On Oct 20, 2009, at 9:17 PM, mdipierro wrote: > > I noticed that: > >>>> 5.1 > 5.0999999999999996 >>>> "%(data)s" % dict(data=5.1) > '5.1' >>>> "%(data)e" % dict(data=5.1) > '5.100000e+00' >>>> "%(data)f" % dict(data=5.1) > '5.100000' >>>> "%(data).3f" % dict(data=5.1) > '5.100' >>>> "%(data).3e" % dict(data=5.1) > '5.100e+00' > > So when the number is printed it is printed fine. The internal > representation is what it is.
It's routine floating point precision issues, I think. The current patch has doctests for everything that changed significantly. But I think I'll add tests for at least some of the other validations. I don't have to to make them thorough, but it'd be nice to have something as a starting point. > > Massimo > > On Oct 20, 11:07 pm, Jonathan Lundell <[email protected]> wrote: >> On Oct 20, 2009, at 8:32 PM, mdipierro wrote: >> >> >> >>> I am ok with the patch. Perhaps for date and datetime we could also >>> have a %(format)s that inserts the actual format string after some >>> beautification, but I can add that later. >> >> I've created some of the doctests. One anomaly (not surprising, >> really) is this one: >> >> >>> IS_FLOAT_IN_RANGE(1,5)(5.1) >> (5.0999999999999996, 'enter a number between 1.0 and 5.0') >> >> Do we care? >> >> >> >>> massimo >> >>> On Oct 20, 9:25 pm, Jonathan Lundell <[email protected]> wrote: >>>> On Oct 20, 2009, at 5:24 PM, mdipierro wrote: >> >>>>> please email it to me when done. Thanks. >> >>>> Are you OK with the patches? I don't want to do the doctests until >>>> they're settled, since the error messages figure in the tests. >> >>>>> On Oct 20, 6:42 pm, Jonathan Lundell <[email protected]> wrote: >>>>>> On Oct 20, 2009, at 2:17 PM, mdipierro wrote: >> >>>>>>> OK. I will take the patch. >> >>>>>> Here's a first cut:http://lobitos.net/web2py-patches/gluon/ >>>>>> validators.py >> >>>>>> This is *not* ready to go; once we agree on the patch, I'll >>>>>> update >>>>>> the >>>>>> doctest strings accordingly. >> >>>>>>> On Oct 20, 3:38 pm, Jonathan Lundell <[email protected]> wrote: >>>>>>>> On Oct 20, 2009, at 12:04 PM, Yarko Tymciurak wrote: >> >>>>>>>>> "integer" is fine and appropriate, and not at all "techie" - >>>>>>>>> if >>>>>>>>> you >>>>>>>>> went to grade school in the US, you would understand "integer" >>>>>>>>> perfectly. I expect the same would be true of any translation >>>>>>>>> of >>>>>>>>> the same. >> >>>>>>>> I don't have a strong feeling on the question, though I'm not >>>>>>>> sure >>>>>>>> that my mother could define 'integer', and she graduated from >>>>>>>> high >>>>>>>> school (quite some time ago). FWIW, 'whole number' is perfectly >>>>>>>> good >>>>>>>> English (and lower grade level than 'integer'; notice that it's >>>>>>>> the >>>>>>>> dictionary definition of 'integer'). >> >>>>>>>> More generally, I was looking at the validation error messages, >>>>>>>> and >>>>>>>> have some suggestions. I'll generate a patch if you like. >>>>>>>> Here's >>>>>>>> the >>>>>>>> (sorted) list: >> >>>>>>>> cannot be empty! >>>>>>>> input a number between %(min)s and %(max)s >>>>>>>> input an integer between %(min)s and %(max)s >>>>>>>> invalid email! >>>>>>>> invalid expression! >>>>>>>> invalid expression! >>>>>>>> invalid filename or extension! >>>>>>>> invalid image! >>>>>>>> invalid IPv4 address! >>>>>>>> invalid length! >>>>>>>> invalid url! >>>>>>>> invalid url! >>>>>>>> invalid url! >>>>>>>> must be alphanumeric! >>>>>>>> must be HH:MM:SS! >>>>>>>> must be YYYY-MM-DD HH:MM:SS! >>>>>>>> must be YYYY-MM-DD! >>>>>>>> value already in database! >>>>>>>> value not allowed! >>>>>>>> value not in database! >> >>>>>>>> Comments: >> >>>>>>>> 1. If it were up to me, I'd remove all the exclamation marks; >>>>>>>> there's >>>>>>>> no need to be excited about a mis-entry; we're just providing >>>>>>>> information. >> >>>>>>>> 2. IS_LENGTH should perhaps give the use the same range info >>>>>>>> that >>>>>>>> the >>>>>>>> number-range messages do. >> >>>>>>>> 3. 'enter a number' rather than 'input a number'. >> >>>>>>>> 4. In general, 'enter <something>' is more helpful and friendly >>>>>>>> than >>>>>>>> 'invalid <whatever>'. It's not possible in every case, but in >>>>>>>> many >>>>>>>> cases we could do it. Example: 'enter a valid email address'. >> >>>>>>>>> On Tue, Oct 20, 2009 at 12:59 PM, mdipierro >>>>>>>>> <[email protected]> wrote: >> >>>>>>>>> You may be right but to me "whole number" does not sound good. >>>>>>>>> Anyway, >>>>>>>>> one can use internationalization to fix this. >> >>>>>>>>> I wish like there are "en-en", "en-us", there were "en-us- >>>>>>>>> tech" >>>>>>>>> and >>>>>>>>> "en-us-non-tech". (Perhaps I should patent this!) >> >>>>>>>>> Massimo >> >>>>>>>>> On Oct 20, 12:38 pm, Jonathan Lundell <[email protected]> >>>>>>>>> wrote: >>>>>>>>>> On Oct 20, 2009, at 10:34 AM, mdipierro wrote: >> >>>>>>>>>>> ok. >> >>>>>>>>>> My intuitive sense is that more users are likely to >>>>>>>>>> understand >>>>>>>>> "whole >>>>>>>>>> number" than "integer". We all take the meaning of integer >>>>>>>>>> for >>>>>>>>>> granted, but I wonder whether it isn't a little on the >>>>>>>>>> technical >>>>>>>>> side >>>>>>>>>> for my Aunt Polly. >> >>>>>>>>>>> On Oct 20, 12:19 pm, Jonathan Lundell <[email protected]> >>>>>>>>>>> wrote: >>>>>>>>>>>> On Oct 20, 2009, at 10:16 AM, mdipierro wrote: >> >>>>>>>>>>>>> I think "integer number" is the proper way but I am not a >>>>>>>>>>>>> native >>>>>>>>>>>>> english speaker. Any english speaker can help us on this? >> >>>>>>>>>>>> Here's the (Oxford American) dictionary entry: >> >>>>>>>>>>>> integer |ˈɪntədʒər| >>>>>>>>>>>> noun >>>>>>>>>>>> 1 a whole number; a number that is not a fraction. > > --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "web2py-users" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [email protected] For more options, visit this group at http://groups.google.com/group/web2py?hl=en -~----------~----~----~----~------~----~------~--~---
