I can confirm this in at least version 0.91.3. The problem seems to be
caused by the midpoint of the stem being included in the path twice in a
row. The following patch removes the midpoint altogether and fixes the
rendering problem at least on my version of xpdf. I'm kind of swamped
right now, so I hope someone else can check whether this breaks any
other uses of arrows, or some other backend. (If this doesn't seem to
happen, please file a bug report in the tracker so it isn't forgotten.)
The patch is against the maintenance branch, but the trunk has
similar-looking code.

Index: lib/matplotlib/patches.py
===================================================================
--- lib/matplotlib/patches.py   (revision 5366)
+++ lib/matplotlib/patches.py   (working copy)
@@ -634,7 +634,13 @@
                 if shape == 'right':
                     coords = right_half_arrow
                 elif shape == 'full':
-                    
coords=npy.concatenate([left_half_arrow,right_half_arrow[::-1]])
+                    # Concatenating the full paths caused the midpoint
+                    # of the stem to be included twice, which was
+                    # rendered badly by xpdf. Since the point is right
+                    # between the corners of the stem, we can drop it
+                    # from both halves.
+                    coords=npy.concatenate([left_half_arrow[:-1],
+                                            right_half_arrow[-2::-1]])
                 else:
                     raise ValueError, "Got unknown shape: %s" % shape
             cx = float(dx)/distance


-- 
Jouni K. Seppänen
http://www.iki.fi/jks


-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
_______________________________________________
Matplotlib-users mailing list
Matplotlib-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/matplotlib-users

Reply via email to