Re: visual-line-mode

[Chong Yidong <cyd@stupidchicken.com>]

Stefan Monnier <monnier@iro.umontreal.ca> writes:

> PS: My patch is just a proof of concept.  It has a few bugs.  One of
> them can be seen when you set truncate-lines, in which case it doesn't
> preserve columns correctly (basically, its temporary-goal-column ends
> up being computed from the beginning of the displayed part of the line,
> so as long as hscroll is not changed, it works OK but as soon as
> hscroll is modified, the result is incorrect).

Does the following patch give good results/make sense?

*** trunk/src/indent.c.~1.209.~	2008-06-29 09:56:50.000000000 -0400
--- trunk/src/indent.c	2008-07-11 17:08:33.000000000 -0400
***************
*** 2058,2067 ****
--- 2058,2070 ----
        int it_start;
        int oselective;
        int it_overshoot_expected;
+       int first_x;
  
        SET_TEXT_POS (pt, PT, PT_BYTE);
        start_display (&it, w, pt);
  
+       first_x = it.first_visible_x;
+ 
        /* Scan from the start of the line containing PT.  If we don't
  	 do this, we start moving with IT->current_x == 0, while PT is
  	 really at some x > 0.  The effect is, in continuation lines, that
***************
*** 2119,2128 ****
  	move_it_by_lines (&it, XINT (lines), 0);
  
        if (!NILP (lcols))
! 	move_it_in_display_line
! 	  (&it, ZV,
! 	   (int)(cols * FRAME_COLUMN_WIDTH (XFRAME (w->frame)) + 0.5),
! 	   MOVE_TO_X);
  
        SET_PT_BOTH (IT_CHARPOS (it), IT_BYTEPOS (it));
      }
--- 2122,2135 ----
  	move_it_by_lines (&it, XINT (lines), 0);
  
        if (!NILP (lcols))
! 	{
! 	  move_it_in_display_line (&it, ZV, first_x, MOVE_TO_X);
! 	  it.current_x = 0;
! 	  move_it_in_display_line
! 	    (&it, ZV,
! 	     (int)(cols * FRAME_COLUMN_WIDTH (XFRAME (w->frame)) + 0.5),
! 	     MOVE_TO_X);
! 	}
  
        SET_PT_BOTH (IT_CHARPOS (it), IT_BYTEPOS (it));
      }