Is there a way/plugin that makes notepad++ run code on a file when we open it?

Claudia Frank

do you mean the same as pressing ALT+0 ?

Cheers
Claudia

Alguem Meugla

Something similar, but it would collapse only the biggest things, ie: see this gif. It shouldn’t collapse more things inside those which are being collapsed.
It would also make it possible to choose which things should be collapsed at start. For example, on an html file I would only want it to close the head and some scripts while keeping others opened.

Claudia Frank

@Alguem-Meugla

Something similar, but it would collapse only the biggest things, ie: see this gif. It shouldn’t collapse more things inside those which are being collapsed.

A python script like this would do this

_number_of_lines = editor.getLineCount()
i = 0
while True:
    if i < _number_of_lines:
        if editor.getFoldLevel(i) & FOLDLEVEL.HEADERFLAG:
            editor.foldLine(i,0)
            i = editor.getLastChild(i,-1)
        else:
            i += 1
    else:
        break

It would also make it possible to choose which things should be collapsed at start. For example, on an html file I would only want it to close the head and some scripts while keeping others opened.

How should this work? A fold flag doesn’t know if it is set for the header tag or something else.
Of course you could anaylze the code for the head tag but how should teh script know
which scripts you want to have collapsed and which expanded?

Cheers
Claudia

Alguem Meugla

@Claudia-Frank said:

How should this work? A fold flag doesn’t know if it is set for the header tag or something else.
Of course you could anaylze the code for the head tag but how should teh script know
which scripts you want to have collapsed and which expanded?

Cheers
Claudia

It would “click” the + that is on a specified line or the line were the flag is.

Claudia Frank

@Alguem-Meugla

No, this is not what I meant.
Let’s assume we have the following code

<!DOCTYPE html>
<html>
    <head>
        <script>
            function func_h1() {
                document.getElementById("h1").innerHTML = "Should I be collapsed?";
            }
        </script>
    </head>
    <body>

        <h2>JavaScript Example</h2>

        <p id="h1">Head1</p>
        <p id="b1">Body1</p>

        <button type="button" onclick="func_h1()">Head1</button>
        <button type="button" onclick="func_b1()">Body1</button>


        <script>
            function func_b1() {
                document.getElementById("b1").innerHTML = "Or Should I be collpased?";
            }
        </script>
    </body>
</html>

How should a script decide which of the script tag needs to be collapsed?
There must be some logic provided by you to collapse the correct tag.

Cheers
Claudia

Alguem Meugla

.

<!DOCTYPE html>
<html>
<head>       <!-- NPP_PLZ_COLLAPSE_THIS_LINE -->
    <script>
        function func_h1() {
            document.getElementById("h1").innerHTML = "Should I be collapsed?";
        }
    </script>
</head>
<body>

    <h2>JavaScript Example</h2>

    <p id="h1">Head1</p>
    <p id="b1">Body1</p>

    <button type="button" onclick="func_h1()">Head1</button>
    <button type="button" onclick="func_b1()">Body1</button>


    <script>       // NPP_PLZ_COLLAPSE_THIS_LINE
        function func_b1() {
            document.getElementById("b1").innerHTML = "Or Should I be collpased?";
        }
    </script>
    <script>
        function func_b1() {
            document.getElementById("b1").innerHTML = "Or Should I be collpased?";
        }
    </script>
</body>
</html>

It would then “click” on the - that is on the right of the number of the lines that contains “NPP_PLZ_COLLAPSE_THIS_LINE” (btw when I said “+” above in a post I meant “-”).

OR

1.<!DOCTYPE html>
2.<html>
3.<head>       <!-- NPP_PLZ_COLLAPSE_THE_LINES(3,21) -->
4.    <script>
5.        function func_h1() {
6.            document.getElementById("h1").innerHTML = "Should I be collapsed?";
7.        }
8.    </script>
9.</head>
10.<body>
11.
12.    <h2>JavaScript Example</h2>
13. 
14.    <p id="h1">Head1</p>
15.    <p id="b1">Body1</p>
16.
17.    <button type="button" onclick="func_h1()">Head1</button>
18.    <button type="button" onclick="func_b1()">Body1</button>
19.
20.
21.    <script>
22.        function func_b1() {
23.            document.getElementById("b1").innerHTML = "Or Should I be collpased?";
24.        }
25.    </script>
26.    <script>
27.        function func_b1() {
28.            document.getElementById("b1").innerHTML = "Or Should I be collpased?";
29.        }
30.    </script>
31.</body>
32.</html>

It would collapse the lines 3 and 21 because of “NPP_PLZ_COLLAPSE_THE_LINES(3,21)”.

Claudia Frank

@Alguem-Meugla

ok, let’s finalize it.
The folding logic is the following

Searching through the document to find either the identifier to collapse
the fold header in this line or the identifier to collapse every root fold header starting
from A to (including) Z.

In the case of folding A to Z the identifier is immediately followed by the start and end line separated by a comma and encased by round parenthesis.

In any case only this level will be folded so if clicking onto the plus sign
the whole structure gets expanded again.

Program logic is,
whenever a file gets activated first it gets scanned about the identifiers and
the parts get collapsed. If a file gets closed and reopened the logic will be applied again.
If only the buffer is switch, meaning you switch from one tab to another one
nothing happens.

Do you agree?

Cheers
Claudia

Alguem Meugla

@Claudia-Frank said:

Program logic is,
whenever a file gets activated first it gets scanned about the identifiers and
the parts get collapsed. If a file gets closed and reopened the logic will be applied again.
If only the buffer is switch, meaning you switch from one tab to another one
nothing happens.

Do you agree?

That’s it!

Claudia Frank

@Alguem-Meugla

ok - I don’t find the time today but I will provide a possible solution tomorrow.

Cheers
Claudia

savage roar

I have a similar problem.
I wrote a script that solves it, but I can’t make it work properly in FILEOPENED callback.
This works just fine.

from Npp import *

# Collapses all blocks annotated with "//!"
def auto_collapse():
	to_collapse = [];
	for i in range(0, editor.getLineCount()-1):
		line = editor.getLine(i).strip()
		if line[:3] == "//!":
			to_collapse.append(i+1)
	for num in to_collapse:
		editor.foldLine(num, 0)

auto_collapse()

But this script behaves buggy as hell (no errors, just incorrect folding). Why? Maybe there is another “FILEREADY” callback to hook on?

from Npp import *

def auto_collapse(args):
	notepad.activateBufferID(args["bufferID"])
	to_collapse = [];
	for i in range(0, editor.getLineCount()-1):
		line = editor.getLine(i).strip()
		if line[:3] == "//!":
			to_collapse.append(i+1)
	for num in to_collapse:
		editor.foldLine(num, 0)
		
notepad.callback(auto_collapse, [NOTIFICATION.FILEOPENED])

savage roar

Okay, now THIS works for me. It seems like that dirty hack with “new/close” forces N++ to repaint the folded document. Now I’m searching for the best way to launch this on N++ startup.

from Npp import *

def auto_collapse(args):
	notepad.activateBufferID(args["bufferID"])
	for i in range(0, editor.getLineCount()-1):
		line = editor.getLine(i).strip()
		if line[:3] == "//!":
			editor.foldLine(i+1, 0)
	notepad.new()
	notepad.close()

notepad.callback(auto_collapse, [NOTIFICATION.FILEOPENED])

Claudia Frank

@Alguem-Meugla

Ok, this wasn’t as straightforward as I thought it is, basically because
the folding flags get set after buffer has been activated, therefore
I’m using now UPDATEUI callback, with some logic to prevent multiple scans, to do the folding

Within the configuration area you can define which file extensions
should be handled. COLLPASE_LINE_IDENTIFIER and COLLPASE_FROM_TO_IDENTIFIER
can be changed as long as the changes do not break regex searches.
In addition for COLLPASE_FROM_TO_IDENTIFIER matches need to be in form any_numberCOMMAany_number (like 3,4 or 14,25).

In order to make this work automatically, meaning that the script gets called when npp starts,
create a new script and name it startup.py (if you not already have a user startup.py - not to be confused with the standard startup.py python script automatically created)
and copy the content into it. Restart npp and test it.

If you do it like that then the logic is the following:
When starting npp, the script gets executed and registers two callbacks.
The BUFFERACTIVED callback gets called whenever you switch to another tab,
then the function checks if this file should be folded and if so if it has been already done.
If it hasn’t be done it set a global flag that it needs to be done.
The second callback UPDATEUI is called whenever something changes within the current document,
if the flag has been set that the document should be scanned, it will do so, collapses
the found lines and set the global flag to false in order to prevent subsequent scans.

import os

# ---------------------------- configuration area -----------------------------

FILE_EXTENSION_LIST = ['.html','.js']
COLLPASE_LINE_IDENTIFIER = 'NPP_COLLAPSE_IT'
COLLPASE_FROM_TO_IDENTIFIER = 'NPP_COLLAPSE\((\d+,\d+)\)'

# -----------------------------------------------------------------------------


def scan(number_of_lines):
    
    def find_line_identifiers(): # NPP_COLLAPSE_IT
        matches = []
        editor.research('\\b{0}\\b'.format(COLLPASE_LINE_IDENTIFIER), lambda m: matches.append(m.span()), 0)   
        return matches
    
    def find_from_to_line_identifiers(): # NPP_COLLAPSE(19,41)
        matches = []
        editor.research('{0}'.format(COLLPASE_FROM_TO_IDENTIFIER), lambda m: matches.append(m.span(1)), 0)   
        return matches
            
    for match in find_line_identifiers():
        _line = editor.lineFromPosition(match[0])
        if editor.getFoldLevel(_line) & FOLDLEVEL.HEADERFLAG:
            editor.foldLine(_line,0)
    
    for match in find_from_to_line_identifiers():
        start_line, end_line = list(map(lambda x: int(x)-1, editor.getTextRange(*match).split(',')))
        while True:
            if start_line <= end_line:
                if editor.getFoldLevel(start_line) & FOLDLEVEL.HEADERFLAG:
                    editor.foldLine(start_line,0)
                    start_line = editor.getLastChild(start_line,-1) + 1
                else:
                    start_line += 1
            else:
                break 

should_be_folded = False
def callback_FOLD_BUFFERACTIVATED(args):
    global should_be_folded
    _state = editor.getProperty('AUTO_FOLDED')
    if _state == '':
        filename, extension = os.path.splitext(notepad.getBufferFilename(args['bufferID']))
        if extension in FILE_EXTENSION_LIST:
            editor.setProperty('AUTO_FOLDED', '1')
            should_be_folded = True
        else:
            editor.setProperty('AUTO_FOLDED', '-1')
            should_be_folded = False
    elif _state == '-1':
        should_be_folded = False

def callback_FOLD_UPDATEUI(args):
    global should_be_folded
    if should_be_folded:
        scan(editor.getLineCount())
        should_be_folded = False

notepad.callback(callback_FOLD_BUFFERACTIVATED, [NOTIFICATION.BUFFERACTIVATED])
editor.callback(callback_FOLD_UPDATEUI, [SCINTILLANOTIFICATION.UPDATEUI])

Did only some basic tests, but hope this is useful for you.

Cheers
Claudia

Claudia Frank

@savage-roar

call the script from user startup.py or put your code into user startup.py.

Cheers
Claudia

Claudia Frank

Minor code cleanup

import os

# ---------------------------- configuration area -----------------------------

FILE_EXTENSION_LIST = ['.html','.js']
COLLPASE_LINE_IDENTIFIER = 'NPP_COLLAPSE_IT'
COLLPASE_FROM_TO_IDENTIFIER = 'NPP_COLLAPSE\((\d+,\d+)\)'

# -----------------------------------------------------------------------------


def scan(number_of_lines):
    
    def find_line_identifiers(): # NPP_COLLAPSE_IT
        matches = []
        editor.research('\\b{0}\\b'.format(COLLPASE_LINE_IDENTIFIER), lambda m: matches.append(m.span()), 0)   
        return matches
    
    def find_from_to_line_identifiers(): # NPP_COLLAPSE(19,41)
        matches = []
        editor.research(COLLPASE_FROM_TO_IDENTIFIER, lambda m: matches.append(m.span(1)), 0)   
        return matches
            
    for match in find_line_identifiers():
        _line = editor.lineFromPosition(match[0])
        if editor.getFoldLevel(_line) & FOLDLEVEL.HEADERFLAG:
            editor.foldLine(_line,0)
    
    for match in find_from_to_line_identifiers():
        start_line, end_line = map(lambda x: int(x)-1, editor.getTextRange(*match).split(','))
        while True:
            if start_line <= end_line:
                if editor.getFoldLevel(start_line) & FOLDLEVEL.HEADERFLAG:
                    editor.foldLine(start_line,0)
                    start_line = editor.getLastChild(start_line,-1) + 1
                else:
                    start_line += 1
            else:
                break 

should_be_folded = False
def callback_FOLD_BUFFERACTIVATED(args):
    global should_be_folded
    _state = editor.getProperty('AUTO_FOLDED')
    if _state == '':
        filename, extension = os.path.splitext(notepad.getBufferFilename(args['bufferID']))
        if extension in FILE_EXTENSION_LIST:
            editor.setProperty('AUTO_FOLDED', '1')
            should_be_folded = True
        else:
            editor.setProperty('AUTO_FOLDED', '-1')
            should_be_folded = False
    else:
        should_be_folded = False

def callback_FOLD_UPDATEUI(args):
    global should_be_folded
    if should_be_folded:
        scan(editor.getLineCount())
        should_be_folded = False

notepad.callback(callback_FOLD_BUFFERACTIVATED, [NOTIFICATION.BUFFERACTIVATED])
editor.callback(callback_FOLD_UPDATEUI, [SCINTILLANOTIFICATION.UPDATEUI])

Cheers
Claudia

Alguem Meugla

@Claudia-Frank said:

create a new script and name it startup.py (if you not already have a user startup.py - not to be confused with the standard startup.py python script automatically created)

Where do I find/put that?

Claudia Frank

@Alguem-Meugla

either install python script plugin via plugin manager or by using the msi package.
I would suggest to use the msi package as it has been reported that installation via plugin manager
fails from time to time.

Once installed, goto Plugins->Python Script->New Script and name it startup.py.
and press Save button.
Copy the posted script into the newly created document, save it and restart npp.

Cheers
Claudia

Claudia Frank

@Alguem-Meugla

sorry - I forgot one step.

before you restart npp, goto
Plugins->Python Script->Configuration
and instead of Initialisation: LAZY choose ATSTARTUP.

Cheers
Claudia

Alguem Meugla

@Claudia-Frank
I copied the script to Plugins>Python Script>Scripts>(Ctrl+)startup and saved as administrator.
I also did what you said on Configuration.
But it is doing nothing when I start my file with “NPP_COLLAPSE_IT” on a line with a -. I then changed the script and added a line “print(‘startup.py done!’)” on the end trying to see if Plugins>Python Script>Show Console showed it, but nothing appears there other than Ready (and some more things above it).

Claudia Frank

@Alguem-Meugla

I don’t understand what you mean by

I copied the script to Plugins>Python Script>Scripts>(Ctrl+)startup and saved as administrator.

you should have done

goto Plugins->Python Script->New Script and name it startup.py
and press Save button.
Copy the posted script into the newly created document, save it and restart npp.

Depending on your installation, the script directory is different,
by using the way I mentioned you make sure that the correct directory is used.

Cheers
Claudia

Alguem Meugla

@Claudia-Frank said:

you should have done

goto Plugins->Python Script->New Script and name it startup.py
and press Save button.
Copy the posted script into the newly created document, save it and restart npp.

Depending on your installation, the script directory is different,
by using the way I mentioned you make sure that the correct directory is used.

When I go on New Script it puts me on Documents. Should I create the file there or in a place I want? I had created a folder there and put the new file named startup.py with the script inside it, but wasn’t working, so I replaced the standard startup.py with the new one. Didn’t work.