Calculate on a DATE object in minilang

[Jacopo Cappellato]

Wow,

thanks Adrian.
I've added a comment to the following Jira issue:

https://issues.apache.org/jira/browse/OFBIZ-1249

Of course, there is no rush at all.

Jacopo

Adrian Crum wrote:

> Jacques,
>
> If you go back to the start of this thread, you will see that there is
> an existing Jira issue.
>
> Jacopo,
>
> I would be happy to work on this. It would help if someone could C&P the
> minilang format we discussed into the Jira issue, then I will work on it
> as soon as I'm done with a project at work.
>
> -Adrian
>
> Jacques Le Roux wrote:
>
>> We should 1st create a Jira issue for this, isn'it (maybe there is
>> already one) ? I guess your example is a good base for that using
>> all adjusters Adrian proposed   years="${year}"... millis="${millis}".
>> Adrian proposed that the field "field=" could be optionnal, so the
>> adjusters fields, but there must be at least one of them
>> (including "field=" in them ?) I don't think a such a rule is already
>> used in minilang.
>>
>> Jacques
>>
>>
>> De : "Jacopo Cappellato" <[hidden email]>
>>
>>> It would be nice to have:
>>>
>>> <adjust-date-time field="tomorrowDate" from-field="nowTimestamp"
>>> days="+1"/>
>>> <format-date field="tomorrowName" from-field="tomorrowDate"
>>> format="EEEE"/>
>>> <log level="always" message="Tomorrow is ${tomorrowName}"/>
>>>
>>> Output:
>>>
>>> "Tomorrow is Wednesday"
>>>
>>> Jacopo
>>
>>
>> Jacques
>>
>> De : "Jacopo Cappellato" <[hidden email]>
>>
>>> I'm keeping this thread alive because I have a wish to see the
>>> <set-calendar> (or <adjust-date-time>) operation implemented: in fact in
>>> the Minilang service I've committed yesterday to close a financial time
>>> period we have the requirement to get a Timestamp (older than an
>>> existing one of a few seconds).
>>> I'm wondering if someone here could help with this task... maybe Adrian
>>> "The Lord Of Time" Crum? ;-)
>>>
>>> Jacopo
>>>
>>>
>>> Jacques Le Roux wrote:
>>>
>>>> I just had a cursory review, yes seems like a good idea. I agree
>>>> with Adrian view : rely on joda team as possible. And maybe
>>
>> help
>>
>>>> them also, though joda looks like being already very well advanced.
>>>>
>>>> +1 for me
>>>>
>>>> Jacques
>>>>
>>>> De : "Jim Barrows" <[hidden email]>
>>>>
>>>>> Or, at least provide a better starting point then the java date
>>>>> stuff :)
>>>>>
>>>>> On Dec 10, 2007 10:22 AM, Adrian Crum <[hidden email]> wrote:
>>>>>
>>>>>> Jim,
>>>>>>
>>>>>> That looks very interesting! We already have the UtilDateTime
>>>>>> class that does pretty much the same
>>>>>> thing. I'll take a look at it - there's a chance we can eliminate
>>>>>> a lot of OFBiz code by going with
>>>>>> the joda-time code.
>>>>>>
>>>>>> -Adrian
>>>>>>
>>>>>>
>>>>>> Jim Barrows wrote:
>>>>>>
>>>>>>
>>>>>>> What about wrapping joda time (joda-time.sourceforge.net) up in
>>>>>>> mini-lang?
>>>>>>>
>>>>>>> On Dec 10, 2007 10:04 AM, Chris Howe <[hidden email]> wrote:
>>>>>>>
>>>>>>>
>>>>>>>> Keep in mind there are additional outputs that one might want
>>>>>>>> from date/time math.  Weekday of first of month, week of year,
>>>>
>>>> etc.  I'm lacking moments of inspiration on the subject at the
>>>> moment, so I'll just put that out there for now instead of
>>
>> offering a
>>
>>>> solution :P
>>>>
>>>>>>>> ----- Original Message ----
>>>>>>>> From: Jacopo Cappellato <[hidden email]>
>>>>>>>> To: [hidden email]
>>>>>>>> Sent: Monday, December 10, 2007 10:24:27 AM
>>>>>>>> Subject: Re: Calculate on a DATE object in minilang
>>>>>>>>
>>>>>>>>
>>>>>>>> Adrian,
>>>>>>>>
>>>>>>>> looks pretty good.
>>>>>>>> One small suggestion: use "field" instead of "to-field" to use
>>>>>>>> the same
>>>>>>>>
>>>>>>>> convention of the set operator:
>>>>>>>>
>>>>>>>> <adjust-date-time field="toDate" from-field="fromDate"
>>>>>>>>  years="+1" months="+1" ... millis="+1"
>>>>>>>>  locale="parameters.locale" time-zone="parameters.timeZone"/>
>>>>>>>>
>>>>>>>> Jacopo
>>>>>>>>
>>>>>>>>
>>>>>>>> Adrian Crum wrote:
>>>>>>>>
>>>>>>>>
>>>>>>>>> Thanks Jacques.
>>>>>>>>>
>>>>>>>>> I've been thinking about this a little, and here is my suggestion:
>>>>>>>>>
>>>>>>>>> <adjust-date-time from-field="fromDate" to-field="toDate"
>>>>>>>>> years="+1" months="+1" ... millis="+1"
>>>>>>>>> locale="parameters.locale" time-zone="parameters.timeZone"/>
>>>>>>>>>
>>>>>>>>> The to-field attribute would be optional, as well as any of the
>>>>>>>>> adjusters. An <adjust-date-time> element that has no adjusters
>>>>>>>>
>>>>>>>> produces
>>>>>>>>
>>>>>>>>
>>>>>>>>> an error.
>>>>>>>>>
>>>>>>>>> Adjustments would be performed using the minilang context's locale
>>>>>>>>
>>>>>>>> and
>>>>>>>>
>>>>>>>>
>>>>>>>>> time zone, unless the locale and time-zone attributes are used.
>>>>>>>>>
>>>>>>>>> Any thoughts?
>>>>>>>>>
>>>>>>>>> -Adrian
>>>>>>>>>
>>>>>>>>> Jacques Le Roux wrote:
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>> Thanks Adrian,
>>>>>>>>>>
>>>>>>>>>> I put your comment and Chris's in the issue. Hopefully Fabien
>>>>>>>>>> will
>>>>>>>>>> take care of it (propose somehting), else we will see later...
>>>>>>>>>>
>>>>>>>>>> Jacques
>>>>>>>>>>
>>>>>>>>>> De : "Adrian Crum" <[hidden email]>
>>>>>>>>>>
>>>>>>>>>>> I agree with Chris. I haven't looked at the issue lately, but
>>>>>>>>>>> when
>>>>>>>>
>>>>>>>> I
>>>>>>>>
>>>>>>>>
>>>>>>>>>>> first looked at it, it simply performed millisecond
>>>>>>>>>>
>>>>>>>>>> arithmetic. That might not be the expected behavior.
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>> Date calculation would be a great addition to mini language.
>>>>>>>>>>> Maybe
>>>>>>>>
>>>>>>>> we
>>>>>>>>
>>>>>>>>
>>>>>>>>>>> should discuss how it would look in mini lang code, then an
>>>>>>>>>>
>>>>>>>>>> implementation could be submitted.
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>> -Adrian
>>>>>>>>>>>
>>>>>>>>>>> Jacques Le Roux <[hidden email]> wrote: Thanks
>>>>>>>>>>> Chris,
>>>>>>>>>>>
>>>>>>>>>>> Interesting remark, seems that the disussion begins to roll...
>>>>>>>>>>>
>>>>>>>>>>> Jacques
>>>>>>>>>>>
>>>>>>>>>>> De : "Chris Howe"
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>> This really needs to go through a Calendar object instead of
>>>>>>>>>>>> being
>>>>>>>>>>>> cast to a long.  Also, I don't think the current
>>>>>>>>>>>
>>>>>>>>>>> element structure is best suited for the operations that
>>>>>>>>>>> would be
>>>>>>>>>>> most beneficial for date/time calculations
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>> ----- Original Message ----
>>>>>>>>>>>> From: Jacques Le Roux
>>>>>>>>>>>> To: [hidden email]
>>>>>>>>>>>> Sent: Saturday, December 8, 2007 5:37:08 PM
>>>>>>>>>>>> Subject: Calculate on a DATE object in minilang
>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>> Hi All,
>>>>>>>>>>>>
>>>>>>>>>>>> 3 months ago Fabien Carrion has contribued some interesting
>>>>>>>>
>>>>>>>> patches.
>>>>>>>>
>>>>>>>>
>>>>>>>>>>>> One of them is https://issues.apache.org/jira/browse/OFBIZ-1249
>>>>>>>>>>>> David asked <
>>>>>>>>>>>> do a design and requirements brainstorming discussion on the
>>>>>>>>>>>> dev
>>>>>>>>
>>>>>>>> list
>>>>>>>>
>>>>>>>>
>>>>>>>>>>>> before putting this in.
>>>>>>>>>>>> The main thing I'm wondering is if there are other similar
>>>>>>>>
>>>>>>>> requirements
>>>>>>>>
>>>>>>>>
>>>>>>>>>>>> that could be satisfied by small changes in the design of
>>>>>>>>>>>> this.>>
>>>>>>>>>>>>
>>>>>>>>>>>> I would like to commit this patch now except if we begin to
>>>>>>>>
>>>>>>>> exchange
>>>>>>>>
>>>>>>>>
>>>>>>>>>>>> about new requirements per David suggestion
>>>>>>>>>>>>
>>>>>>>>>>>> Thanks
>>>>>>>>>>>>
>>>>>>>>>>>> Jacques
>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>> ---------------------------------
>>>>>>>>>>> Looking for last minute shopping deals?  Find them fast with
>>>>>>>>>>> Yahoo!
>>>>>>>>>>> Search.
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>
>>>>>
>>>>> --
>>>>> James A Barrows
>>>>>
>>>
>>
>>

[Jacques Le Roux]

Adrian,

Actually on my side the point was more about optionnal but one fields rule. I wanted to underline this as "I don't think such a rule
is already used in minilang."

Jacques

----- Message d'origine -----
De : "Adrian Crum" <[hidden email]>
À : <[hidden email]>
Envoyé : lundi 17 décembre 2007 17:13
Objet : Re: Calculate on a DATE object in minilang

> Jacques,
>
> If you go back to the start of this thread, you will see that there is an existing Jira issue.
>
> Jacopo,
>
> I would be happy to work on this. It would help if someone could C&P the minilang format we
> discussed into the Jira issue, then I will work on it as soon as I'm done with a project at work.
>
> -Adrian
>
> Jacques Le Roux wrote:
>
> > We should 1st create a Jira issue for this, isn'it (maybe there is already one) ? I guess your example is a good base for that

using

> > all adjusters Adrian proposed   years="${year}"... millis="${millis}".
> > Adrian proposed that the field "field=" could be optionnal, so the adjusters fields, but there must be at least one of them
> > (including "field=" in them ?) I don't think a such a rule is already used in minilang.
> >
> > Jacques
> >
> >
> > De : "Jacopo Cappellato" <[hidden email]>
> >
> >>It would be nice to have:
> >>
> >><adjust-date-time field="tomorrowDate" from-field="nowTimestamp"
> >>days="+1"/>
> >><format-date field="tomorrowName" from-field="tomorrowDate" format="EEEE"/>
> >><log level="always" message="Tomorrow is ${tomorrowName}"/>
> >>
> >>Output:
> >>
> >>"Tomorrow is Wednesday"
> >>
> >>Jacopo
> >
> >
> > Jacques
> >
> > De : "Jacopo Cappellato" <[hidden email]>
> >
> >>I'm keeping this thread alive because I have a wish to see the
> >><set-calendar> (or <adjust-date-time>) operation implemented: in fact in
> >>the Minilang service I've committed yesterday to close a financial time
> >>period we have the requirement to get a Timestamp (older than an
> >>existing one of a few seconds).
> >>I'm wondering if someone here could help with this task... maybe Adrian
> >>"The Lord Of Time" Crum? ;-)
> >>
> >>Jacopo
> >>
> >>
> >>Jacques Le Roux wrote:
> >>
> >>>I just had a cursory review, yes seems like a good idea. I agree with Adrian view : rely on joda team as possible. And maybe
> >
> > help
> >
> >>>them also, though joda looks like being already very well advanced.
> >>>
> >>>+1 for me
> >>>
> >>>Jacques
> >>>
> >>>De : "Jim Barrows" <[hidden email]>
> >>>
> >>>>Or, at least provide a better starting point then the java date stuff :)
> >>>>
> >>>>On Dec 10, 2007 10:22 AM, Adrian Crum <[hidden email]> wrote:
> >>>>
> >>>>>Jim,
> >>>>>
> >>>>>That looks very interesting! We already have the UtilDateTime class that does pretty much the same
> >>>>>thing. I'll take a look at it - there's a chance we can eliminate a lot of OFBiz code by going with
> >>>>>the joda-time code.
> >>>>>
> >>>>>-Adrian
> >>>>>
> >>>>>
> >>>>>Jim Barrows wrote:
> >>>>>
> >>>>>
> >>>>>>What about wrapping joda time (joda-time.sourceforge.net) up in mini-lang?
> >>>>>>
> >>>>>>On Dec 10, 2007 10:04 AM, Chris Howe <[hidden email]> wrote:
> >>>>>>
> >>>>>>
> >>>>>>>Keep in mind there are additional outputs that one might want from date/time math.  Weekday of first of month, week of

year,

> >>>
> >>>etc.  I'm lacking moments of inspiration on the subject at the moment, so I'll just put that out there for now instead of
> >
> > offering a
> >
> >>>solution :P
> >>>
> >>>>>>>----- Original Message ----
> >>>>>>>From: Jacopo Cappellato <[hidden email]>
> >>>>>>>To: [hidden email]
> >>>>>>>Sent: Monday, December 10, 2007 10:24:27 AM
> >>>>>>>Subject: Re: Calculate on a DATE object in minilang
> >>>>>>>
> >>>>>>>
> >>>>>>>Adrian,
> >>>>>>>
> >>>>>>>looks pretty good.
> >>>>>>>One small suggestion: use "field" instead of "to-field" to use the same
> >>>>>>>
> >>>>>>>convention of the set operator:
> >>>>>>>
> >>>>>>><adjust-date-time field="toDate" from-field="fromDate"
> >>>>>>>  years="+1" months="+1" ... millis="+1"
> >>>>>>>  locale="parameters.locale" time-zone="parameters.timeZone"/>
> >>>>>>>
> >>>>>>>Jacopo
> >>>>>>>
> >>>>>>>
> >>>>>>> Adrian Crum wrote:
> >>>>>>>
> >>>>>>>
> >>>>>>>>Thanks Jacques.
> >>>>>>>>
> >>>>>>>>I've been thinking about this a little, and here is my suggestion:
> >>>>>>>>
> >>>>>>>><adjust-date-time from-field="fromDate" to-field="toDate"
> >>>>>>>> years="+1" months="+1" ... millis="+1"
> >>>>>>>> locale="parameters.locale" time-zone="parameters.timeZone"/>
> >>>>>>>>
> >>>>>>>>The to-field attribute would be optional, as well as any of the
> >>>>>>>>adjusters. An <adjust-date-time> element that has no adjusters
> >>>>>>>
> >>>>>>>produces
> >>>>>>>
> >>>>>>>
> >>>>>>>>an error.
> >>>>>>>>
> >>>>>>>>Adjustments would be performed using the minilang context's locale
> >>>>>>>
> >>>>>>>and
> >>>>>>>
> >>>>>>>
> >>>>>>>>time zone, unless the locale and time-zone attributes are used.
> >>>>>>>>
> >>>>>>>>Any thoughts?
> >>>>>>>>
> >>>>>>>>-Adrian
> >>>>>>>>
> >>>>>>>>Jacques Le Roux wrote:
> >>>>>>>>
> >>>>>>>>
> >>>>>>>>
> >>>>>>>>>Thanks Adrian,
> >>>>>>>>>
> >>>>>>>>>I put your comment and Chris's in the issue. Hopefully Fabien will
> >>>>>>>>>take care of it (propose somehting), else we will see later...
> >>>>>>>>>
> >>>>>>>>>Jacques
> >>>>>>>>>
> >>>>>>>>>De : "Adrian Crum" <[hidden email]>
> >>>>>>>>>
> >>>>>>>>>>I agree with Chris. I haven't looked at the issue lately, but when
> >>>>>>>
> >>>>>>>I
> >>>>>>>
> >>>>>>>
> >>>>>>>>>>first looked at it, it simply performed millisecond
> >>>>>>>>>
> >>>>>>>>>arithmetic. That might not be the expected behavior.
> >>>>>>>>>
> >>>>>>>>>
> >>>>>>>>>
> >>>>>>>>>>Date calculation would be a great addition to mini language. Maybe
> >>>>>>>
> >>>>>>>we
> >>>>>>>
> >>>>>>>
> >>>>>>>>>>should discuss how it would look in mini lang code, then an
> >>>>>>>>>
> >>>>>>>>>implementation could be submitted.
> >>>>>>>>>
> >>>>>>>>>
> >>>>>>>>>
> >>>>>>>>>>-Adrian
> >>>>>>>>>>
> >>>>>>>>>>Jacques Le Roux <[hidden email]> wrote: Thanks Chris,
> >>>>>>>>>>
> >>>>>>>>>>Interesting remark, seems that the disussion begins to roll...
> >>>>>>>>>>
> >>>>>>>>>>Jacques
> >>>>>>>>>>
> >>>>>>>>>>De : "Chris Howe"
> >>>>>>>>>>
> >>>>>>>>>>
> >>>>>>>>>>
> >>>>>>>>>>>This really needs to go through a Calendar object instead of being
> >>>>>>>>>>>cast to a long.  Also, I don't think the current
> >>>>>>>>>>
> >>>>>>>>>>element structure is best suited for the operations that would be
> >>>>>>>>>>most beneficial for date/time calculations
> >>>>>>>>>>
> >>>>>>>>>>
> >>>>>>>>>>
> >>>>>>>>>>>----- Original Message ----
> >>>>>>>>>>>From: Jacques Le Roux
> >>>>>>>>>>>To: [hidden email]
> >>>>>>>>>>>Sent: Saturday, December 8, 2007 5:37:08 PM
> >>>>>>>>>>>Subject: Calculate on a DATE object in minilang
> >>>>>>>>>>>
> >>>>>>>>>>>
> >>>>>>>>>>>Hi All,
> >>>>>>>>>>>
> >>>>>>>>>>>3 months ago Fabien Carrion has contribued some interesting
> >>>>>>>
> >>>>>>>patches.
> >>>>>>>
> >>>>>>>
> >>>>>>>>>>>One of them is https://issues.apache.org/jira/browse/OFBIZ-1249
> >>>>>>>>>>>David asked <
> >>>>>>>>>>>do a design and requirements brainstorming discussion on the dev
> >>>>>>>
> >>>>>>>list
> >>>>>>>
> >>>>>>>
> >>>>>>>>>>>before putting this in.
> >>>>>>>>>>>The main thing I'm wondering is if there are other similar
> >>>>>>>
> >>>>>>>requirements
> >>>>>>>
> >>>>>>>
> >>>>>>>>>>>that could be satisfied by small changes in the design of this.>>
> >>>>>>>>>>>
> >>>>>>>>>>>I would like to commit this patch now except if we begin to
> >>>>>>>
> >>>>>>>exchange
> >>>>>>>
> >>>>>>>
> >>>>>>>>>>>about new requirements per David suggestion
> >>>>>>>>>>>
> >>>>>>>>>>>Thanks
> >>>>>>>>>>>
> >>>>>>>>>>>Jacques
> >>>>>>>>>>>
> >>>>>>>>>>>
> >>>>>>>>>>
> >>>>>>>>>>
> >>>>>>>>>>
> >>>>>>>>>>---------------------------------
> >>>>>>>>>>Looking for last minute shopping deals?  Find them fast with Yahoo!
> >>>>>>>>>>Search.
> >>>>>>>>>
> >>>>>>>>>
> >>>>>>>
> >>>>>>>
> >>>>>>
> >>>>
> >>>>--
> >>>>James A Barrows
> >>>>
> >>
> >
> >
>