Character count bug

Eduardo Pinheiro Guerra

I believe that when there is a “–” character it is not counted.
It is not the “hyphen”, it is the “dash”

For example, as the counting in the software is shown:

EITHALOG LOGISTICA E TRANSPORTES EI = 35 characters
EITHALO- LOGISTICA E TRANSPORTES EI = 35 characters
EITHALO– LOGISTICA E TRANSPORTES EI = 34 characters

If you copy and paste from the forum, it does not show the error, so I’m making a file available for viewing.

It is a file that is generated automatically through a logistic control software.

https://www.dropbox.com/s/6bl5951wf6faaep/Notepad%2B%2B.zip?dl=0

Claudia Frank

@Eduardo-Pinheiro-Guerra

you are correct, and it looks like more chars are affected - every char within 0x80 and 0xBF.
But if you use “convert to utf-8”, than it is correctly counted.
The issue seems to be in ScintillaEditView.h

long getUnicodeSelectedLength() const
{
	// return -1 if it's multi-selection or rectangle selection
	if ((execute(SCI_GETSELECTIONS) > 1) || execute(SCI_SELECTIONISRECTANGLE))
		return -1;
	auto size_selected = execute(SCI_GETSELTEXT);
	char *selected = new char[size_selected + 1];
	execute(SCI_GETSELTEXT, 0, reinterpret_cast<LPARAM>(selected));
	char *c = selected;
	long length = 0;
	while(*c != '\0')
	{
		if( (*c & 0xC0) != 0x80)
			++length;
		++c;
	}
	delete [] selected;
	return length;
}

but I’m unsure why this is done.

Cheers
Claudia