[GRASS-dev] discussion: replacing ps.map

[Glynn Clements]

Jachym Cepicky wrote:

> hi,
> tahanks for you answer
> 
> 2007/4/6, Glynn Clements <glynn at gclements.plus.com>:
> >
> > J-Báchym Čepický wrote:-A
> >
> > > the PS driver works nice, thanks!
> > >
> > > What about d.text.freetype ? I wonder how complicated it would be to
> > > implement too..?
> >
> > Huh?
> >
> > d.text.freetype should work with the PS driver, although it will
> > rasterise the bitmaps at a resolution of 1 point, then embed the
> > bitmaps.
> 
> d.mon PS
> d.text.freetype font=luximr text="hallo, world" at=10,10
> d.mon stop=PS
> 
> gv map.ps
> 
> document is not opend

Oops. There are a couple of bugs in the bitmap code.

The first is in the prolog; the string is allocated with one byte per
pixel, when it should be one bit, causing too much data to be read.

	Index: lib/psdriver/psdriver.ps
	===================================================================
	RCS file: /grassrepository/grass6/lib/psdriver/psdriver.ps,v
	retrieving revision 1.2
	diff -u -r1.2 psdriver.ps
	--- lib/psdriver/psdriver.ps	30 Mar 2007 06:29:35 -0000	1.2
	+++ lib/psdriver/psdriver.ps	6 Apr 2007 13:48:00 -0000
	@@ -39,7 +39,7 @@
	 /BITMAP {
	   gsave
	     4 2 roll translate
	-    1 index string /tmpstr exch def
	+    1 index 7 add 8 idiv string /tmpstr exch def
	     true [1 0 0 1 0 0] {currentfile tmpstr readhexstring pop} imagemask
	   grestore
	 } bind def

If you make this change to the map.ps file, the file should display,
but you'll notice the second bug; the accumulator wasn't being
cleared, resulting in "echoes":

	Index: lib/psdriver/Draw_bitmap.c
	===================================================================
	RCS file: /grassrepository/grass6/lib/psdriver/Draw_bitmap.c,v
	retrieving revision 1.1
	diff -u -r1.1 Draw_bitmap.c
	--- lib/psdriver/Draw_bitmap.c	30 Mar 2007 04:51:23 -0000	1.1
	+++ lib/psdriver/Draw_bitmap.c	6 Apr 2007 13:48:00 -0000
	@@ -25,6 +25,7 @@
	 			{
	 				output("%02X", acc);
	 				bit = 0x80;
	+				acc = 0;
	 			}
	 		}
 
Both of these are fixed in CVS, and it appears to work.

> same for d.font && d.text
> 
> What am I missing?

That I said "should work" rather than "does work" ;)

This time, I have actually tested it.

-- 
Glynn Clements <glynn at gclements.plus.com>