Hi,

If I do

import re
pattern = re.compile(r'(?P<days>\d*)(-?)(?P<hours>\d\d):(?P<minutes>\d\d):(?P<seconds>\d\d)')
s = '104-02:47:06'
match = pattern.search(s)
match_dict = match.groupdict('0')

I get

match_dict
{'days': '104', 'hours': '02', 'minutes': '47', 'seconds': '06'}

However, if the string has no initial part (corresponding to the number of
days), e.g.

s = '02:47:06'
match = pattern.search(s)
match_dict = match.groupdict('0')

I get

match_dict
{'days': '', 'hours': '02', 'minutes': '47', 'seconds': '06'}

I thought that 'days' would default to '0'.

What am I doing wrong?

Cheers,

Loris
--
This signature is currently under construction.

"Loris Bennett" <loris.bennett@fu-berlin.de> writes:
>I thought that 'days' would default to '0'.

It will get the value '0' if (?P<days>\d*) does
/not/ participate in the match.

In your case, it /does/ participate in the match,
\d* matching the empty string.

Try (?P<days>\d+)?.

On Friday, 29 April 2022 at 09:50:08 UTC+2, Loris Bennett wrote:
> Hi,
>
> If I do
>
> import re
> pattern = re.compile(r'(?P<days>\d*)(-?)(?P<hours>\d\d):(?P<minutes>\d\d):(?P<seconds>\d\d)')
> s = '104-02:47:06'
> match = pattern.search(s)
> match_dict = match.groupdict('0')
>
> I get
>
> match_dict
> {'days': '104', 'hours': '02', 'minutes': '47', 'seconds': '06'}
>
> However, if the string has no initial part (corresponding to the number of
> days), e.g.
>
> s = '02:47:06'
> match = pattern.search(s)
> match_dict = match.groupdict('0')
>
> I get
>
> match_dict
> {'days': '', 'hours': '02', 'minutes': '47', 'seconds': '06'}
>
> I thought that 'days' would default to '0'.
>
> What am I doing wrong?

You tell, but it's quite obvious that you (just) run a regex on a string and captures are going to be strings: indeed, '02' is not a number either...

Julio

ram@zedat.fu-berlin.de (Stefan Ram) writes:

> "Loris Bennett" <loris.bennett@fu-berlin.de> writes:
>>I thought that 'days' would default to '0'.
>
> It will get the value '0' if (?P<days>\d*) does
> /not/ participate in the match.
>
> In your case, it /does/ participate in the match,
> \d* matching the empty string.
>
> Try (?P<days>\d+)?.

Ah, thanks. I was misunderstanding the meaning of 'participate'.

Cheers,

Loris

--
This signature is currently under construction.

"Loris Bennett" <loris.bennett@fu-berlin.de> writes:

> ram@zedat.fu-berlin.de (Stefan Ram) writes:
>
>> "Loris Bennett" <loris.bennett@fu-berlin.de> writes:
>>>I thought that 'days' would default to '0'.
>>
>> It will get the value '0' if (?P<days>\d*) does
>> /not/ participate in the match.
>>
>> In your case, it /does/ participate in the match,
>> \d* matching the empty string.
>>
>> Try (?P<days>\d+)?.
>
> Ah, thanks. I was misunderstanding the meaning of 'participate'.

What I actually need is

((?P<days>\d+)(-?))?(?P<hours>\d\d):(?P<minutes>\d\d):(?P<seconds>\d\d)

so that I can match both

99-11:22:33

and

11:22:33

and have 'days' be '0' in the later case.
Thanks for pointing me in the right direction.

Cheers,

Loris

--
This signature is currently under construction.

Julio Di Egidio <julio@diegidio.name> writes:

> On Friday, 29 April 2022 at 09:50:08 UTC+2, Loris Bennett wrote:
>> Hi,
>>
>> If I do
>>
>> import re
>> pattern = re.compile(r'(?P<days>\d*)(-?)(?P<hours>\d\d):(?P<minutes>\d\d):(?P<seconds>\d\d)')
>> s = '104-02:47:06'
>> match = pattern.search(s)
>> match_dict = match.groupdict('0')
>>
>> I get
>>
>> match_dict
>> {'days': '104', 'hours': '02', 'minutes': '47', 'seconds': '06'}
>>
>> However, if the string has no initial part (corresponding to the number of
>> days), e.g.
>>
>> s = '02:47:06'
>> match = pattern.search(s)
>> match_dict = match.groupdict('0')
>>
>> I get
>>
>> match_dict
>> {'days': '', 'hours': '02', 'minutes': '47', 'seconds': '06'}
>>
>> I thought that 'days' would default to '0'.
>>
>> What am I doing wrong?
>
> You tell, but it's quite obvious that you (just) run a regex on a string and captures are going to be strings: indeed, '02' is not a number either...
>
> Julio

I am not sure what you are trying to tell me. I wasn't expecting
anything other than strings. The problem was, as Stefan helped me to
understand, that I misunderstood what 'participating in the match'
means.

Cheers,

Loris

--
This signature is currently under construction.

On Tuesday, 3 May 2022 at 16:44:10 UTC+2, Loris Bennett wrote:
> Julio Di Egidio <ju...@diegidio.name> writes:
> > On Friday, 29 April 2022 at 09:50:08 UTC+2, Loris Bennett wrote:
> >> Hi,
> >>
> >> If I do
> >>
> >> import re
> >> pattern = re.compile(r'(?P<days>\d*)(-?)(?P<hours>\d\d):(?P<minutes>\d\d):(?P<seconds>\d\d)')
> >> s = '104-02:47:06'
> >> match = pattern.search(s)
> >> match_dict = match.groupdict('0')
> >>
> >> I get
> >>
> >> match_dict
> >> {'days': '104', 'hours': '02', 'minutes': '47', 'seconds': '06'}
> >>
> >> However, if the string has no initial part (corresponding to the number of
> >> days), e.g.
> >>
> >> s = '02:47:06'
> >> match = pattern.search(s)
> >> match_dict = match.groupdict('0')
> >>
> >> I get
> >>
> >> match_dict
> >> {'days': '', 'hours': '02', 'minutes': '47', 'seconds': '06'}
> >>
> >> I thought that 'days' would default to '0'.
> >>
> >> What am I doing wrong?
> >
> > You tell, but it's quite obvious that you (just) run a regex on a string and captures are going to be strings: indeed, '02' is not a number either....
>
> I am not sure what you are trying to tell me. I wasn't expecting
> anything other than strings. The problem was, as Stefan helped me to
> understand, that I misunderstood what 'participating in the match'
> means.

Stefan's "fix" makes it fail if there is not at least one digit in that position. And not only I don't se how that is in fact a fix, nor I still can guess where the string '0' is supposed to originate from in case we do not have the days part in the input. Maybe I am overlooking something...?

Julio

On Tuesday, 3 May 2022 at 19:18:30 UTC+2, Julio Di Egidio wrote:
> On Tuesday, 3 May 2022 at 16:44:10 UTC+2, Loris Bennett wrote:
> > Julio Di Egidio <ju...@diegidio.name> writes:
> > > On Friday, 29 April 2022 at 09:50:08 UTC+2, Loris Bennett wrote:
> > >> Hi,
> > >>
> > >> If I do
> > >>
> > >> import re
> > >> pattern = re.compile(r'(?P<days>\d*)(-?)(?P<hours>\d\d):(?P<minutes>\d\d):(?P<seconds>\d\d)')
> > >> s = '104-02:47:06'
> > >> match = pattern.search(s)
> > >> match_dict = match.groupdict('0')
> > >>
> > >> I get
> > >>
> > >> match_dict
> > >> {'days': '104', 'hours': '02', 'minutes': '47', 'seconds': '06'}
> > >>
> > >> However, if the string has no initial part (corresponding to the number of
> > >> days), e.g.
> > >>
> > >> s = '02:47:06'
> > >> match = pattern.search(s)
> > >> match_dict = match.groupdict('0')
> > >>
> > >> I get
> > >>
> > >> match_dict
> > >> {'days': '', 'hours': '02', 'minutes': '47', 'seconds': '06'}
> > >>
> > >> I thought that 'days' would default to '0'.
> > >>
> > >> What am I doing wrong?
> > >
> > > You tell, but it's quite obvious that you (just) run a regex on a string and captures are going to be strings: indeed, '02' is not a number either...
> >
> > I am not sure what you are trying to tell me. I wasn't expecting
> > anything other than strings. The problem was, as Stefan helped me to
> > understand, that I misunderstood what 'participating in the match'
> > means.
> Stefan's "fix" makes it fail if there is not at least one digit in that position. And not only I don't se how that is in fact a fix, nor I still can guess where the string '0' is supposed to originate from in case we do not have the days part in the input. Maybe I am overlooking something...?

Yes, I am missing the meaning of the parameter in the call to groupdict.
<https://docs.python.org/2.7/library/re.html#re.MatchObject.groupdict>

And then Stefan's solution makes sense, as it's the default for the *non-optional* parts that are missing...

Right? Sorry for the confusion.

Julio

On Tuesday, 3 May 2022 at 19:24:14 UTC+2, Julio Di Egidio wrote:
> On Tuesday, 3 May 2022 at 19:18:30 UTC+2, Julio Di Egidio wrote:
> > On Tuesday, 3 May 2022 at 16:44:10 UTC+2, Loris Bennett wrote:
> > > Julio Di Egidio <ju...@diegidio.name> writes:
> > > > On Friday, 29 April 2022 at 09:50:08 UTC+2, Loris Bennett wrote:
> > > >> Hi,
> > > >>
> > > >> If I do
> > > >>
> > > >> import re
> > > >> pattern = re.compile(r'(?P<days>\d*)(-?)(?P<hours>\d\d):(?P<minutes>\d\d):(?P<seconds>\d\d)')
> > > >> s = '104-02:47:06'
> > > >> match = pattern.search(s)
> > > >> match_dict = match.groupdict('0')
> > > >>
> > > >> I get
> > > >>
> > > >> match_dict
> > > >> {'days': '104', 'hours': '02', 'minutes': '47', 'seconds': '06'}
> > > >>
> > > >> However, if the string has no initial part (corresponding to the number of
> > > >> days), e.g.
> > > >>
> > > >> s = '02:47:06'
> > > >> match = pattern.search(s)
> > > >> match_dict = match.groupdict('0')
> > > >>
> > > >> I get
> > > >>
> > > >> match_dict
> > > >> {'days': '', 'hours': '02', 'minutes': '47', 'seconds': '06'}
> > > >>
> > > >> I thought that 'days' would default to '0'.
> > > >>
> > > >> What am I doing wrong?
> > > >
> > > > You tell, but it's quite obvious that you (just) run a regex on a string and captures are going to be strings: indeed, '02' is not a number either...
> > >
> > > I am not sure what you are trying to tell me. I wasn't expecting
> > > anything other than strings. The problem was, as Stefan helped me to
> > > understand, that I misunderstood what 'participating in the match'
> > > means.
> > Stefan's "fix" makes it fail if there is not at least one digit in that position. And not only I don't se how that is in fact a fix, nor I still can guess where the string '0' is supposed to originate from in case we do not have the days part in the input. Maybe I am overlooking something...?
> Yes, I am missing the meaning of the parameter in the call to groupdict.
> <https://docs.python.org/2.7/library/re.html#re.MatchObject.groupdict>
>
> And then Stefan's solution makes sense, as it's the default for the *non-optional* parts that are missing...
>
> Right? Sorry for the confusion.

P.S. I don't see Stefan's posts, and I doubt he sees mine...

Julio