How to highlight the underlined squiggle words

Michael Vincent

@Michael-Vincent said in How to highlight the underlined squiggle words:

However, it only seems to get the next squiggly indicator if it’s visible on screen

Now thinking this is an optimization in DSpellCheck for larger documents - only spell check the visible portions of a large document so the squigglys aren’t added to misspelled words off screen until those words are scrolled on screen which is why a top to bottom document search doesn’t turn up squigglies off screen.

Investigating more with “PerlScript”.

Cheers.

Michael Vincent

@Michael-Vincent said in How to highlight the underlined squiggle words:

Investigating more with “PerlScript”.

I give up. Using the script to “scroll” through the doc does find more squigglies, but not all of them. I’m pretty sure this is due to a DSpellCheck optimization as I alluded to in previous post. Need to put my own indicators on a large document and see if the SCI_INDICATORSTART/END API finds them all.

I tried 2 ways, with SCI_INDICATORSTART/END:

#!perl

use strict;
use warnings;

use Time::HiRes qw/sleep/;
use Win32::Mechanize::NotepadPlusPlus ':main';

my $indicator = 19;
my $increment = 750;
my ( $indicStart, $indicEnd ) = ( 0, 0 );
my $loc = 0;
my $end = editor->getLength();
my $HELP = 0;
while ( $loc < $end ) {

    if ( not $loc % $increment ) {
        editor->gotoPos( $loc );
        sleep 0.5;
    }

    $indicStart = editor->indicatorEnd( 19, $loc );
    if ( $indicStart == 0 ) {
        $HELP = 1;
        next;
    }
    $indicEnd = editor->indicatorEnd( 19, $indicStart );

    if ( $indicEnd == $end ) {
        $HELP = 1;
        next;
    }
    $loc = $indicEnd;
    print "$indicStart - $indicEnd\n";
} continue {
    if ( $HELP ) {
        $HELP = 0;
        $loc = int( $loc / $increment +1 ) * $increment;
        if ( $loc >= $end ) {
            last;
        }
    }
}

and with SCI_INDICATORVALUEAT:

#!perl

use strict;
use warnings;

use Time::HiRes qw/sleep/;
use Win32::Mechanize::NotepadPlusPlus ':main';

my $indicator = 19;

my $FOUND = 0;
for my $loc ( 0 .. editor->getLength() ) {

    if ( not $loc % 500 ) {
        editor->gotoPos( $loc );
        sleep 0.50;
    }
    my $iva = editor->indicatorValueAt( $indicator, $loc );
    if ( $iva and not $FOUND ) {
        print "$loc";
        $FOUND = 1;
    } elsif ( not $iva and $FOUND ) {
        printf " - %i\n", $loc - 1;
        $FOUND = 0;
    }
}

Cheers.

guy038

Hello, @mia-nguyen, @peterjones, @alan-kilborn, @ekopalypse, @michael-vincent and All,

Two months ago, I asked Sergey Smushin ( aka @Predelnik ), the DSpellCheck’s author, for the following request :

https://github.com/Predelnik/DSpellCheck/issues/201

I saw that a commit, referencing that issue, has been added :

https://github.com/Predelnik/DSpellCheck/commit/9b5e740ce4440669d6daf29ca98d77edd99d466c

From the moment that any line containing a misspelled word can be bookmarked, it would be easy to gather all these lines for subsequent processes !

@michael-vincent said :

Now thinking this is an optimization in DSpellCheck for larger documents - only spell check the visible portions of a large document so the squigglys aren’t added to misspelled words off screen until those words are scrolled on screen which is why a top to bottom document search doesn’t turn up squigglies off screen.

If so, may be, studying the commit, above, would give some hues about the way, used by @predelnik, to bookmark, in one go, all lines with misspelled words ?

Just a non-coder thought ;-))

BR

guy038

Ekopalypse

@Michael-Vincent

I did a quick test and it seems that logic below works

INDICATOR_TO_SEARCH = 31
length = editor.getLength()
start = editor.indicatorStart(INDICATOR_TO_SEARCH, -1)
end = editor.indicatorEnd(INDICATOR_TO_SEARCH, -1)

real_start = start if editor.indicatorValueAt(INDICATOR_TO_SEARCH, start) else end
print(start, end, length, real_start)

i = 100
while real_start != length:
    i -= 1
    if i == 0:
        print('ahhh endless loop - breaking')
        break
    end = editor.indicatorEnd(INDICATOR_TO_SEARCH, real_start)
    print(editor.getTextRange(real_start, end))
    print()
    real_start = editor.indicatorEnd(INDICATOR_TO_SEARCH, end)

Ekopalypse

Maybe it is a good idea to check if start != end before executing
the while loop to see if there has been an indicator found at all
and to prevent endless loops. :)

Michael Vincent

@Ekopalypse said in How to highlight the underlined squiggle words:

logic below works

Thanks. Converted to PerlScript:

#!perl

use strict;
use warnings;

use Win32::Mechanize::NotepadPlusPlus ':main';
my $npp = notepad();

my $indicator = 19;
my $length    = editor->getLength();
my $start     = editor->indicatorStart( $indicator, -1 );
my $end       = editor->indicatorEnd( $indicator, -1 );

my $real_start;

if ( editor->indicatorValueAt( $indicator, $start ) ) {
    $real_start = $start;
} else {
    $real_start = $end;
}

print "$start, $end, $length, $real_start\n";

my $i = 100;
while ( $real_start != $length ) {
    $i -= 1;
    if ( $i == 0 ) {
        print "ahhh endless loop - breaking\n";
        last;
    }
    $end = editor->indicatorEnd( $indicator, $real_start );
    print editor->getTextRange( $real_start, $end ); 
    print "\n";
    $real_start = editor->indicatorEnd( $indicator, $end );
}

It “works” but same issue, only finding the squigglies on screen. If I scroll down to some new squigglies, it only finds those on screen again (i.e., not the previous ones that were on screen but are no longer). Becoming more convinced this is a DSpellCheck optimization - only squiggly-ing the misspellings on screen.

Cheers.

Ekopalypse

@Michael-Vincent

hmm, did the test using the mark indicator from find dialog,
which worked, maybe I should have used/installed dspell plugin.
Sorry for the noise.

Michael Vincent

@Ekopalypse said in How to highlight the underlined squiggle words:

Sorry for the noise.

No noise at all! In fact you pretty much confirmed it’s a DSpellCheck optimization . I was going to get around to writing a script to add indicators but didn’t know the Find => Mark could do that. Using the same long file I’ve been testing on, I Find => Mark a word I know recurs over 100 times throughout the file and all my scripts (2 previous and the one I converted from your Python example) found more than just those indicators on screen in view.

Cheers.

Ekopalypse

I guess for a plugin like DSpellCheck it makes sense to do its work
only for the current screen but on the other side it might be possible
to do it in the background as well - but threading and scintilla and npp …
maybe an async method would be safer.

Ekopalypse

@Michael-Vincent said in How to highlight the underlined squiggle words:

Becoming more convinced this is a DSpellCheck optimization

I guess you are correct, as to this it checks only the visible area.

Mia Nguyen

Hi @all ,
Me again another doubt. If the word is correct should be legitimate and I right clicked to add into Dictionary, it will be added into my own Npp only.
Is there anyway after all i extract the plug in and gave it for other people so they can use on their own Npp ?

PeterJones

@Mia-Nguyen said in How to highlight the underlined squiggle words:

Hi @all ,
Is there anyway after all i extract the plug in and gave it for other people so they can use on their own Npp ?

Assuming your language is US-English, it puts it into %AppData%\Notepad++\Plugins\Config\Hunspell\en_US.usr … thus, I would assume that it goes into that same folder, but with the appropriate language code (for whatever spelling language you have selected) as the file basename, with .usr as the extension. So you could distribute that file, which the other users can then merge with their dictionary (assuming they’ve also added words).

Mia Nguyen

@PeterJones If i send them my .usr file. Can they use directly or have to do anything to merge?

PeterJones

@Mia-Nguyen said in How to highlight the underlined squiggle words:

@PeterJones If i send them my .usr file. Can they use directly or have to do anything to merge?

If they didn’t have any custom words yet, or if you can guarantee that your .usr file has every custom word they’ve already added, then they can just use your file directly. But if they save your .usr file overtop of their old one, then any custom words they had before would be lost.

Alan Kilborn

…any custom words they had before would be lost.

If the .usr file is in text format (don’t know, don’t use the spell-check plugin), perhaps the Compare plugin or a script to perform an automatic merge is called for. Just sayin.