(I’m also really surprised that Michael’s code worked for him a few years ago; I don’t believe the definition of the MASK field of SCI_MARKERNEXT has changed in the last three years.)

Nevermind. The definition of Notepad++'s MARK_BOOKMARK has changed in the last three years. Specifically, in v8.4.6 in Sept 2022, it was changed from 24 to 20 to work with Scintilla 5.3.0.

So just changing that single line to

SET LOCAL MASK ~ 1<<20

would have been sufficient.

In case you don’t understand: 3 years ago, when Michael wrote the script, bookmarks used mark#24, so SET LOCAL MASK ~ 1<<24 set the mask to 0x1000000, so if bit #24 matched, it would say the line number.

With v8.4.6 and newer, that marker is in bit #20, so if you only are looking at bit#24, it’s not going to find it. Your code, with 0xFFFFFF on the other hand, was looking at all marker numbers from 0 to 23, so it would find marker#20 as the bookmark line. But it would also find any of the other markers on those lines, even the ones that aren’t bookmarks – so it would find markers you weren’t looking for. If you file happened to have other marker types – like if you had any hidden lines due to View > Hide Lines or code folding in your active lexer – then it would report false matches.

Marker Number References

• Scintilla reserves #25-31 for folding margins
• Scintilla reserves #21-24 for Change History
• Notepad++ reserves #16-20 for “internal use” (specifically, 18 & 19 for hidden lines, and 20 for bookmarks, with two more for future use)
• Notepad++ allows #0-15 for plugins to use, but makes no guarantee that two plugins won’t try to make use of the same marker.

I’m not sure why you completely reworked @Michael-Vincent’s script, instead of just fixing the third line to

SET LOCAL MASK ~ (1<<24) -1

It seems that would have been a lot less work for you than redoing all the logic, slightly differently.

(I’m also really surprised that Michael’s code worked for him a few years ago; I don’t believe the definition of the MASK field of SCI_MARKERNEXT has changed in the last three years.)

As far as redirecting the NppExec output to a file, did you look in Plugins > NppExec > Help/Manual, looking at section 4.5? Because section 4.5 is completely devoted to how to redirect output in NppExec, so that would seem like a natural place to start. As a hint, if you previously had

echo HI

you could replace it with

cmd /c echo HI > outputfile.txt

to redirect it to overwrite outputfile.txt,
or

cmd /c echo HI >> outputfile.txt

to redirect it to append to outpputfile.txt

cls
set local currentline = 0
set local breakpoints = ""
:repeat
sci_sendmsg SCI_MARKERNEXT $(currentline) 0xFFFFFF
// If marker next fails then we exit.
if $(MSG_RESULT) == -1) 
    exit
// Next marker exists.
else 
    // Check if the market has looped back to 0.
    if $(MSG_RESULT) < $(currentline)
        echo $(breakpoints)
        exit
    else
        // Increment the resulting line by 1.
        set local currentline ~ $(MSG_RESULT) + 1
        set local breakpoints = ${breakpoints} $(FILE_NAME) $(currentline) \r\n
    endif
    goto repeat
endif

I couldn’t quite figure out how to make NppExec write to a file, so if anyone can add that i’d be really grateful.

BTW, I wasn’t complaining about Notepad++'s default bookmark symbol image in any way

Didn’t think you were :-) it was more of an exercise to see if I could do it. I have an NppExec ‘marker’ script that adds different colored “bookmarks” by placing normal bookmarks and then converting them. So I though it’d be possible to just change the default symbol / color.

Cheers.

Yes, that is good information.
The Pythonscript implementation is very similar and straightforward.

BTW, I wasn’t complaining about Notepad++'s default bookmark symbol image in any way. Just wondering/commenting on its possible origins. It hadn’t occurred to me that the origins might be with an earlier VS version.

The obvious stock Scintilla choice (had it been used) would have been:

Do you have NppExec (or Python Script) that allows a script to run at startup?

SCI_SENDMSG SCI_MARKERDEFINE 24 SC_MARK_BOOKMARK
SCI_SENDMSG SCI_MARKERSETFORE 24 255
SCI_SENDMSG SCI_MARKERSETBACK 24 255

(Python Scrip implementation will vary ;-)

Cheers.

I started working with an older version of Visual Studio recently, due to what is used for certain projects at work.

I notice that its breakpointing symbol is VERY similar to Notepad’s bookmark image, the main difference being coloring:

In the past I’ve had a passing thought of “where did the design for the N++ bookmark come from?” as it isn’t a stock Scintilla indicator symbol like those shown HERE (with a little “scrolldown” from where that link goes).

Perhaps this “red” older-Visual-Studio BP symbol (shown above) was the origin of the N++ bookmark symbol.

The obvious stock Scintilla choice (had it been used) would have been:

BTW, current VS (2019) BP symboling is much less “fancy”; appears rather flat:

Then right code is:

Yup - methinks that’s correct.

Cheers.

The Scintilla docs @Alan-Kilborn linked are crucial since the Notepad++ editing component is based on Scintilla. As for “translating” my NppExec script to C++:

I’m sure about that, but what Alan Kilborn doesn’t understand is that my problem is not about Scintilla api that is documented but with the c++ conversion.
Then right code is:

				long line = 1;
				int mask = (1 << 24);
				int result=::SendMessage(curScintilla, SCI_MARKERNEXT, line, mask);

You helped me, thanks.

SendMessage(getCurScintilla(), SCI_MARKERNEXT, line, mask);

sorry:

int markerAtLine = ::SendMessage(getCurScintilla(), SCI_MARKERNEXT, line, mask);

Cheers.

I think maybe you’re not connecting the dots. The Scintilla docs @Alan-Kilborn linked are crucial since the Notepad++ editing component is based on Scintilla. As for “translating” my NppExec script to C++:

HWND getCurScintilla()
{
    int which = -1;
    ::SendMessage( nppData._nppHandle, NPPM_GETCURRENTSCINTILLA, 0,
                   ( LPARAM )&which );
    return ( which == 0 ) ? nppData._scintillaMainHandle :
           nppData._scintillaSecondHandle;
}

[...]
int line = 0;
int mask = 1<<24;
SendMessage(getCurScintilla(), SCI_MARKERNEXT, line, mask);

That should get you the first bookmark since N++ uses 24 as the bookmark marker and thus bitshitfted 24 is the mask the Scintilla call requires. You’ll need to put your own C++ code / logic around that to determine if you’ve found them all and how to find the next one, but again, my NppExec “pseudo” code should help you with that - as will the Scintilla docs.

Cheers.

long line = 0;
int marker = 0;
::SendMessage(curScintilla, SCI_MARKERGET, line, (LPARAM)&marker);

This is :

long line = 0;
int marker=::SendMessage(curScintilla, SCI_MARKERGET, line, 0);

Right ?

@Alan-Kilborn said in How to get breakpoint list ?:

I did point you to the documentation, and a specific function in those docs, but then in your follow-on posting you seemed to ignore that and went with a different function

Sorry I followed your link and see that SCI_MARKERGET returns the bookmark at a specific line, perfect for a test. I tested and it doesn’t work. But even if you’re convinced I’m a bad programmer, you think I would have more chance with SCI_MARKERNEXT that seems more advanced.
If you tell me why I get “0”, I’ll can understand and make both work, but you doesn’t seem to know c++, then just tell me that.

you should have just said, "no it’s not too much complicated, you just have to get the current scintilla, like that [little code] and retrieve the bookmark for each line like that [little code].

I don’t know…
Nobody here is going to “spoon feed” you every little detail.
If you’re a plugin writer, a certain level of “sophistication” is assumed.
Pointing you to the documentation should be enough.
Of course, I did point you to the documentation, and a specific function in those docs, but then in your follow-on posting you seemed to ignore that and went with a different function…
I’m not sure what to think.
I wish you good luck with your endeavors.

In my notes on bookmark manipulation I have this thread and specifically THIS POST noted.
The logic flow there can probably help you.

Ok, that part work, I get the scintilla for the current document, but for the next part I don’t find example, I tried that:

Here is some NppExec code to iterate the bookmarks in the active document.

SET LOCAL POS = 0
SET LOCAL FOUND = 0
SET LOCAL MASK ~ 1<<24

:LISTNEXT
SCI_SENDMSG SCI_MARKERNEXT $(POS) $(MASK)
IF $(MSG_RESULT)==-1 GOTO STOP

NPP_CONSOLE on
SET LOCAL FOUND = 1
SET LOCAL LINE ~ $(MSG_RESULT) + 1
SCI_SENDMSG SCI_GETLINE $(MSG_RESULT) @""
ECHO $(FULL_CURRENT_PATH):$(LINE):	$(MSG_LPARAM)

SET LOCAL POS = $(LINE)
GOTO LISTNEXT

:STOP
IF $(FOUND)==0 THEN
    MESSAGEBOX "No Bookmarks Found"
ENDIF

The API calls are pretty much the same (SCI_SENDMSG = ::SendMessage(getCurrScintilla(), <MESSAGE>, wParam, lParam)) and you can do C++ looping much cleaner than the GOTO required in NppExec.

I usually try to mock something up in NppExec (or “PerlScript” or PythonScript) before trying to add to a plugin.

Cheers.

int which = -1;
::SendMessage(nppData._nppHandle, NPPM_GETCURRENTSCINTILLA, 0, (LPARAM)&which);
if (which == -1)
	return FALSE;
HWND curScintilla = (which == 0) ? nppData._scintillaMainHandle : nppData._scintillaSecondHandle;

Ok, that part work, I get the scintilla for the current document, but for the next part I don’t find example, I tried that:

long line = 0;
int marker = 0;
::SendMessage(curScintilla, SCI_MARKERGET, line, (LPARAM)&marker);

But it always gave 0, can you enlighten me ?

Do you mean it’s complicated to get a list of bookmark ?

To ask such a question in that way leads me to believe that it might be complicated for you.
Haha.
Just a joke, don’t get offended!

But, I pointed you to the documentation, which I presume you looked at before you replied with the above question.
And did you see a function there that instantly provides you with a list of bookmarks?
No, because there isn’t such a function.
To get the bookmarks you have to iterate and build any list you want yourself.
You would probably use THIS to do that.
And you would also need to know that the marker number that N++ uses for bookmarks is decimal 24.

So, overall, is it complicated?
I suppose that depends upon your knowledge and your perspective.

You can read about their implementation HERE although that won’t make you an instant expert on how a plugin can gain access.

Do you mean it’s complicated to get a list of bookmark ?

is it possible to get a list of these circles fro a plugin

Are you asking if it is possible for a plugin to know if a line has a bookmark or not?

If so then the answer is yes.
The bookmark is manipulated as something called a “marker”.
You can read about their implementation HERE although that won’t make you an instant expert on how a plugin can gain access.

I suppose it is a breakpoint,

Not a breakpoint, as Notepad++ is not a debugger/IDE.

It’s a Bookmark ; see the Mark tab in the Find window (for setting bookmarks via a search), and also the Search menu’s Bookmark menu (for manipulating bookmarked content).

For those unfamiliar, this is what the OP is seeing: