Update colour writing to use from_array

GitOrigin-RevId: 891ab17e101d369d4ce2e06707ece1139de1eb4b
Change-Id: I110a0328e8f061b4ea2dfe18412a63bc24421da2

man/ex.rst

diff --git a/man/ex.rst b/man/ex.rst
index e0d8b04..6aa558e 100644
--- a/man/ex.rst
+++ b/man/ex.rst
@@ -106,19 +106,19 @@
 Colour
 ^^^^^^
 
-For colour images the input rows are generally 3 times as long as
-for greyscale, because there are 3 channels, RGB, instead of just
-one, grey.  Below, the *p* literal has 2 rows of 9 values (3 RGB
+For colour images the input rows are like
+``[ R, G, B, R, G, B, R, G, B, ... ]``;
+a row for *w* pixels will have ``3 * w`` values in it,
+because there are 3 channels, RGB.
+Below, the *rows* literal has 2 rows of 9 values (3 RGB
 pixels per row).  The spaces are just for your benefit, to mark out
 the separate pixels; they have no meaning in the code. ::
 
-  import png
-  p = [(255,0,0, 0,255,0, 0,0,255),
-       (128,0,0, 0,128,0, 0,0,128)]
-  f = open('swatch.png', 'wb')
-  w = png.Writer(3, 2, greyscale=False)
-  w.write(f, p)
-  f.close()
+  import png    
+  rows = [(255,0,0, 0,255,0, 0,0,255),    
+          (128,0,0, 0,128,0, 0,0,128)]    
+  image = png.from_array(rows, "RGB")
+  image.save("swatch.png")
 
 
 More Colour