Benchmarks for bencodepy.encode and buxtor.bencode.

gistfile1.txt

# Speed comparison of bencodepy and buxtor.bencode
# bencodepy is consistently 0.3 - 0.5 seconds faster. Why?

# bencodepy: https://github.com/eweast/BencodePy/blob/master/bencodepy/encode.py
#          4914005 function calls (3618005 primitive calls) in 2.830 seconds
#
#
# buxtor.bencode: https://github.com/davidwtbuxton/buxtor/blob/master/buxtor/bencode.py
#          5592005 function calls (4548005 primitive calls) in 3.183 seconds


# This is the output of bencodepy's benchmark, using Python 3.5.2 on Mac OS X 10.11.6.
# (with a whole bunch of setup logging removed).

$ python bench_encoding.py
Number of objects to decode: 30000.
Benchmark using bencodepy.encode
Match
Total encode size: 1718.662 MB.
Wed Oct 12 19:11:55 2016    encode_stats

         4914005 function calls (3618005 primitive calls) in 2.830 seconds

   Ordered by: internal time

   ncalls  tottime  percall  cumtime  percall filename:lineno(function)
   630000    0.893    0.000    0.985    0.000 encode.py:13(__encode_byte_str)
    30000    0.480    0.000    0.480    0.000 {method 'join' of 'bytes' objects}
1002000/30000    0.367    0.000    2.085    0.000 encode.py:56(__select_encoder)
120000/30000    0.241    0.000    2.059    0.000 encode.py:37(__encode_dict)
    30000    0.205    0.000    2.770    0.000 encode.py:5(encode)
   630000    0.141    0.000    1.126    0.000 encode.py:47(<lambda>)
   156000    0.120    0.000    0.134    0.000 encode.py:19(__encode_int)
  1218000    0.100    0.000    0.100    0.000 {method 'append' of 'list' objects}
96000/24000    0.082    0.000    0.670    0.000 encode.py:29(__encode_list)
        1    0.060    0.060    2.830    2.830 bench_encoding.py:30(<listcomp>)
   156000    0.041    0.000    0.175    0.000 encode.py:52(<lambda>)
   630000    0.041    0.000    0.041    0.000 {built-in method builtins.len}
120000/30000    0.033    0.000    2.067    0.000 encode.py:48(<lambda>)
96000/24000    0.027    0.000    0.677    0.000 encode.py:50(<lambda>)
        1    0.000    0.000    2.830    2.830 {built-in method builtins.exec}
        1    0.000    0.000    2.830    2.830 bench_encoding.py:28(bench)
        1    0.000    0.000    2.830    2.830 <string>:1(<module>)
        1    0.000    0.000    0.000    0.000 {method 'disable' of '_lsprof.Profiler' objects}



# And the output of buxtor.bencode using the same Python, same script but modified with
# from buxtor.bencode import encode, decode


$ python bench_encoding_buxtor.py 
Number of objects to decode: 30000.
Benchmark using buxtor.bencode
Match
Total encode size: 1718.662 MB.
Wed Oct 12 19:11:04 2016    encode_stats

         5592005 function calls (4548005 primitive calls) in 3.183 seconds

   Ordered by: internal time

   ncalls  tottime  percall  cumtime  percall filename:lineno(function)
1002000/30000    2.464    0.000    3.160    0.000 bencode.py:66(encode)
  2712000    0.259    0.000    0.259    0.000 {built-in method builtins.isinstance}
   216000    0.199    0.000    0.199    0.000 {method 'join' of 'bytes' objects}
   816000    0.075    0.000    0.075    0.000 {method 'append' of 'list' objects}
   120000    0.065    0.000    0.065    0.000 {built-in method builtins.sorted}
96000/24000    0.057    0.000    0.755    0.000 bencode.py:83(<listcomp>)
   630000    0.041    0.000    0.041    0.000 {built-in method builtins.len}
        1    0.022    0.022    3.182    3.182 bench_encoding_buxtor.py:28(<listcomp>)
        1    0.000    0.000    3.183    3.183 {built-in method builtins.exec}
        1    0.000    0.000    3.182    3.182 <string>:1(<module>)
        1    0.000    0.000    3.182    3.182 bench_encoding_buxtor.py:26(bench)
        1    0.000    0.000    0.000    0.000 {method 'disable' of '_lsprof.Profiler' objects}