Peter Hickman
1/10/2006 3:36:00 PM
Christer Nilsson wrote:
>There must be a better way of converting dates than this ugly mess ?
>
>def convert_date format, s
> case format
> when "yymmddhhmmss"
> year = s[0,2].to_i + 2000
> month= s[2,2].to_i
> day = s[4,2].to_i
> hour = s[6,2].to_i
> minute=s[8,2].to_i
> second=s[10,2].to_i
> Time.local(year,month,day,hour,minute,second)
> when "yyyymmddhhmm"
> year = s[0,4].to_i
> month= s[4,2].to_i
> day = s[6,2].to_i
> hour = s[8,2].to_i
> minute=s[10,2].to_i
> Time.local(year,month,day,hour,minute)
> when "ddmmyyhhmm"
> day = s[0,2].to_i
> month= s[2,2].to_i
> year = s[4,2].to_i + 2000
> hour = s[6,2].to_i
> minute=s[8,2].to_i
> Time.local(year,month,day,hour,minute)
> when "yyyymmdd"
> year = s[0,4].to_i
> month= s[4,2].to_i
> day = s[6,2].to_i
> Time.local(year,month,day)
> end
>end
>
>Christer
>
>
>
I'm not too sure about the + 2000 part of the year calculation. It seems
to be missing the pivot.
The real problem is that there are so many different date formats and
none of them are *sensible*. You could write some fsm to parse the
string of "ymdhms" symbols and give yourself a pat on the back for your
coding skills or you could just use the code and get on with something
more interesting. As a point of note there is the question of the year
values, there should be an else clause to raise an error and there is no
checking that the values are sensible (or even numbers), does
format.size == s.size etc.
Yes to code looks ugly but I would be more concerned about the possible
errors than it not being /nice/.
Sometimes ugly is the best you can get.