo-users(a)lists.pharo.org> wrote: Thanks Bernardo That construct works as required, however Iâm not sure what the semantic difference is to my original. I seems counterintuitive. Could you expand on what is going on for my benefit (if you have time). Again thanks. Also do you have a view on my description of the asDate result, and whether it should self-validate the receiver and raise an exception? Regards R From: Bernardo Ezequiel Contreras <vonbecmann(a)gmail.com> Reply to: Any question about pharo is welcome <pharo-users(a)lists.pharo.org> Date: Tuesday, 1 March 2022 at 15:05 To: Any question about pharo is welcome <pharo-users(a)lists.pharo.org> Subject: [Pharo-users] Re: ByteString asDate try to send the #on:do: message to a block. try this ^ [ (Date readFrom: self pattern: âdd/mm/yyyyâ). true ] on: DateError do: [ false]. On Tue, Mar 1, 2022 at 11:52 AM Robert Briggs via Pharo-users <pharo-users(a)lists.pharo.org> wrote: I may be being stupid but the current Date Class does not appear to protect itself against a ByteString that is not in valid date format, e.g. by raising an Error. e.g. the code âABCâ asDate opens the Debugger on #isLetter was sent to nil because it appears to expect more letters after $C but the peek returns nil instead. This seems to me to require a more defensive programming approach. I would like to validate a user input as a valid date in a number of different formats before converting it to a Date and storing it in my model. Iâve tried using the Date parser, e.g. Date readFrom: '04/02/2013' readStream pattern: 'dd/mm/yyyy'. This is fine for simply parsing a valid string but rates a DateError if it doesnât match the . So entering â04/13/2013â cleverly raises âDateError: There is no 13th monthâ. My thought was I could use this to return a simple true/false result in a ByteString extension in my application for example as follows: ByteString >> isValdiDate (Date readFrom: self pattern: âdd/mm/yyyyâ) on: DateError do: [^ false]. ^ true However this doesnât intercept the DateError which is simply reported in the Debugger as before. This may be my lack of uncderstaning of the Exception handling in Pharo. Any and all thoughts welcome. R -- Bernardo E.C. Sent from a cheap desktop computer in South America. -- Bernardo E.C. Sent from a cheap desktop computer in South America.

March 2, 2022

Re: ByteString asDate

by Bernardo Ezequiel Contreras

see chapter 13 in http://files.pharo.org/books-pdfs/deep-into-pharo/2013-DeepIntoPharo-EN.pdf it's a chapter dedicated to exceptions. the #on:do: message is in section 13.4. regarding the asDate message, i really don't know. you need a kind of validator. sorry, if i'm not much help. On Wed, Mar 2, 2022 at 7:35 AM Robert Briggs via Pharo-users < pharo-users(a)lists.pharo.org> wrote: > Thanks Bernardo > > > > That construct works as required, however Iâm not sure what the semantic > difference is to my original. I seems counterintuitive. Could you expand > on what is going on for my benefit (if you have time). > > Again thanks. > > > > Also do you have a view on my description of the asDate result, and > whether it should self-validate the receiver and raise an exception? > > > > Regards > > R > > > > *From: *Bernardo Ezequiel Contreras <vonbecmann(a)gmail.com> > *Reply to: *Any question about pharo is welcome < > pharo-users(a)lists.pharo.org> > *Date: *Tuesday, 1 March 2022 at 15:05 > *To: *Any question about pharo is welcome <pharo-users(a)lists.pharo.org> > *Subject: *[Pharo-users] Re: ByteString asDate > > > > try to send the #on:do: message to a block. try this > > > > ^ [ (Date readFrom: self pattern: âdd/mm/yyyyâ). > > true ] on: DateError do: [ false]. > > > > > > > > On Tue, Mar 1, 2022 at 11:52 AM Robert Briggs via Pharo-users < > pharo-users(a)lists.pharo.org> wrote: > > I may be being stupid but the current Date Class does not appear to > protect itself against a ByteString that is not in valid date format, e.g. > by raising an Error. > > > > e.g. the code âABCâ asDate opens the Debugger on #isLetter was sent to > nil because it appears to expect more letters after $C but the peek > returns nil instead. > > This seems to me to require a more defensive programming approach. > > > > I would like to validate a user input as a valid date in a number of > different formats before converting it to a Date and storing it in my model. > > > > Iâve tried using the Date parser, e.g. > > > > Date readFrom: '04/02/2013' readStream pattern: 'dd/mm/yyyy'. > > > > > > This is fine for simply parsing a valid string but rates a DateError if it > doesnât match the . So entering â04/13/2013â cleverly raises âDateError: > There is no 13th monthâ. > > > > My thought was I could use this to return a simple true/false result in a > ByteString extension in my application for example as follows: > > > > ByteString >> isValdiDate > > > > (Date readFrom: self pattern: âdd/mm/yyyyâ) on: DateError do: [^ false]. > > ^ true > > > > However this doesnât intercept the DateError which is simply reported in > the Debugger as before. This may be my lack of uncderstaning of the > Exception handling in Pharo. > > > > Any and all thoughts welcome. > > > > R > > > > > > > > > > > > > > > -- > > Bernardo E.C. > > > > Sent from a cheap desktop computer in South America. > -- Bernardo E.C. Sent from a cheap desktop computer in South America.

March 2, 2022