date and fuzzy calculations

[C de-Avillez]

Hello folks.

We got a bug on Ubuntu [1] where the OP (Mike, copied here) proposes
a rework of the fuzzy 'date' calculations; I am copying his last
comment here:

---- start comment ----

"This explanation does not quite square with the behavior. Witness:

          date
        Mon Jan 31 08:51:30 EST 2011

          date --date='this month' +%j
        031

          date --date='next month' +%j
        062

          date --date='next month'
        Thu Mar 3 08:57:01 EST 2011

So it _appears_ that the algorithm is
        next month -> +1 month
        month -> this month's number of days
        this month's number of days -> 31

        Julian date += 31

I find it hard to accept that the given behavior is "correct" even
if it is
"as documented".

So now the question becomes "am I motivated enough to dig into the code,
understand it thoroughly enough to see what _is_ going on, and then do
something about it."

The simplest correct solution would seem to be:

 manipulations of "month" must take place as manipulations of
 the month field as expressed by the date format %m .

 _if_ the result is an illegal date ( i.e. 31 February )
 _and_ the day of month ( %d ) or julian day ( %j ) will be output,
 _then_
                output an illegal date error
    date --date='next month'
  date: invalid date `31 February'
 _else_
                output
                  date --date='next month' +%B
  February

Would that be deemed acceptable behavior by the keepers of /bin/date ?"

---- end comment ----

Please note I closed the Ubuntu bug WONTFIX, since this is well
documented in both the info pages and on the FAQ.

Still, since Mike is willing to code, I considered it more
productive to contact you folks for a better analysis.

Please copy Mike on the replies, I am not sure he subscribes to the ML.

Cheers,

..C..

[1] https://bugs.launchpad.net/bugs/710368

signature.asc
Description: OpenPGP digital signature