update the if-query so that it spans all possible contact with cells cases.

[elmbeech]

the possible cases:

n:  contact with cell type
n + 1:  contact with live cell
n + 2:  contact with dead cell  <= here the current implementation stops
n + 3:  contact with apoptotic cell  <= bug fix
n + 4:  contact with necrotic cell  <= bug fix
n + 5:  contact with other dead cell  <= bug fix

[drbergman]

great catch! Seems like the right fix. I found two other bugs to fix:

1. At the end of get_cell_contact_signals

std::string search_for = "contact with " + cell_definitions_by_type[0]->name; 
static int scaling_start_index = find_signal_index( search_for ); 
for( int i=0; i < n+2 ; i++ )
{ output[i] /= signal_scales[scaling_start_index+i]; }

we should loop until i < n+5

2. In this block in get_selected_signals

for( int i=0; i < indices.size() ; i++ )
{
	int ind = indices[i]; 
	if( ind >= contact_start_index && ind < contact_start_index+n+2)
	{ signals[i] = contact_signals[ind-contact_start_index]; }
	else
	{ signals[i] = get_single_signal( pCell , ind ); }
}

the check should instead be ind < contact_start_index+n+5 (the 5 is the difference). @vincent-noel It looks like PhysiBoSS uses get_selected_signals. Ultimately, that falls back to get_single_signal which would spam a warning message if this bug was affecting your users. So no extra concern here for this function.

[elmbeech]

@drbergman , i wonder, would it maybe not be better to make constant:

int contact_with = 5;

and then just use this constant everywhere.
so the value has to be set only once.
(i am not sure about what would be the best name for this constant.)

[drbergman]

sure! Makes sense as all these changes are within one file.

[elmbeech]

@drbergman finally, i came back to this.
i added a contact_with_built_in variable, as discussed.