Marker number

Michael Vincent

Is this a feature or a bug?

Neither - you’re stepping on markers already assigned.

You need to pick a marker number that is not used by something else and thus already has it’s mask applied to a margin. 6 was indeed a good choice. 17 is bad since Notepad++ already uses that for MARK_HIDELINESUNDERLINE:

https://github.com/notepad-plus-plus/notepad-plus-plus/blob/11f7ba2e0a80dae890c5162a92635d6c6151af64/PowerEditor/src/ScintillaComponent/ScintillaEditView.h#L116-L119

25 is another bad choice since Scintilla uses that for folding margins .

Stick to the low numbered markers, like under 10 should be OK, but don’t be surprised of some wonkiness there too if a plugin you are using has already claimed one. Unfortunately there is no interface to see what markers are already assigned.

Cheers.

Alan Kilborn

@Paul-Wormer said in Marker number:

On my screen at home and in the preview the sizes of the examples are reasonable, after submit they become horrendous

I don’t know that they are all that horrendously large this time. So, it’s OK.

Paul Wormer

@Michael-Vincent Thank you. I suspected something like that. Because the Scintilla manual says that 0 to 20 are free, I had some doubts. It is indeed a pity that the marker numbers cannot be inspected.

Alan Kilborn

@Paul-Wormer said in Marker number:

Because the Scintilla manual says that 0 to 20 are free

They are free from Scintilla’s perspective.
Notepad++, as a user of Scintilla, is allowed to use the free ones as it sees fit.
You, as a user, have to avoid markers used by both Scintilla and Notepad++.
(pretty sure this is obvious by now)

a pity that the marker numbers cannot be inspected

The N++ “bookmark” marker ID can be retrieved, see HERE (and also HERE), but that is of limited usefulness because it only gives you one ID value that is tied up.

But yes, there should be a “cough up a free marker ID, and then don’t give that same marker ID to anyone else that asks for a free one” plugin message/function.

Then plugins that play nice would use that interface when they need markers for their use. This includes the PythonScript plugin. Although PS has no use for markers on its own, such a function would be made available to scripters (like you) in order to obtain a unique marker ID.

Paul Wormer

@Alan-Kilborn Yes, I ~~understand~~ that is obvious to me now. I’m groping my way into the Notepad++/Scintilla jungle and start getting a feel for its philosophy.

BTW, how about your promise of a few day ago?

Alan Kilborn

there should be a “cough up a free marker ID, and then don’t give that same marker ID to anyone else that asks for a free one” plugin message/function.

Oh wait, it appears there IS the plugin message I’m talking about. Doh!

See HERE in N++ code.

And then this in the PS docs:

I suppose I haven’t noticed this because in all of my scripting endeavors, I haven’t needed to use a custom marker (or, I just don’t remember – failing memory disclaimer!).

Paul Wormer

@Alan-Kilborn I tried in the PythonScript console:

>>> print(notepad.allocateMarker(25))
None
>>> print(notepad.allocateMarker(0))
None

What does that mean?

Alan Kilborn

@Paul-Wormer said in Marker number:

BTW, how about your promise of a few day ago?

See HERE for “status”.
:-)

Alan Kilborn

@Paul-Wormer said in Marker number:

print(notepad.allocateMarker(25))
None
print(notepad.allocateMarker(0))
None

What does that mean?

So the argument is numberRequested.

In the first case you are asking for 25, presumably contiguous, markers to be allocated. That’s way too many; that many aren’t going to be found.

In the second case you are asking for 0 markers to be allocated. That’s also pretty nonsensical.

If you want one marker, try notepad.allocateMarker(1). Each time you call that, you should get a new number returned.

(It appears you are thinking that you tell it which marker ID you want, at least from your usage of 25 – that isn’t how it works)

Paul Wormer

@Alan-Kilborn I expected a boolean upon return: if true markerNr is allocated; if false markerNr is free. That is not how it works apparently.

Alan Kilborn

@Paul-Wormer said in Marker number:

I expected a boolean upon return: if true markerNr is allocated; if false markerNr is free. That is not how it works apparently.

No, it isn’t, but…why do you care what the actual number is?
In your code, go with:

my_marker_id = notepad.allocateMarker(1)

and then wherever you were using a hard-coded-magic-number for a marker number, use my_marker_id instead:

    my_marker_id = notepad.allocateMarker(1)
    editor.markerDefine(my_marker_id, MARKERSYMBOL.CIRCLEPLUS)  
    nSymbol =  editor.markerSymbolDefined(my_marker_id)    
    editor.setMarginWidthN(marginNr, 16)               
    editor.setMarginTypeN(marginNr,  MARGINTYPE.SYMBOL)
    editor.setMarginMaskN(marginNr, 1<<my_marker_id )      
    editor.markerAdd(line-1,my_marker_id)

Note that this kind of code will allocate a new number each time you run the script. So, as you’re working on the script, you’ll probably find you’ll be out of markers fairly quickly. :-(

To solve that, you could do:

try:                                              # <-- important change!
    my_marker_id                                  # <-- important change!
except NameError:                                 # <-- important change!
    my_marker_id = notepad.allocateMarker(1)      # <-- important change!
editor.markerDefine(my_marker_id, MARKERSYMBOL.CIRCLEPLUS)  
nSymbol =  editor.markerSymbolDefined(my_marker_id)    
editor.setMarginWidthN(marginNr, 16)               
editor.setMarginTypeN(marginNr,  MARGINTYPE.SYMBOL)
editor.setMarginMaskN(marginNr, 1<<my_marker_id )      
editor.markerAdd(line-1,my_marker_id)

Paul Wormer

@Alan-Kilborn Yes, of course. It seems that the description of notepad.allocateMarker is wrong by stating --> bool. Instead it should state --> int. That put me off.

Alan Kilborn

@Paul-Wormer said in Marker number:

It seems that the description of notepad.allocateMarker is wrong by stating --> bool. Instead it should state --> int. That put me off.

Good catch!
See HERE.

Paul Wormer

@Alan-Kilborn Very good. Will Bruderstein listen, what is your experience?

Alan Kilborn

@Paul-Wormer said in Marker number:

Will Bruderstein listen, what is your experience?

Dave Bruderstein and @chcg (Christian) are active on the PythonScript project; maybe not “super” active, but active.

I’d think they’d listen, it’s just an easy documentation error.

Alan Kilborn

I wrote some test code:

for j in range(100):
    my_marker_id = notepad.allocateMarker(1)
    print(my_marker_id)
    if my_marker_id is None: break

and obtained as output:

Not sure where 0, 15, and 16 are…?

Paul Wormer

@Alan-Kilborn When I run my original script with markerNr (is marker_id) = 0 or 15, all is OK. The circleplus appears where you would expect it.

When I run your little script I get:

So during my session I claimed 0 thru 6 and 15. You probably claimed only 0 and 15. Apparently 16 and higher are taken by some plugins.

Alan Kilborn

@Paul-Wormer said in Marker number:

You probably claimed only 0 and 15. Apparently 16 and higher are taken by some plugins.

I ran on a virgin system with only the PythonScript plugin (aside from the plugins that come with Notepad++). I’m fairly confident (as I hinted before) that PS doesn’t claim any itself.