Loading ...
Sorry, an error occurred while loading the content.

Re: [Clip] Code page/character issues

Expand Messages
  • Axel Berger
    ... I second that. A general setting turning off all UTF capabilities and detection would be extremely nice to have. It will then be up to you to detect and
    Message 1 of 14 , Jan 11, 2013
    • 0 Attachment
      John Shotsky wrote:
      > What I am looking for is a way to intercept this problem
      > so it doesn't happen, or a modification to NoteTab, if necessary,
      > to make it work right.

      I second that. A general setting turning off all UTF capabilities and
      detection would be extremely nice to have. It will then be up to you to
      detect and deal with such characters, but for myself I've found that to
      be the easy part, stopping NoteTab from interfering is the hard bit.

      Axel
    • John Shotsky
      Here is another interesting thing about this problem. If I export the file which creates the condition I m discussing, and do not even open the file in
      Message 2 of 14 , Jan 11, 2013
      • 0 Attachment
        Here is another interesting thing about this problem.
        If I export the file which creates the condition I'm discussing, and do not even open the file in NoteTab, then copy the
        file to another folder, then open the copy, it reads fine. If I open the original, it has the problem. That is, the
        simple act of copying an original file to another folder using Windows 7 'fixes' the problem. That is so bizarre, but at
        least I can tell my users to copy it to a different work folder before opening it.

        Regards,
        John
        RecipeTools Web Site: http://recipetools.gotdns.com/


        -----Original Message-----
        From: ntb-clips@yahoogroups.com [mailto:ntb-clips@yahoogroups.com] On Behalf Of John Shotsky
        Sent: Friday, January 11, 2013 10:16
        To: ntb-clips@yahoogroups.com
        Subject: RE: [Clip] Code page/character issues

        As I stated, I can't ask my users to know how to do that. They want to open a file, edit it a little, save and run my
        clips on it. Once they open, edit and save, the damage is done. What I am looking for is a way to intercept this problem
        so it doesn't happen, or a modification to NoteTab, if necessary, to make it work right. As I said, if I export this
        unedited file to either ANSI or UTF-8, it displays correctly, so why doesn't it just open it in that format to start
        with, instead of making bogus characters that aren't correct in *ANY* situation?
        (Speķa Pīr�gi) http://www.saveur.com/article/Recipes/Speka-Piragi-Bacon-Turnovers

        Regards,
        John
        RecipeTools Web Site: <http://recipetools.gotdns.com/> http://recipetools.gotdns.com/

        From: ntb-clips@yahoogroups.com [mailto:ntb-clips@yahoogroups.com] On Behalf Of Axel Berger
        Sent: Friday, January 11, 2013 10:00
        To: ntb-clips@yahoogroups.com
        Subject: Re: [Clip] Code page/character issues


        John Shotsky wrote:
        > usually, a character preceded by an 'A' with a carat on top of it,
        > but there are also other variations.

        NoteTab can only work with one 8-bit character set at a time. It can
        read and use UTF-8 notation if all characters used can be mapped to one
        character page, it can't deal with more than 256 different letters.
        So the safe way to go is to load as raw UTF-8 and run a clip to convert
        to something compatible like HTML's ? notation.

        Axel



        [Non-text portions of this message have been removed]



        ------------------------------------

        Fookes Software: http://www.fookes.com/
        NoteTab website: http://www.notetab.com/
        NoteTab Discussion Lists: http://www.notetab.com/groups.php

        ***
        Yahoo! Groups Links
      • Axel Berger
        ... Almost certainly not. Your original has already been opened and probably saved by NT at least once before. NT saves info about recent files in Note*.fpr.
        Message 3 of 14 , Jan 11, 2013
        • 0 Attachment
          John Shotsky wrote:
          > If I open the original, it has the problem.

          Almost certainly not. Your "original" has already been opened and
          probably saved by NT at least once before. NT saves info about recent
          files in Note*.fpr. When I encounter intractable UTF problems, I rename
          the file and open it then, does the same trick.

          > at least I can tell my users to copy it to a different work folder
          > before opening it.

          If they do that before opening it for the first time, they'll achieve
          nothing.

          But you can include the technique in a clip: Save the file, Rename it
          through ^!Dos (wait a second for cacheing), and open the "new" one. Even
          better, if you can make your users open NT with the /RawUTF8 parameter,
          your problems ought to be solved.

          (I should have remembered. When introduced I refrained from using that,
          because I wanted to keep the way back to version 5.8 and 6.2 open. By
          now I feel quite safe in 7.x and ought to reconsider. Now, where and how
          to find each and every place where NT is evoked? Registry, Programs
          folder, Links, Commands in dozens of shell-like programs, ... -- looks a
          lot like a big fool's errand, and needs to be done on at least three
          machines: Eric, a command line parameter is good, a setting would be
          much much better.)

          Axel
        • John Shotsky
          When the text file is exported from Firefox, then opened in NoteTab, it has the problem I am discussing. Every time, for the same export. If I export from
          Message 4 of 14 , Jan 11, 2013
          • 0 Attachment
            When the text file is exported from Firefox, then opened in NoteTab, it has the problem I am discussing. Every time, for
            the same export.
            If I export from FireFox, then reboot my computer, and then open the file with NoteTab, it has the problem I am
            discussing.
            If I open the file from FireFox, then save it with another name, it still has the same problem.
            If I open the file from FireFox, copy the contents, then paste into a new file and save THAT, it still has the same
            problem.
            If I export from FireFox, then copy the file to another folder, and then open the copy in NoteTab, it does not have the
            problem I am discussing. The original still has the problem.
            If I export from FireFox, then rename the file in the original folder using Windows, then open the file in NoteTab, it
            does not have the problem I am discussing. There is no original at this point.
            I can show it with screenshots, but if I copy or rename the file, the problem vanishes. Win7-64.
            Regards,
            John
            RecipeTools Web Site: <http://recipetools.gotdns.com/> http://recipetools.gotdns.com/

            From: ntb-clips@yahoogroups.com [mailto:ntb-clips@yahoogroups.com] On Behalf Of Axel Berger
            Sent: Friday, January 11, 2013 13:47
            To: ntb-clips@yahoogroups.com
            Subject: Re: [Clip] Code page/character issues


            John Shotsky wrote:
            > If I open the original, it has the problem.

            Almost certainly not. Your "original" has already been opened and
            probably saved by NT at least once before. NT saves info about recent
            files in Note*.fpr. When I encounter intractable UTF problems, I rename
            the file and open it then, does the same trick.

            > at least I can tell my users to copy it to a different work folder
            > before opening it.

            If they do that before opening it for the first time, they'll achieve
            nothing.

            But you can include the technique in a clip: Save the file, Rename it
            through ^!Dos (wait a second for cacheing), and open the "new" one. Even
            better, if you can make your users open NT with the /RawUTF8 parameter,
            your problems ought to be solved.

            (I should have remembered. When introduced I refrained from using that,
            because I wanted to keep the way back to version 5.8 and 6.2 open. By
            now I feel quite safe in 7.x and ought to reconsider. Now, where and how
            to find each and every place where NT is evoked? Registry, Programs
            folder, Links, Commands in dozens of shell-like programs, ... -- looks a
            lot like a big fool's errand, and needs to be done on at least three
            machines: Eric, a command line parameter is good, a setting would be
            much much better.)

            Axel



            [Non-text portions of this message have been removed]
          • Axel Berger
            ... This is really, really weird. There is no explanation I can think of except that Windows itself is saving extra meta-information about files and NT can
            Message 5 of 14 , Jan 12, 2013
            • 0 Attachment
              John Shotsky wrote:
              > If I export from FireFox, then copy the file to another folder,
              > and then open the copy in NoteTab, it does not have the
              > problem I am discussing. The original still has the problem.
              > If I export from FireFox, then rename the file in the original
              > folder using Windows, then open the file in NoteTab, it
              > does not have the problem I am discussing.

              This is really, really weird. There is no explanation I can think of
              except that Windows itself is saving extra meta-information about files
              and NT can access that. You should do a file compare in hex
              (Totalcommander can do that). If they are identical, they ought to be
              after a simple copy and must be after rename, then it has to be external
              meta-information.

              Axel
            • John Shotsky
              Well, if it was mainstream, I wouldn t have brought it up here. At first, I thought maybe FireFox wasn t properly closing the file, which is why I rebooted
              Message 6 of 14 , Jan 12, 2013
              • 0 Attachment
                Well, if it was mainstream, I wouldn't have brought it up here. At first, I thought maybe FireFox wasn't properly
                closing the file, which is why I rebooted between, but that didn't help. I've inspected properties, but that didn't show
                anything different between the before and after. The byte count doesn't change. But somehow, NoteTab is seeing
                'something' that I can't find, which is why I asked here. I'll see if I can find a more robust file metadata analyzer.
                FYI � EditPad Pro opens it fine, but it is also a Unicode editor, and I often use it to see what is 'really' in a text
                file.

                Now for the even weirder part: EditPad Pro DOES display it correctly in text view, but in hex/Ascii view, it shows the
                same thing as NoteTab!
                Text View: Speka Piragi
                Hex/Ascii View: Spe��a P��r��gi

                NoteTab correctly detects it as utf-8. But when I force it to Windows 1252, it displays as in NoteTab � incorrectly.
                That means that NoteTab is detecting it as Win1252 before it is renamed, and as UTF8 afterwards. Further, that means
                that either Windows or NoteTab is incorrectly detecting the code page. But, since EditPad Pro detects it correctly, I
                don't think it's Windows. This reminds me that, over time, I have often had problems with characters in NoteTab, and
                have done everything I could think of to resolve those issues. But this one seems unresolvable. (except for moving the
                file)

                Recipes, it turns out, can be very difficult, since they come from every country, and even when Anglicized, they often
                retain oddball characters. Web pages can display these characters fine, but text editors may have problems. I can't even
                guess why NoteTab detects the code page differently, but I am pretty certain that it is a NoteTab problem alone. I do
                not know what mechanism editors use to detect code pages.

                Regards,
                John
                RecipeTools Web Site: <http://recipetools.gotdns.com/> http://recipetools.gotdns.com/

                From: ntb-clips@yahoogroups.com [mailto:ntb-clips@yahoogroups.com] On Behalf Of Axel Berger
                Sent: Saturday, January 12, 2013 00:39
                To: ntb-clips@yahoogroups.com
                Subject: Re: [Clip] Code page/character issues


                John Shotsky wrote:
                > If I export from FireFox, then copy the file to another folder,
                > and then open the copy in NoteTab, it does not have the
                > problem I am discussing. The original still has the problem.
                > If I export from FireFox, then rename the file in the original
                > folder using Windows, then open the file in NoteTab, it
                > does not have the problem I am discussing.

                This is really, really weird. There is no explanation I can think of
                except that Windows itself is saving extra meta-information about files
                and NT can access that. You should do a file compare in hex
                (Totalcommander can do that). If they are identical, they ought to be
                after a simple copy and must be after rename, then it has to be external
                meta-information.

                Axel



                [Non-text portions of this message have been removed]
              • John Shotsky
                One last bit of information about this problem: When the file has been moved or renamed, it displays correctly in NoteTab, although it doesn t display the
                Message 7 of 14 , Jan 12, 2013
                • 0 Attachment
                  One last bit of information about this problem:
                  When the file has been moved or renamed, it displays correctly in NoteTab, although it doesn't display the accents. It
                  displays standard English characters. However, when this same file is viewed in EditPad Pro, the original accents are
                  still present. It is only the DISPLAY that is different between these two instances. So, in one case NoteTab displays
                  one way, and in the other case a different way.

                  From what I've read about code page detection, there are commands that search through a document and report on
                  characters found. I still have no idea why moving or renaming a file provides different results to NoteTab. But it is
                  apparently something about this search or the decision about how to display that is not working correctly.

                  In my research, I found this fascinating tale of the history and basis of characters on computers:
                  http://www.joelonsoftware.com/printerFriendly/articles/Unicode.html

                  Regards,
                  John
                  RecipeTools Web Site: http://recipetools.gotdns.com/


                  -----Original Message-----
                  From: ntb-clips@yahoogroups.com [mailto:ntb-clips@yahoogroups.com] On Behalf Of John Shotsky
                  Sent: Saturday, January 12, 2013 05:06
                  To: ntb-clips@yahoogroups.com
                  Subject: RE: [Clip] Code page/character issues

                  Well, if it was mainstream, I wouldn't have brought it up here. At first, I thought maybe FireFox wasn't properly
                  closing the file, which is why I rebooted between, but that didn't help. I've inspected properties, but that didn't show
                  anything different between the before and after. The byte count doesn't change. But somehow, NoteTab is seeing
                  'something' that I can't find, which is why I asked here. I'll see if I can find a more robust file metadata analyzer.
                  FYI – EditPad Pro opens it fine, but it is also a Unicode editor, and I often use it to see what is 'really' in a text
                  file.

                  Now for the even weirder part: EditPad Pro DOES display it correctly in text view, but in hex/Ascii view, it shows the
                  same thing as NoteTab!
                  Text View: Speka Piragi
                  Hex/Ascii View: Speķa Pīr�gi

                  NoteTab correctly detects it as utf-8. But when I force it to Windows 1252, it displays as in NoteTab – incorrectly.
                  That means that NoteTab is detecting it as Win1252 before it is renamed, and as UTF8 afterwards. Further, that means
                  that either Windows or NoteTab is incorrectly detecting the code page. But, since EditPad Pro detects it correctly, I
                  don't think it's Windows. This reminds me that, over time, I have often had problems with characters in NoteTab, and
                  have done everything I could think of to resolve those issues. But this one seems unresolvable. (except for moving the
                  file)

                  Recipes, it turns out, can be very difficult, since they come from every country, and even when Anglicized, they often
                  retain oddball characters. Web pages can display these characters fine, but text editors may have problems. I can't even
                  guess why NoteTab detects the code page differently, but I am pretty certain that it is a NoteTab problem alone. I do
                  not know what mechanism editors use to detect code pages.

                  Regards,
                  John
                  RecipeTools Web Site: <http://recipetools.gotdns.com/> http://recipetools.gotdns.com/

                  From: ntb-clips@yahoogroups.com [mailto:ntb-clips@yahoogroups.com] On Behalf Of Axel Berger
                  Sent: Saturday, January 12, 2013 00:39
                  To: ntb-clips@yahoogroups.com
                  Subject: Re: [Clip] Code page/character issues


                  John Shotsky wrote:
                  > If I export from FireFox, then copy the file to another folder,
                  > and then open the copy in NoteTab, it does not have the
                  > problem I am discussing. The original still has the problem.
                  > If I export from FireFox, then rename the file in the original
                  > folder using Windows, then open the file in NoteTab, it
                  > does not have the problem I am discussing.

                  This is really, really weird. There is no explanation I can think of
                  except that Windows itself is saving extra meta-information about files
                  and NT can access that. You should do a file compare in hex
                  (Totalcommander can do that). If they are identical, they ought to be
                  after a simple copy and must be after rename, then it has to be external
                  meta-information.

                  Axel



                  [Non-text portions of this message have been removed]



                  ------------------------------------

                  Fookes Software: http://www.fookes.com/
                  NoteTab website: http://www.notetab.com/
                  NoteTab Discussion Lists: http://www.notetab.com/groups.php

                  ***
                  Yahoo! Groups Links
                • John Wallace
                  What would happen if you saved it as a .html file instead of .txt file? ... From: ntb-clips@yahoogroups.com [mailto:ntb-clips@yahoogroups.com] On Behalf Of
                  Message 8 of 14 , Jan 12, 2013
                  • 0 Attachment
                    What would happen if you saved it as a .html file instead of .txt file?


                    -----Original Message-----
                    From: ntb-clips@yahoogroups.com [mailto:ntb-clips@yahoogroups.com] On Behalf Of John Shotsky
                    Sent: Saturday, January 12, 2013 05:06
                    To: ntb-clips@yahoogroups.com
                    Subject: RE: [Clip] Code page/character issues

                    Well, if it was mainstream, I wouldn't have brought it up here. At first, I thought maybe FireFox wasn't properly closing the file,
                    which is why I rebooted between, but that didn't help. I've inspected properties, but that didn't show anything different between
                    the before and after. The byte count doesn't change. But somehow, NoteTab is seeing 'something' that I can't find, which is why I
                    asked here. I'll see if I can find a more robust file metadata analyzer.
                    FYI – EditPad Pro opens it fine, but it is also a Unicode editor, and I often use it to see what is 'really' in a text file.

                    Now for the even weirder part: EditPad Pro DOES display it correctly in text view, but in hex/Ascii view, it shows the same thing as
                    NoteTab!
                    Text View: Speka Piragi
                    Hex/Ascii View: Speķa Pīrāgi

                    NoteTab correctly detects it as utf-8. But when I force it to Windows 1252, it displays as in NoteTab – incorrectly.
                    That means that NoteTab is detecting it as Win1252 before it is renamed, and as UTF8 afterwards. Further, that means that either
                    Windows or NoteTab is incorrectly detecting the code page. But, since EditPad Pro detects it correctly, I don't think it's Windows.
                    This reminds me that, over time, I have often had problems with characters in NoteTab, and have done everything I could think of to
                    resolve those issues. But this one seems unresolvable. (except for moving the
                    file)

                    Recipes, it turns out, can be very difficult, since they come from every country, and even when Anglicized, they often retain
                    oddball characters. Web pages can display these characters fine, but text editors may have problems. I can't even guess why NoteTab
                    detects the code page differently, but I am pretty certain that it is a NoteTab problem alone. I do not know what mechanism editors
                    use to detect code pages.

                    Regards,
                    John
                    RecipeTools Web Site: <http://recipetools.gotdns.com/> http://recipetools.gotdns.com/

                    From: ntb-clips@yahoogroups.com [mailto:ntb-clips@yahoogroups.com] On Behalf Of Axel Berger
                    Sent: Saturday, January 12, 2013 00:39
                    To: ntb-clips@yahoogroups.com
                    Subject: Re: [Clip] Code page/character issues


                    John Shotsky wrote:
                    > If I export from FireFox, then copy the file to another folder, and
                    > then open the copy in NoteTab, it does not have the problem I am
                    > discussing. The original still has the problem.
                    > If I export from FireFox, then rename the file in the original folder
                    > using Windows, then open the file in NoteTab, it does not have the
                    > problem I am discussing.

                    This is really, really weird. There is no explanation I can think of except that Windows itself is saving extra meta-information
                    about files and NT can access that. You should do a file compare in hex (Totalcommander can do that). If they are identical, they
                    ought to be after a simple copy and must be after rename, then it has to be external meta-information.

                    Axel
                  • Axel Berger
                    ... It has to, those characters are not in CP1252. Converting your sample and assuming mail transfer has not broken anything I get: Speķa Pīrāgi These are
                    Message 9 of 14 , Jan 12, 2013
                    • 0 Attachment
                      John Shotsky wrote:
                      > Text View: Speka Piragi
                      > Hex/Ascii View: Speķa Pīrāgi
                      > NoteTab correctly detects it as utf-8. But when I force it to
                      > Windows 1252, it displays as in NoteTab – incorrectly.

                      It has to, those characters are not in CP1252. Converting your sample
                      and assuming mail transfer has not broken anything I get:

                      Speķa Pīrāgi

                      These are from the "extended block A"
                      http://www.sql-und-xml.de/unicode-database/latin-extended-a.html

                      NoteTab will never be able to deal with them satisfactorily. What I
                      don't get at all is how Win7 interferes with them, but then I have so
                      far refrained from using eXPerimental and stick to Win98. Even that
                      tries to interfere and impose its preferences over mine, but there I can
                      more or less control it. Your identical byte count might result from
                      using UTF-16, don't newer Windoses do that? If so the byte count should
                      be twice the letter count.

                      > But, since EditPad Pro detects it correctly, I
                      > don't think it's Windows.

                      If editpad is true UTF, as you say, then it need not detect anything.
                      Notetab is stricly 8-bit and strictly codepage based, all it can do is
                      read letters from inside that single chosen codepage when encoded as
                      UTF-8. Letters from more than one codepage inside the same document will
                      never work.

                      Axel
                    • John Shotsky
                      EditPad Pro is a Unicode editor, so yes, it displays Unicode and utf-8 and many other code pages correctly. But that file is not Unicode, it is 8-bit UTF. When
                      Message 10 of 14 , Jan 12, 2013
                      • 0 Attachment
                        EditPad Pro is a Unicode editor, so yes, it displays Unicode and utf-8 and many other code pages correctly. But that
                        file is not Unicode, it is 8-bit UTF. When one of these files is moved, NoteTab not only displays it correctly, but it
                        also saves it correctly, that is, without the accents. So, that is the workaround for now. What is not acceptable is the
                        file as first opened, which does not result in a question mark or any valid character in any code page. It is just
                        garbage. Previously, NoteTab displayed a question mark for any character out of its map. Now, it doesn't.

                        But that's not actually the point anyway. The file is UTF-8 when it is written, and after it is copied. Nothing is
                        different about the file except that there is a copy in another location. The copy displays correctly in NoteTab, but
                        the original doesn't. The copy works with my clip library, the original doesn't. If I export the original in NoteTab to
                        UTF-8 it displays correctly, but of course just copying it works, as does renaming it, so I can't say the export
                        actually does anything. However, if I export it to Ascii, question marks show up for those characters, as expected. The
                        clip library can't work with a bunch of question marks either, of course, as there is no way to guess what the missing
                        character is except through a very, very complex word map which replaces question marks with characters if the word is
                        otherwise recognized. So, for the words you correctly detected below, I would simply substitute the unaccented
                        characters for accented ones and that would be fine. But I can't do that with the original, because it displays EXTRA
                        characters, as indicated in my 'Hex/Ascii' view below.

                        So, for now, my instructions will include moving the FireFox-exported file to a work folder, and we'll go with that as
                        long as it continues to work. As to the problem, I will leave it in the category of unresolvable.

                        Regards,
                        John
                        RecipeTools Web Site: <http://recipetools.gotdns.com/> http://recipetools.gotdns.com/

                        From: ntb-clips@yahoogroups.com [mailto:ntb-clips@yahoogroups.com] On Behalf Of Axel Berger
                        Sent: Saturday, January 12, 2013 07:23
                        To: ntb-clips@yahoogroups.com
                        Subject: Re: [Clip] Code page/character issues


                        John Shotsky wrote:
                        > Text View: Speka Piragi
                        > Hex/Ascii View: Spe��a P��r��gi
                        > NoteTab correctly detects it as utf-8. But when I force it to
                        > Windows 1252, it displays as in NoteTab � incorrectly.

                        It has to, those characters are not in CP1252. Converting your sample
                        and assuming mail transfer has not broken anything I get:

                        Speka Piragi

                        These are from the "extended block A"
                        http://www.sql-und-xml.de/unicode-database/latin-extended-a.html

                        NoteTab will never be able to deal with them satisfactorily. What I
                        don't get at all is how Win7 interferes with them, but then I have so
                        far refrained from using eXPerimental and stick to Win98. Even that
                        tries to interfere and impose its preferences over mine, but there I can
                        more or less control it. Your identical byte count might result from
                        using UTF-16, don't newer Windoses do that? If so the byte count should
                        be twice the letter count.

                        > But, since EditPad Pro detects it correctly, I
                        > don't think it's Windows.

                        If editpad is true UTF, as you say, then it need not detect anything.
                        Notetab is stricly 8-bit and strictly codepage based, all it can do is
                        read letters from inside that single chosen codepage when encoded as
                        UTF-8. Letters from more than one codepage inside the same document will
                        never work.

                        Axel



                        [Non-text portions of this message have been removed]
                      • Axel Berger
                        ... To my understanding UTF-8 as a specific encoding is a subset, or rather one of several possible versions, of Unicode. ... Sorry, but if those letters do
                        Message 11 of 14 , Jan 12, 2013
                        • 0 Attachment
                          John Shotsky wrote:
                          > But that file is not Unicode, it is 8-bit UTF.

                          To my understanding UTF-8 as a specific encoding is a subset, or rather
                          one of several possible versions, of Unicode.

                          > When one of these files is moved, NoteTab not only displays it
                          > correctly, but it also saves it correctly, that is, without the
                          > accents.

                          Sorry, but if those letters do have accents, then anything without is
                          INcorrect. It may be an acceptable workaround, like Muller or Mueller
                          instead of Müller, but never correct.

                          > So, that is the workaround for now.

                          Right

                          > But that's not actually the point anyway.

                          Agreed. Win7 does something strange here and I'm very happy I need not
                          concern myself with that.

                          > As to the problem, I will leave it in the category of unresolvable.

                          Probably best.
                        Your message has been successfully submitted and would be delivered to recipients shortly.