Solution #0d32fca6-aa87-42bd-a7e0-ebdc381a7ed1

1def solve(input):
2    text = input.get("text", "")
3    if not text:
4        return {"encoded_length": 0, "decoded": ""}
5
6    freqs = tuple(map(text.count, set(text)))
7    if len(freqs) == 1:
8        return {"encoded_length": len(text), "decoded": text}
9
10    def merge_cost(seq):
11        n = len(seq)
12        if n == 1:
13            return 0
14        if n == 2:
15            return seq[0] + seq[1]
16
17        a = seq[0]
18        b = seq[1]
19        s = a + b
20
21        rest = seq[2:]
22        lo = tuple(filter(lambda x: x < s, rest))
23        hi = tuple(filter(lambda x: x >= s, rest))
24        merged = lo + (s,) + hi
25        return s + merge_cost(merged)
26
27    return {"encoded_length": merge_cost(tuple(sorted(freqs))), "decoded": text}

1def solve(input):
2    text = input.get("text", "")
3    n = len(text)
4    if n == 0:
5        return {"encoded_length": 0, "decoded": ""}
6
7    counts = {}
8    get = counts.get
9    for ch in text:
10        counts[ch] = get(ch, 0) + 1
11
12    freqs = list(counts.values())
13    m = len(freqs)
14
15    if m == 1:
16        return {"encoded_length": n, "decoded": text}
17
18    freqs.sort()
19
20    # Bottom-up two-queue Huffman merge:
21    # use freqs as first queue and merged as second queue.
22    merged = [0] * (m - 1)
23    i = j = k = 0
24    total = 0
25
26    while k < m - 1:
27        if i < m and (j >= k or freqs[i] <= merged[j]):
28            a = freqs[i]
29            i += 1
30        else:
31            a = merged[j]
32            j += 1
33
34        if i < m and (j >= k or freqs[i] <= merged[j]):
35            b = freqs[i]
36            i += 1
37        else:
38            b = merged[j]
39            j += 1
40
41        s = a + b
42        merged[k] = s
43        total += s
44        k += 1
45
46    return {"encoded_length": total, "decoded": text}