aboutsummaryrefslogtreecommitdiff
path: root/lib/CSplit.py
diff options
context:
space:
mode:
Diffstat (limited to 'lib/CSplit.py')
-rw-r--r--lib/CSplit.py124
1 files changed, 62 insertions, 62 deletions
diff --git a/lib/CSplit.py b/lib/CSplit.py
index a28b1c8..03559c1 100644
--- a/lib/CSplit.py
+++ b/lib/CSplit.py
@@ -1,70 +1,70 @@
# A CSplit is a Clock-shaped split: the children are grouped in a circle.
# The numbering is a little different from a real clock: the 12 o'clock
-# position is called 0, not 12. This is a little easier since Python
-# usually counts from zero. (BTW, there needn't be exactly 12 children.)
+# position is called 0, not 12. This is a little easier since Python
+# usually counts from zero. (BTW, there needn't be exactly 12 children.)
from math import pi, sin, cos
from Split import Split
class CSplit() = Split():
- #
- def minsize(self, m):
- # Since things look best if the children are spaced evenly
- # along the circle (and often all children have the same
- # size anyway) we compute the max child size and assume
- # this is each child's size.
- width, height = 0, 0
- for child in self.children:
- wi, he = child.minsize(m)
- width = max(width, wi)
- height = max(height, he)
- # In approximation, the diameter of the circle we need is
- # (diameter of box) * (#children) / pi.
- # We approximate pi by 3 (so we slightly overestimate
- # our minimal size requirements -- not so bad).
- # Because the boxes stick out of the circle we add the
- # box size to each dimension.
- # Because we really deal with ellipses, do everything
- # separate in each dimension.
- n = len(self.children)
- return width + (width*n + 2)/3, height + (height*n + 2)/3
- #
- def getbounds(self):
- return self.bounds
- #
- def setbounds(self, bounds):
- self.bounds = bounds
- # Place the children. This involves some math.
- # Compute center positions for children as if they were
- # ellipses with a diameter about 1/N times the
- # circumference of the big ellipse.
- # (There is some rounding involved to make it look
- # reasonable for small and large N alike.)
- # XXX One day Python will have automatic conversions...
- n = len(self.children)
- fn = float(n)
- if n = 0: return
- (left, top), (right, bottom) = bounds
- width, height = right-left, bottom-top
- child_width, child_height = width*3/(n+4), height*3/(n+4)
- half_width, half_height = \
- float(width-child_width)/2.0, \
- float(height-child_height)/2.0
- center_h, center_v = center = (left+right)/2, (top+bottom)/2
- fch, fcv = float(center_h), float(center_v)
- alpha = 2.0 * pi / fn
- for i in range(n):
- child = self.children[i]
- fi = float(i)
- fh, fv = \
- fch + half_width*sin(fi*alpha), \
- fcv - half_height*cos(fi*alpha)
- left, top = \
- int(fh) - child_width/2, \
- int(fv) - child_height/2
- right, bottom = \
- left + child_width, \
- top + child_height
- child.setbounds((left, top), (right, bottom))
- #
+ #
+ def minsize(self, m):
+ # Since things look best if the children are spaced evenly
+ # along the circle (and often all children have the same
+ # size anyway) we compute the max child size and assume
+ # this is each child's size.
+ width, height = 0, 0
+ for child in self.children:
+ wi, he = child.minsize(m)
+ width = max(width, wi)
+ height = max(height, he)
+ # In approximation, the diameter of the circle we need is
+ # (diameter of box) * (#children) / pi.
+ # We approximate pi by 3 (so we slightly overestimate
+ # our minimal size requirements -- not so bad).
+ # Because the boxes stick out of the circle we add the
+ # box size to each dimension.
+ # Because we really deal with ellipses, do everything
+ # separate in each dimension.
+ n = len(self.children)
+ return width + (width*n + 2)/3, height + (height*n + 2)/3
+ #
+ def getbounds(self):
+ return self.bounds
+ #
+ def setbounds(self, bounds):
+ self.bounds = bounds
+ # Place the children. This involves some math.
+ # Compute center positions for children as if they were
+ # ellipses with a diameter about 1/N times the
+ # circumference of the big ellipse.
+ # (There is some rounding involved to make it look
+ # reasonable for small and large N alike.)
+ # XXX One day Python will have automatic conversions...
+ n = len(self.children)
+ fn = float(n)
+ if n = 0: return
+ (left, top), (right, bottom) = bounds
+ width, height = right-left, bottom-top
+ child_width, child_height = width*3/(n+4), height*3/(n+4)
+ half_width, half_height = \
+ float(width-child_width)/2.0, \
+ float(height-child_height)/2.0
+ center_h, center_v = center = (left+right)/2, (top+bottom)/2
+ fch, fcv = float(center_h), float(center_v)
+ alpha = 2.0 * pi / fn
+ for i in range(n):
+ child = self.children[i]
+ fi = float(i)
+ fh, fv = \
+ fch + half_width*sin(fi*alpha), \
+ fcv - half_height*cos(fi*alpha)
+ left, top = \
+ int(fh) - child_width/2, \
+ int(fv) - child_height/2
+ right, bottom = \
+ left + child_width, \
+ top + child_height
+ child.setbounds((left, top), (right, bottom))
+ #