DK sort D opted for python

#SortD  - JGMiner - April 8, 2014 
# a python optimized version of Knuth's Sort D demo 
# parses ip address by useing Py's native for each char in a string
# then uses except catch on try if int to detect the dots
# of course you could just use split.

def parseIpId(s):
    a=[0]
    for i in s:
        try : a[-1]=10*a[-1]+int(i)
        except : a.append(0)
    return a

print parseIpId("192.162.36.1")

print "Sorting algorithm D page 78, Knuth"
def recKey(Rec):
    return Rec/10                                                      # simple key for integers of 10  : divide by 10 making key domain one tenth record domain
    
# ------------ Data Structures ---------------------------

Recs = [10, 10, 30, 50, 70, 90, 20, 40, 40, 60, 0, 100, 100, 40, 30, 40]
print " Records: ", Recs

Keys = [recKey(i) for i in Recs]                                       # Define a set of keys using a method for repeatability/reuse
print " Keys: ", Keys

Sorted = [0,]*len(Recs)                                                # Clear a new array for sorted, needs to be fully defined due to random placement of results 
# ------------ Algorithm ---------------------------------

counts = [0,]*len(Keys)                                                # zero a counts array
print " D1: counts:", counts

for key in Keys: counts[key]+=1                                        # use each value in K as an index to counts
print " D2 & D3: counts:", counts

for key in range(min(Keys),max(Keys)): counts[key+1] += counts[(key)]  # sum the counts by key value order
print " D4: counst:", counts

for idx, rec in enumerate(reversed(Recs)):                             # Using records in reverse order allows the summed counts dictate sorted order
   key=recKey(rec)                                                     # calc key
   Sorted[counts[key]-1] = rec                                         # need to offset by 1 because its is a true count being used as an index
   counts[key] -= 1
print " D5 & D6: S:", Sorted

Python Fiddle

Python Cloud IDE