• Hi, @zeff-wheelock and @scott-sumner,

    Scott, your regexes are exact, of course ! In addition, searching for the last occurrence makes sense, as we want to know the final state of all these servers.

    However, I must miss something ! To my mind, we should search for the last line, with the word connected, for each server, ONLY IF a previous line, with the word disconnected, for the according server, has not been found !

    For instance, if we consider the part of the Zeff's log, below :

    05/23/2017 05:04:09 AM SMTP Server: qwerty321f.domain.com (xxx.xxx.xxx.xxx) connected 05/23/2017 05:04:10 AM Router: Transferred 1 messages to SERVER004/DOMAIN via Notes 05/23/2017 05:04:10 AM SMTP Server: Message 0031D206 (…) received 05/23/2017 05:04:10 AM SMTP Server: qwerty321f.domain.com (xxx.xxx.xxx.xxx) disconnected. 1 message[s] received

    To my mind, it looks like, on 05/23/2017 05:04:10 AM, the server "qwerty321f" is disconnected. No ? Of course, I suppose that the string xxx.xxx.xxx.xxx, in the first and last line, just above, represents the same IPV4 address !



  • @guy038 said:

    In addition, it’s rather annoying because you cannot determine, easily, how many files are replaced and how many are not !

    I have always used the tab colors for this. Prior to a change/replace on all opened files I do Ctrl-Shift-S to do a save-all. That both makes sure I have a backup of the files available should I make a mistake with that mass search/replace and it changes the colors all of the tabs to be blue. I then do the search-replace using replace-all-in-all-opened-documents. The tabs of the files that were changed turn red.

    Granted - that does not give you a count of the number of files that were changed. I suspect it would be easy to add the count to the status line message though adding a count of the number of files not changed seems like overkill. For example “Replaced 400 occurrences in 304 files. (Search text not found in 196 open files)”

    Related to this is I’d also like to see “Close all that are not dirty” added. Sometimes I have hundreds of files open and use replace-all-in-all-opened-documents. While it’s easy to scroll through the document list using Ctrl-PageDown or Ctrl-PageUp it’s hard to spot the blue/red tabs when scrolling rapidly. Thus I’ve wished for at times a “Close all that are not dirty” feature that would close all files that have not been changed leaving me with just the red tabs. I could then inspect those, using Ctrl-Z (undo), F3 a bunch of times to view the results of the search-text, and Ctrl-Y (redo) as needed to inspect the changes in that file. For now what I’ve been doing is to go to a clean (blue) tab and to hold Ctrl-W (close) down letting the keyboard auto-repeat. That will close long strings of blue tabs. When it hits a red tab the save yes/no? warning pops up. I let go of the Ctrl-W, hit ESC to cancel the warning, and Ctrl-PageDown to the next blue tab. That works but is painful when it’s a random mix of clean/dirty files.

  • A script could look like this, so if you are interested in
    a python script solution let me know and we can work it out.

    # -*- coding: utf-8 -*- import os FIND_WHAT = u'FIND_WHAT_REGULAR_EXPRESSION' REPLACE_WITH = u'REPLACE_WITH_EXPRESSION' DIRECTORY_OF_FILES = u'THE_DIRECTORY' # like c:\\temp\\directory - doublebackslash needed os.chdir(DIRECTORY_OF_FILES) list_of_files = [x for x in os.listdir('.') if x.endswith('.html')] for file in list_of_files: notepad.open(file) editor.rereplace(FIND_WHAT, REPLACE_WITH) notepad.save() notepad.close()

    This script looks for html files in a defined directory.
    If your files are under certain different subdirectories we need to modify the code.


  • @Rafael-Lopes-Vivian

    I think part of what is in play here is that UTF-8 is the default encoding for an HTML-5 file.

    The characters that make up a “text” file are code points. A character encoding scheme maps the code points it understands into numbers that are stored in the file.

    UTF-8 and ISO-8859-1 are two different character encoding schemes. A file can technically be encoded in one or the other (or some other scheme entirely) but not in both.

    However, when a file contains only code points that are encoded identically by 2 or more encoding schemes, then, unless there is some special meta-data in the file to indicate which encoding scheme is being used, the “proper” scheme is not knowable from the file’s contents.

    UTF-8 is a variable-length encoding (the size of the number each code point is mapped to varies); ISO-8859-1 is a fixed length encoding (the size of the number each code point is mapped to is 1 byte).

    Since ISO-8859-1 represents each code point in a byte, it can only encode 256 code points, the first 256 code points of the Unicode character set. UTF-8 can be used to encode most (if not all) code points of the UNICODE character set.

    Code points 0 - 127 are encoded identically by the UTF-8 and ISO-8859-1 schemes. Code points 128 - 255 differ by becoming a 2-byte sequence with UTF-8 whereas they are single bytes with ISO-8859-1.

    So, if an ISO-8859-1 encoded file contains any code point from 128-255, then it will be a different size than a UTF-8 encoded file that contains the same code points.

  • @Aleksei-Samov

    not sure if this is a possible solution but
    what if you define a delimiter with


    and allow nesting with keywords and numbers and additional delimiters if needed?


  • @guy038

    I can confirm that this happens at my installation as well.
    And it affects replace all in all opened documents as well as find in files.


  • @Jim-Dailey
    Hello Jim,

    to be honest I don’t understand how it works at all
    and to be even more honest, I use duckduckgo/google etc…
    limiting search to the community web site address and do searches.


  • @Harry-Binswanger

    glad you solved it.


  • @Nicole-Sigaud

    as you already found out, macro recording sometimes doesn’t record all needed steps.
    If this happens you can use the …\notepad++\localization\english.xml file to find the
    message ids needed and edit shortcuts.xml like this

    <Macro name="Convert to ansi" Ctrl="no" Alt="no" Shift="no" Key="0"> <Action type="0" message="2013" wParam="0" lParam="0" sParam="" /> <Action type="0" message="2178" wParam="0" lParam="0" sParam="" /> <Action type="2" message="0" wParam="45009" lParam="0" sParam=""/> <Action type="0" message="2179" wParam="0" lParam="0" sParam="" /> </Macro>

    The message 45009 is doing the conversion, the others were recorded.


  • Okay, I’ll compare the versions tonight.

  • @Robert-P

    looks like the underlying scintilla control doesn’t have a dedicated xaml lexer
    but question is does it have to?
    Ignore my ignorance but is xaml not just xml?
    If it is, why do you want to have a different lexer, any special reason for it?
    Or do you want to have the xaml extension assigned to the xml lexer,which btw.
    is already assigned to it in the default theme.


  • Thanks for sharing such amazing info
    كل ما يهمك سيدتي

  • For CCompletion issue, see https://www.exploit-db.com/exploits/31895/. The last version of the plugin seems to be 1.19 without an update since at least 2011-08-01.

  • @johann-woeckinger

    No idea. Have you tried adding <path to npp> to your PATH environment variable? I don’t have any reason to believe it will help, but it is something to try.

  • Gotta LOVE how windows tries to muscle out the competition!

    Okay, here is a work around:

    Create a dummy file with the extension you wish to default to your program. The file can be empty. Right click on the dummy file and select Properties. On the General tab there is a section that states:
    Type of file: <YOUR_SUFFIX>
    Opens with <CURRENT_DEFAULT_PROGRAM> and a button that says <CHANGE>. Click Change. How do you want to open <YOUR_SUFFIX> files from now on? Choose <More apps>. Scroll to the bottom and select <Look for another app on this PC> Go to C:\Program Files (x86)\Notepad++ and select <notepad++.exe> and select <OPEN> Close Properties and your done!
  • @Thom-Orahim It’s already fixed in v7.4.1

  • @go2to said:
    Several other editors accept direct copy and paste in the replace-with box, be it CR+LF, LF or combination of both. They are probably written in C++. But I think it can be done in any language since CR and LF have the same codes.\r\n\r\nDon is busy now with other issues in 7.4.1. I saw there are problems with replacing again in several open files.\r\n

    It is now one line after deleting CR+LF and adding \r\n. Everything is pasted in the replace box.
    It works flawlessly in the current file, or all open files, or in find in files.

Internal error.

Oops! Looks like something went wrong!