homepage

from io import BytesIO
from tokenize import tokenize, tok_name
sample = 'éléphants = "un éléphant, deux éléphants, ..."\nprint(éléphants)\n'
sampleb = sample.encode('utf-8')
exec(sample)
# output: un éléphant, deux éléphants, ...
exec(sampleb)
# output: un éléphant, deux éléphants, ...
module = BytesIO()
module.write(sampleb)
module.seek(0)
for line in tokenize(module.readline):
 print(tok_name[line.type], line)
# output:
ENCODING TokenInfo(type=57, string='utf-8', start=(0, 0), end=(0, 0), line='')
ERRORTOKEN TokenInfo(type=54, string='é', start=(1, 0), end=(1, 1), line='éléphants = "un éléphant, deux éléphants, ..."\n')
NAME TokenInfo(type=1, string='léphants', start=(1, 1), end=(1, 9), line='éléphants = "un éléphant, deux éléphants, ..."\n')
OP TokenInfo(type=53, string='=', start=(1, 10), end=(1, 11), line='éléphants = "un éléphant, deux éléphants, ..."\n')
STRING TokenInfo(type=3, string='"un éléphant, deux éléphants, ..."', start=(1, 12), end=(1, 46), line='éléphants = "un éléphant, deux éléphants, ..."\n')
NEWLINE TokenInfo(type=4, string='\n', start=(1, 46), end=(1, 47), line='éléphants = "un éléphant, deux éléphants, ..."\n')
NAME TokenInfo(type=1, string='print', start=(2, 0), end=(2, 5), line='print(éléphants)\n')
OP TokenInfo(type=53, string='(', start=(2, 5), end=(2, 6), line='print(éléphants)\n')
ERRORTOKEN TokenInfo(type=54, string='é', start=(2, 6), end=(2, 7), line='print(éléphants)\n')
NAME TokenInfo(type=1, string='léphants', start=(2, 7), end=(2, 15), line='print(éléphants)\n')
OP TokenInfo(type=53, string=')', start=(2, 15), end=(2, 16), line='print(éléphants)\n')
NEWLINE TokenInfo(type=4, string='\n', start=(2, 16), end=(2, 17), line='print(éléphants)\n')
ENDMARKER TokenInfo(type=0, string='', start=(3, 0), end=(3, 0), line='')

r84364 

This doesn't seem to be a complete fix; the regex used does not include Other_ID_Start or Other_ID_Continue from
https://docs.python.org/3.5/reference/lexical_analysis.html#identifiers
Hence tokenize does not accept '℘·'.
Credit to modchan from http://stackoverflow.com/a/29586366/1763356.

Joshua opened #24194 as a duplicate of this because he could not reopen this. I am leaving it open as the superseder for this as Serhiy has already added two dependencies there, and because this seems to be a duplicate in turn of #1693050 (which I will close along with #32987).

Actually, #1693050 and #12731, about \w, are duplicates.