generator runout

[Skaperen]

this can get annoying:

Output:
>>> regions_fns = (f'{pre}{my}{aws}{ec2}region{s}{ext}'  \
...               for my in ('my_','my','')              \
...               for aws in ('aws_','aws','')           \
...               for ec2 in ('ec2_','ec2','')           \
...               for pre in ('','.','.config/')         \
...               for s in ('s','')                      \
...               for ext in ('.list','.txt','.conf','') \
... )
>>> len([x for x in regions_fns])
648
>>> len([x for x in regions_fns])
0
>>>

[buran]

this can get annoying:

Not sure if there is question here, but generators are exhausted after you iterate over them once.

Also why not use itertools.product to shorten the comprehension?

[perfringo]

this can get annoying

I wouldn't call it 'annoying' i would call it 'silent'. There is no error raised when using exhausted iterator or generator only strange behavior and missing values.

There are many ways to mitigate this situation. One of them is to define function which returns new generator/iterator every time called.

[Skaperen]

it's not an error. it's the way things a supposed to be. it was just annoying to me because i thought it was better to use a generator like this wherever i could. today i ran into an issue because i needed to count the number of iterations a generator had and i didn't think about it being exhausted. i just changed it to a list.

show me what my line-wrapped comprehension would look like with itertools.product. i think the way i have now (changed to be a list) looks clear and easy to read. my imagination of itertools.product is not like that.

[buran]

how me what my line-wrapped comprehension would look like with itertools.product.

from itertools import product
my_product = product(('my_','my',''), ('aws_','aws',''), ('ec2_','ec2',''),
                     ('','.','.config/'), ('s',''), ('.list','.txt','.conf',''))
regions_fns = (f'{pre}{my}{aws}{ec2}region{s}{ext}' for my, aws, ec2, pre, s, ext in my_product)
# or with str.format()
# regions_fns = ('{3}{0}{1}{2}region{4}{5}'.format(*item) for item in my_product)

---

at least don't use line continuation char \. following is valid code

regions_fns = (f'{pre}{my}{aws}{ec2}region{s}{ext}'
               for my in ('my_','my','')
               for aws in ('aws_','aws','')
               for ec2 in ('ec2_','ec2','')
               for pre in ('','.','.config/')
               for s in ('s','')
               for ext in ('.list','.txt','.conf',''))

[DeaD_EyE]

from itertools import tee

iterator1, iterator2 = tee(your_generator())

More efficient is to save the result in a list or tuple.
The function tee keeps track of yielded elements.

[Skaperen]

at least don't use line continuation char \.

i don't always remember when the line continuation char is needed. i use it when unsure, to be sure the code works. i know ... i should just try it and see.