September 2017 Update

This commit is contained in:
Ingy döt Net 2017-09-23 10:01:46 +02:00
parent bba7bfd280
commit ba8067c3b7
14570 changed files with 153136 additions and 63871 deletions

View file

@ -1,85 +0,0 @@
huffman_encoding_table = (counts) ->
# counts is a hash where keys are characters and
# values are frequencies;
# return a hash where keys are codes and values
# are characters
build_huffman_tree = ->
# returns a Huffman tree. Each node has
# cnt: total frequency of all chars in subtree
# c: character to be encoded (leafs only)
# children: children nodes (branches only)
q = min_queue()
for c, cnt of counts
q.enqueue cnt,
cnt: cnt
c: c
while q.size() >= 2
a = q.dequeue()
b = q.dequeue()
cnt = a.cnt + b.cnt
node =
cnt: cnt
children: [a, b]
q.enqueue cnt, node
root = q.dequeue()
root = build_huffman_tree()
codes = {}
encode = (node, code) ->
if node.c?
codes[code] = node.c
else
encode node.children[0], code + "0"
encode node.children[1], code + "1"
encode(root, "")
codes
min_queue = ->
# This is very non-optimized; you could use a binary heap for better
# performance. Items with smaller priority get dequeued first.
arr = []
enqueue: (priority, data) ->
i = 0
while i < arr.length
if priority < arr[i].priority
break
i += 1
arr.splice i, 0,
priority: priority
data: data
dequeue: ->
arr.shift().data
size: -> arr.length
_internal: ->
arr
freq_count = (s) ->
cnts = {}
for c in s
cnts[c] ?= 0
cnts[c] += 1
cnts
rpad = (s, n) ->
while s.length < n
s += ' '
s
examples = [
"this is an example for huffman encoding"
"abcd"
"abbccccddddddddeeeeeeeee"
]
for s in examples
console.log "---- #{s}"
counts = freq_count(s)
huffman_table = huffman_encoding_table(counts)
codes = (code for code of huffman_table).sort()
for code in codes
c = huffman_table[code]
console.log "#{rpad(code, 5)}: #{c} (#{counts[c]})"
console.log()

View file

@ -1,34 +0,0 @@
> coffee huffman.coffee
---- this is an example for huffman encoding
000 : n (4)
0010 : s (2)
0011 : m (2)
0100 : o (2)
01010: t (1)
01011: x (1)
01100: p (1)
01101: l (1)
01110: r (1)
01111: u (1)
10000: c (1)
10001: d (1)
1001 : i (3)
101 : (6)
1100 : a (3)
1101 : e (3)
1110 : f (3)
11110: g (1)
11111: h (2)
---- abcd
00 : a (1)
01 : b (1)
10 : c (1)
11 : d (1)
---- abbccccddddddddeeeeeeeee
0 : e (9)
1000 : a (1)
1001 : b (2)
101 : c (4)
11 : d (8)