Hexadecimal optimizations

[BoboTiG]

What was wrong?

Nothing wrong, simply trying to improve performances here, and there.

How was it fixed?

I used that script to naively benchmark changes for the 2nd commit:

import re
from timeit import timeit

def original(value, regex=re.compile("(0x)?[0-9a-f]*", re.IGNORECASE | re.ASCII)):
    return regex.fullmatch(value) is not None

def new(value, regex=re.compile("(0[xX])?[0-9a-fA-F]*")):
    return regex.fullmatch(value) is not None

print("original(True)", timeit(
    "original('0x5831e66b7de5df2d8d4705c86c844ff60a50efcec53f9449b16dc43216a783f7')", "from __main__ import original"))
print("new(True)", timeit(
    "new('0x5831e66b7de5df2d8d4705c86c844ff60a50efcec53f9449b16dc43216a783f7')", "from __main__ import new"))
print()
print("original(False)", timeit(
    "original('0x5831e66b7de5df2d8d4705c86c844ff60a50efcec53f9449b16dc43216a783f7g')", "from __main__ import original"))
print("new(False)", timeit(
    "new('0x5831e66b7de5df2d8d4705c86c844ff60a50efcec53f9449b16dc43216a783f7g')", "from __main__ import new"))

Improvements unlocked by the first commit:

• When value has 0x prefix:

  • before: 0.1212 sec
  • after : 0.1132 sec (-6.6 %)

• When value has 0X prefix:

  • before: 0.1976 sec
  • after : 0.1167 sec (-40.9 %)

• When value has no 0x prefix:

  • before: 0.1918 sec
  • after : 0.1165 sec (-39.2 %)

Improvements unlocked by the 2nd commit:

• When value is hexadecimal string (TX hash):

  • before: 1.0965 sec
  • after : 0.3652 sec (-66.7 %)

• When value is not hexadecimal string (TX hash + 'g'):

  • before: 1.4546 sec
  • after : 0.7022 sec (-51.7 %)

To-Do

• [x] Clean up commit history

• [x] Add entry to the release notes

Cute Animal Picture

[BoboTiG]

I wasn't sure it deserved a news fragment. Here it is!

[carver]

Thanks for the contribution!