Notepad++ v5.4.1 new features and fixed bugs (from v5.4) :
1. Add ChangeMarkers plugin in installer.
2. Add theme description in installer.
3. Make installer update contextMenu.xml (backup the old one then override it with the new one).
4. Fix lastRecentFileList localization issue.
5. Add readonly change status notification for plugin system.
6. Fix Styler Configurator user extension handler.
Issue: The option "Don't use %APPDATA%" does not work properly anymore in 5.4.1 (Unicode).
Platform: Vista 32.
Symptoms:
NP++ does not remember:
- tabs (opened files);
- settings changes;
- does not make recently opened file entries;
--> unless it is run as administrator (which is undesirable).
This option used to work fine for me until 5.3.x or at least 5.2.x.
The problem is solved by re-installing without using this option.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
1) The problem with the "Recent Files List" seems to be system-specific, or locale-specific. I can't reproduce it on different versions of Windows XP (tried SP2 and SP3, English and Russian).
2) Problems with searching for Cyrillic (in general, non-Latin) characters in UCS-2 or UTF-8 files, including the "Match case" problem, is caused by not-completely-patched version of SciLexer.dll in Notepad++ 5.4.1.
To solve it, replace the file "SciLexer.dll" with its previous version from Notepad++ 5.3.1. It's the only thing required.
However, both Notepad++ 5.3.1 and 5.4.1 have a problem with non-Latin text shown in Notepad++'es "Find results" pane - some characters of non-Latin text found in UCS-2 or UTF-8 files are shown incorrectly in the "Find results" pane. I'll try to provide a patch for this.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I can confirm only the 3rd bug: cyrillic text in UCS-2 and UTF-8 files is ALWAYS searched case-sensitivetly, ignoring the option "Match case". However, cyrillic text in ANSI files is correctly searched case-sensitivetly when "Match case" is off. So there seems to be some error with "Match case" when searching non-ANSI (i.e. UCS-2 or UTF-8) cyrillic (I think, non-Latin in general) text.
Regarding the 1st and 2nd problems mentioned, I think it can be caused by non-Russian (or non-Ukrainian, for example) locale (Control Panel -> Regional and Language Options).
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
There seems to be bug that persists in the latest version from previous ones. When you have two or more documents open, upon changing to one of them by clicking on its tab for this, it is opened in another instance of Notepad++. Anyone else experiencing this?
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I was able to reproduce this behavior only when a document was cloned; and then only sometimes. If anyone has found some sure fire steps to reproduce this behavior as well the the actions that cause it to suddenly work correctly please post!
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Regarding "Match case": I meant, cyrillic text in ANSI files is correctly searched case-insensitivetly when "Match case" is off. In non-ANSI file it is always searched case-sensitivetly, ignoring the "Match case" option.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
... ant the cause is inside Document::FindText(). One big part of this function has been removed for some reason, and it makes absolutely imposible to search for non-Latin text in UTF-8 case-insensitively.
The cause is in the following: non-Latin characters in UTF-8 are represented by 2 or 3 bytes.
Current implementation of Document::FindText() performs char by char (i.e. byte by byte) comparison which can not be applied to non-Latin characters case-insensitively becasue the same non-Latin character can be of 2 bytes in Upper Case and of 3 bytes in Lower Case or otherwise. Therefore we need some additional temporary buffer which will initially store the searched UTF-8 string in Upper Case, and then compare it character by character with characters form current document also converted to Upper Case. As an example of such approach, you can look at implementation of Document::FindText() in the sources of Scintilla used in Notepad++ 5.0.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
1/ The displayed selected character count is still wrong (3 lines by 3 columns can lead to 475!). Bug reported a while ago on SF tracker.
2/ Perform a rectangular selection using Alt+mouse, and then press Shift+Down arrow. The whole line below the selected rectangle shows as selected as well, and those that follow if pressing repeatedly are stream selected, in a strange mixed selection mode. And it is not just a display thing: hitting Del removes both the rectangle and the full line(s).
Let's call this a feature, it could be useful.
3/ Typing characters in sequence in a rectanular selection is a nice feature. But, on a mistake, backspace does not remove the last typed column: it jumps to the rightmost typed char on the first selected row and special typing mode is cancelled. This one I'd call a bug. Undo does the right thing though.
CChris
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Notepad++ v5.4.1 just came out :
https://sourceforge.net/project/showfiles.php?group_id=95717
Notepad++ v5.4.1 new features and fixed bugs (from v5.4) :
1. Add ChangeMarkers plugin in installer.
2. Add theme description in installer.
3. Make installer update contextMenu.xml (backup the old one then override it with the new one).
4. Fix lastRecentFileList localization issue.
5. Add readonly change status notification for plugin system.
6. Fix Styler Configurator user extension handler.
For the v5.4.x main feature please refer to :
https://sourceforge.net/forum/forum.php?thread_id=3283129&forum_id=331753
As usual, post any regression and critical bug you found here.
Don
Issue: The option "Don't use %APPDATA%" does not work properly anymore in 5.4.1 (Unicode).
Platform: Vista 32.
Symptoms:
NP++ does not remember:
- tabs (opened files);
- settings changes;
- does not make recently opened file entries;
--> unless it is run as administrator (which is undesirable).
This option used to work fine for me until 5.3.x or at least 5.2.x.
The problem is solved by re-installing without using this option.
Whoops. I just raised a bug for Styler Configurator user extension. Looks like we should close it .. :)
Thanks; it's closed now.
Thank you for release, but still 2 bugs
1 - My problem persists (Russian)
http://pic.ipicture.ru/uploads/090526/1v3Wt07pyJ.jpg
2 - Replace on: 20H D;03 C -ever this text
'2 - Replace on: 20H D;03 C -ever this text' -
You stated the same thing in the 5.4 release thread; I still don't see the 'bug' here. Perhaps you could explain the steps to re-create?
almostautomated
And last BUG
The option "Match case" dont work properly. (he always work)
Could anyone else confirm this? I've tried multiple times and it just keeps working the way it should!
Ctrl+M no longer works and there's no shortcut on the keyboard :(. i found that really useful.
i can push the dropdown menu key and select the highlight tho. but it would be nice if it was available as a shortcut.
prozaker,
This feature was removed in 5.4 in leui of the new 5 highlighters.
You can vote for a re-implementation here:
http://apps.sourceforge.net/ideatorrent/notepad-plus/ideatorrent/idea/7/
Sorry, the problem was nppcm.dll after the upgrade file is older version.
1 - But the problem with the "Recent Files List"
now it is constant
http://pic.ipicture.ru/uploads/090526/1v3Wt07pyJ.jpg
(me no need to do another shot with the same problem)
2 - Search UCS-2 on the Cyrillic is a work strange way
I Search:
Хан
He finds:
дан
Хао
etc
--------------------------------------
3 - The option "Match case" dont work properly. (he always work)
It's all!
OK, finally:
1) The problem with the "Recent Files List" seems to be system-specific, or locale-specific. I can't reproduce it on different versions of Windows XP (tried SP2 and SP3, English and Russian).
2) Problems with searching for Cyrillic (in general, non-Latin) characters in UCS-2 or UTF-8 files, including the "Match case" problem, is caused by not-completely-patched version of SciLexer.dll in Notepad++ 5.4.1.
To solve it, replace the file "SciLexer.dll" with its previous version from Notepad++ 5.3.1. It's the only thing required.
However, both Notepad++ 5.3.1 and 5.4.1 have a problem with non-Latin text shown in Notepad++'es "Find results" pane - some characters of non-Latin text found in UCS-2 or UTF-8 files are shown incorrectly in the "Find results" pane. I'll try to provide a patch for this.
I'll test it tomorrow
I can confirm only the 3rd bug: cyrillic text in UCS-2 and UTF-8 files is ALWAYS searched case-sensitivetly, ignoring the option "Match case". However, cyrillic text in ANSI files is correctly searched case-sensitivetly when "Match case" is off. So there seems to be some error with "Match case" when searching non-ANSI (i.e. UCS-2 or UTF-8) cyrillic (I think, non-Latin in general) text.
Regarding the 1st and 2nd problems mentioned, I think it can be caused by non-Russian (or non-Ukrainian, for example) locale (Control Panel -> Regional and Language Options).
There seems to be bug that persists in the latest version from previous ones. When you have two or more documents open, upon changing to one of them by clicking on its tab for this, it is opened in another instance of Notepad++. Anyone else experiencing this?
I was able to reproduce this behavior only when a document was cloned; and then only sometimes. If anyone has found some sure fire steps to reproduce this behavior as well the the actions that cause it to suddenly work correctly please post!
4. Fix lastRecentFileList localization issue.
Sorry,the bug disappeared again
http://img413.imageshack.us/img413/7440/bugqhd.jpg
Regarding "Match case": I meant, cyrillic text in ANSI files is correctly searched case-insensitivetly when "Match case" is off. In non-ANSI file it is always searched case-sensitivetly, ignoring the "Match case" option.
... ant the cause is inside Document::FindText(). One big part of this function has been removed for some reason, and it makes absolutely imposible to search for non-Latin text in UTF-8 case-insensitively.
The cause is in the following: non-Latin characters in UTF-8 are represented by 2 or 3 bytes.
Current implementation of Document::FindText() performs char by char (i.e. byte by byte) comparison which can not be applied to non-Latin characters case-insensitively becasue the same non-Latin character can be of 2 bytes in Upper Case and of 3 bytes in Lower Case or otherwise. Therefore we need some additional temporary buffer which will initially store the searched UTF-8 string in Upper Case, and then compare it character by character with characters form current document also converted to Upper Case. As an example of such approach, you can look at implementation of Document::FindText() in the sources of Scintilla used in Notepad++ 5.0.
http://pic.ipicture.ru/uploads/090529/r7gd73WS06.jpg
Russian version/ XP sp3
So! What you tell now? ))))
2 - http://pic.ipicture.ru/uploads/090529/DADWPbXiSD.jpg - O_o
Search UCS-2 on the Cyrillic is a work strange way
Search:
Хан
He finds:
дан
Хао
etc
3 - The option "Match case" in the menu Find dont work properly. (he always work)
This problem too UCS-2 and UTF-8
BUG № 2
UCS-2 and UTF-8 too
When will the auto-updater start finding this release ? (It's still at 5.3.1 now)
This is a great editor. I am missing proper syntax highlighting for FORTRAN 90 code. How can I do that? Any idea?
Thanks
Well, if you are asking for ANY IDEA, how about this one: create your own lexer plugin.
On retangular selections:
1/ The displayed selected character count is still wrong (3 lines by 3 columns can lead to 475!). Bug reported a while ago on SF tracker.
2/ Perform a rectangular selection using Alt+mouse, and then press Shift+Down arrow. The whole line below the selected rectangle shows as selected as well, and those that follow if pressing repeatedly are stream selected, in a strange mixed selection mode. And it is not just a display thing: hitting Del removes both the rectangle and the full line(s).
Let's call this a feature, it could be useful.
3/ Typing characters in sequence in a rectanular selection is a nice feature. But, on a mistake, backspace does not remove the last typed column: it jumps to the rightmost typed char on the first selected row and special typing mode is cancelled. This one I'd call a bug. Undo does the right thing though.
CChris