Improve evaluation experience for large values

[Cnly]

The other day I had a cider-nrepl server started inside a project that doesn't have iced-nrepl in it. When I :IcedConnect to it, it sent back a 1,201,722-byte bencoded string (I'm not familiar with the nrepl protocol so have no idea what it is; it began with a d18:changed-namespaces dict) which froze my nvim for 79 seconds... that's a throughput of 15KB/s and sounds a bit too slow. Profiling shows:

FUNCTIONS SORTED ON SELF TIME
count  total (s)   self (s)  function
   10  78.780231  78.630764  provider#python3#Call()
    8             78.627463  <SNR>292_decode_via_python()

    327019 function calls (255933 primitive calls) in 70.975 seconds

    Ordered by: standard name
 
    ncalls  tottime  percall  cumtime  percall filename:lineno(function)
         1    0.000    0.000    0.000    0.000 _bootlocale.py:33(getpreferredencoding)
         1    0.000    0.000    0.000    0.000 bencode.py:18(__decode_integer)
         1    0.000    0.000    0.000    0.000 bencode.py:29(__decode_list)
         1    0.004    0.004   70.975   70.975 bencode.py:3(<module>)
     51182   13.666    0.000   61.401    0.001 bencode.py:3(__decode_string)
    9952/1    7.920    0.001   70.969   70.969 bencode.py:46(__decode_dict)
   61136/1    1.647    0.000   70.969   70.969 bencode.py:62(__decode)
         1    0.000    0.000   70.969   70.969 bencode.py:75(iced_bencode_decode)
         1    0.000    0.000    0.000    0.000 codecs.py:260(__init__)
         1    0.000    0.000    0.000    0.000 codecs.py:309(__init__)
         1    0.000    0.000    0.002    0.002 codecs.py:319(decode)
         1    0.002    0.002    0.002    0.002 {built-in method _codecs.utf_8_decode}
         1    0.000    0.000    0.000    0.000 {built-in method _locale.nl_langinfo}
         1    0.000    0.000   70.975   70.975 {built-in method builtins.exec}
         2    0.000    0.000    0.000    0.000 {built-in method builtins.len}
         1    0.000    0.000    0.000    0.000 {built-in method builtins.print}
         1    0.000    0.000    0.000    0.000 {built-in method io.open}
         1    0.000    0.000    0.000    0.000 {method '__exit__' of '_io._IOBase' objects}
         2    0.000    0.000    0.000    0.000 {method 'append' of 'list' objects}
    102364   14.094    0.000   14.094    0.000 {method 'decode' of 'bytes' objects}
         1    0.000    0.000    0.000    0.000 {method 'disable' of '_lsprof.Profiler' objects}
     51182   33.599    0.001   33.599    0.001 {method 'encode' of 'str' objects}
     51183    0.043    0.000    0.043    0.000 {method 'find' of 'str' objects}
         1    0.001    0.001    0.003    0.003 {method 'read' of '_io.TextIOWrapper' objects}

So this PR improves that by

• Reducing the excessive number of calls to .encode() and .decode().
• Avoiding creating new strings by slicing on every return.

Profiling after the changes:

FUNCTIONS SORTED ON SELF TIME
count  total (s)   self (s)  function
   10   3.374407   3.205914  provider#python3#Call()
    8              3.203132  <SNR>292_decode_via_python()

         326315 function calls (255229 primitive calls) in 0.140 seconds

   Ordered by: standard name

   ncalls  tottime  percall  cumtime  percall filename:lineno(function)
        1    0.000    0.000    0.000    0.000 _bootlocale.py:33(getpreferredencoding)
        1    0.000    0.000    0.000    0.000 bencode.py:16(__decode_integer)
        1    0.000    0.000    0.000    0.000 bencode.py:26(__decode_list)
        1    0.001    0.001    0.140    0.140 bencode.py:3(<module>)
    51182    0.049    0.000    0.073    0.000 bencode.py:3(__decode_string)
   9952/1    0.027    0.000    0.135    0.135 bencode.py:43(__decode_dict)
  61136/1    0.028    0.000    0.135    0.135 bencode.py:59(__decode)
        1    0.000    0.000    0.136    0.136 bencode.py:71(iced_bencode_decode)
        1    0.000    0.000    0.000    0.000 codecs.py:260(__init__)
        1    0.000    0.000    0.000    0.000 codecs.py:309(__init__)
        1    0.000    0.000    0.001    0.001 codecs.py:319(decode)
        1    0.001    0.001    0.001    0.001 {built-in method _codecs.utf_8_decode}
        1    0.000    0.000    0.000    0.000 {built-in method _locale.nl_langinfo}
   101657    0.008    0.000    0.008    0.000 {built-in method builtins.chr}
        1    0.000    0.000    0.140    0.140 {built-in method builtins.exec}
        4    0.000    0.000    0.000    0.000 {built-in method builtins.len}
        1    0.000    0.000    0.000    0.000 {built-in method builtins.print}
        1    0.000    0.000    0.000    0.000 {built-in method io.open}
        1    0.000    0.000    0.000    0.000 {method '__exit__' of '_io._IOBase' objects}
        2    0.000    0.000    0.000    0.000 {method 'append' of 'list' objects}
    51182    0.011    0.000    0.011    0.000 {method 'decode' of 'bytes' objects}
        1    0.000    0.000    0.000    0.000 {method 'disable' of '_lsprof.Profiler' objects}
        1    0.001    0.001    0.001    0.001 {method 'encode' of 'str' objects}
    51183    0.012    0.000    0.012    0.000 {method 'find' of 'bytes' objects}
        1    0.001    0.001    0.003    0.003 {method 'read' of '_io.TextIOWrapper' objects}

which is ~500x improvement. The code does look a little bit harder to deal with than before, but I still hope this can be useful to iced.

[liquidz]

@Cnly Thanks!
I'll have a look tomorrow.

When I :IcedConnect to it, it sent back a 1,201,722-byte bencoded string

The cause may be track-state middleware.
#170

which is ~500x improvement.

Wow, awesome!

[liquidz]

@Cnly Thanks for your contribution!
Overall the PR looks good.
Could you add only tests I commented on?

[Cnly]

Not sure if the version problem is something we want to fix in this PR. Maybe not?

[liquidz]

@Cnly Great work! Thanks!

[liquidz]

Not sure if the version problem is something we want to fix in this PR. Maybe not?

This may result in a solution, but the Vim script version is still slow and not fixed.
https://github.com/liquidz/vim-iced/blob/300552623923cbda4918f3b7c1bea41dae63008a/autoload/iced/component/bencode/vim.vim

Anyway, thank you so much for your work!