Javascript and working with dates

[rhaazy]

I need to write some javascript that will return a date string in the
form mm/dd/yyyy.

The date needs to be today's date - 30 days.

Is there a relatively straight forward way to do this?

So far all I can find is a mess of variables for month, day, and year,
and combining some date functions together, etc etc. Seems like a lot
of work for what should be very simple.

[Dr_KralNOSPAM]

On Sun, 21 Sep 2008 04:02:36 +0200, Lasse Reichstein Nielsen
<lr********@gma il.comwrote in <r6**********@g mail.com>:

>Dr***********@ nyc.rr.com writes:

>Let JS do all the complex logic of various month lengths, leap years and
year boundaries.

Agree ...

>var past = new Date(new Date().getTime( ) - 30*24*60*60000)

which is why this is not the way to do it. You are asking for a point
in time that is 30 * 24 hours earlier. That's not always the same as
30 days. Some days are 25 hours, some are 23, due to daylight saving
transition. This code will fail to give the correct day in some cases.

An interesting point. This error will occur if and only if one includes a
23 or 25 hour day AND is in a boundry period of the first or last hour of
the day. That can happen only two hours a year.

One can correct for this by shifting the current time to noon:
var past = new Date( new Date().setHours (12) - 30*24*60*60000 )

The OP did not indicate if the question was 30 24hour days or 30 calendar
days.

>Use:
var past = new Date();
past.setDate(pa st.getDate()-30);
That gives the same time, 30 days earlier.

NO!

That will work only if the previous month is 30 days (some are 28, 29, 31).
So it is correct only in three months - June, October and December. It
fails in most of May because in the USA DST happens in April.

BTW, if you start in January, then you will go forward some 11 months
unless you adjust the year also but even then the December day of month
will be one low.

[Dr_KralNOSPAM]

please ignore this post -- shifting the date does work

sorry

[rhaazy]

On Sep 20, 11:25*pm, Dr_KralNOS...@n yc.rr.com wrote:

please ignore this post -- shifting the date does work

sorry

Thank you everyone for the discussion. This code is a lot cleaner
than what I had, and I learned a lot about the various date functions.

[Thomas 'PointedEars' Lahn]

sasuke wrote:

Thomas 'PointedEars' Lahn wrote:

Date.prototype. fmt = function(format ) {
var d = this;
return format.replace(/%([CdFfHMmnRSTYy%])/g,
function(m, p1) {
[...]
});
};

Very good modifications. But won't it be better if we use a cached
copy of the RegExp native object instead of creating a new one on each
invocation?

It would, if what you describe were the case. With a RegExp literal
no new object is created, but an already existing object is being
referred to. So there is no need for optimization through "caching":

,-[ECMAScript Language Specification, Edition 3 Final]
|
| 7 Lexical Conventions
|
| The source text of an ECMAScript program is first converted into
| a sequence of input elements, which are either tokens, line
terminators,
| comments, or white space. The source text is scanned from left to
right,
| repeatedly taking the longest possible sequence of characters as
| the next input element.
|
| [...]
|
| 7.8.5 Regular Expression Literals
|
| A regular expression literal is an input element that is converted
to
| a RegExp object (section 15.10) when it is scanned. The object is
| created before evaluation of the containing program or function
begins.
| Evaluation of the literal produces a reference to that object; it
does
| not create a new object.
|
| [...]
| Semantics
|
| A regular expression literal stands for a value of the Object type.
| This value is determined in two steps: first, the characters
comprising
| the regular expression's RegularExpressi onBody and
RegularExpressi onFlags
| production expansions are collected uninterpreted into two strings
Pattern
| and Flags, respectively. Then the new RegExp constructor is called
with
| two arguments Pattern and Flags and the result becomes the value of
the
| RegularExpressi onLiteral. If the call to new RegExp generates an
error,
| an implementation may, at its discretion, either report the error
| immediately while scanning the program, or it may defer the error
until
| the regular expression literal is evaluated in the course of program
| execution.
PointedEars

[Thomas 'PointedEars' Lahn]

[supersedes GG-miswrapped version]

sasuke wrote:

Thomas 'PointedEars' Lahn wrote:

Date.prototype. fmt = function(format ) {
var d = this;
return format.replace(/%([CdFfHMmnRSTYy%])/g,
function(m, p1) {
[...]
});
};

Very good modifications.

Thanks.

But won't it be better if we use a cached copy of the RegExp
native object instead of creating a new one on each invocation?

It would, if what you describe were the case. With a RegExp literal
no new object is created, but an already existing object is being
referred to. So there is no need for optimization through "caching":

,-[ECMAScript Language Specification, Edition 3 Final]
|
| 7 Lexical Conventions
|
| The source text of an ECMAScript program is first converted into
| a sequence of input elements, which are either tokens, line
| terminators, comments, or white space. The source text is scanned
| from left to right, repeatedly taking the longest possible sequence
| of characters as the next input element.
|
| [...]
|
| 7.8.5 Regular Expression Literals
|
| A regular expression literal is an input element that is converted
| to a RegExp object (section 15.10) when it is scanned. The object
| is created before evaluation of the containing program or function
| begins. Evaluation of the literal produces a reference to that
| object; it does not create a new object.
|
| [...]
| Semantics
|
| A regular expression literal stands for a value of the Object type.
| This value is determined in two steps: first, the characters
| comprising the regular expression's RegularExpressi onBody and
| RegularExpressi onFlags production expansions are collected
| uninterpreted into two strings Pattern and Flags, respectively.
| Then the new RegExp constructor is called with two arguments
| Pattern and Flags and the result becomes the value of the
| RegularExpressi onLiteral. If the call to new RegExp generates an
| error, an implementation may, at its discretion, either report
| the error immediately while scanning the program, or it may defer
| the error until the regular expression literal is evaluated in
| the course of program execution.
PointedEars

[sasuke]

On Sep 22, 7:15*pm, "Thomas 'PointedEars' Lahn" <PointedE...@we b.de>
wrote:

[supersedes GG-miswrapped version]

sasuke wrote:

Thomas 'PointedEars' Lahn wrote:

Date.prototype. fmt = function(format ) {
* var d = this;
* return format.replace(/%([CdFfHMmnRSTYy%])/g,
* * function(m, p1) {
* * * [...]
* * });
};

Very good modifications.

Thanks.

You are most welcome.

But won't it be better if we use a cached copy of the RegExp
native object instead of creating a new one on each invocation?

It would, if what you describe were the case. *With a RegExp literal
no new object is created, but an already existing object is being
referred to. *So there is no need for optimization through "caching":

,-[ECMAScript Language Specification, Edition 3 Final]

[enlightening piece of text]

My bad; I did search the specification document for words like
'cache', 'cached' etc. in terms of Regular expressions but to no
avail. I should have known better than trying to use implementation
dependent terms as search keywords. I have always used global RegExp
literals for two reasons: to prevent new RegExp object creation and
for the sake of maintainability ; I guess I need to strike out the
former.

It would be interesting to note the behavior of the implementation in
memory limited environment when faced with a extremely large number of
RegExp literals; would it then go in favor of implementing LFU/LRU
policy or prefer to throw an OOM (out of memory) error. Just a thought
though.

/sasuke

[Dr J R Stockton]

In comp.lang.javas cript message <48************ *********@news. orange.fr>
, Sat, 20 Sep 2008 00:34:29, SAM <st************ *********@wanad oo.fr.inv
alidposted:

>
Example with a trap over February:
============== =============== =====
var y = 2005, m = 03, d = 30; // 30/03/2005 (in dd/mm/yyyy)
var D = new Date(y, +m-1, d); // the orginal JS date
D.setDate(D.ge tDate()-30); // 30 days less
// that is done
alert( d+'/'+m+'/'+y+
' = original date\n'+
D.getDate()+'/'+(+D.getMonth( )+1)+'/'+D.getFullYear ()+
' = date - 30 days'):

As new Date() is not used, then it is not a "real-time" question. For
such, it is more efficient, and in some cases may avoid obscure error,
to work entirely in UTC.

var D = new Date(Date.UTC(y , m-1, d - 30))
// then output D using UTC methods.

If there are still browsers that don't like non-positive day-of-month,
then move the 30 to between the parentheses, multiplying it by 864e5.

--
(c) John Stockton, nr London, UK. ?@merlyn.demon. co.uk Turnpike v6.05 MIME.
Web <URL:http://www.merlyn.demo n.co.uk/- FAQish topics, acronyms, & links.

Usenet News services are currently unreliable; I may not see all articles here.

[Dr J R Stockton]

In comp.lang.javas cript message <gb************ *@news.t-online.com>,
Fri, 19 Sep 2008 23:44:11, Stevo <[e-mail] dposted:

>rhaazy wrote:

>On Sep 19, 3:04 pm, SAM <stephanemoriau x.NoAd...@wanad oo.fr.invalid>
wrote:

>>var D = new Date();
D.setMonth(D. getMonth()-1)
alert( D.getDate()+'/'+(+D.getMonth( )+1)+'/'+D.getFullYear () );
--19/8/2008

Thats fine if all I want is a date, but to be able to subtract 30 days
from any given date introduces a huge mess of logic that needs to be
coded.

I can't recall any language that makes dealing with dates easy.

Icelandic Javascript. UTC all year round, unless they join the EU.
VBScript, for those few who are Americans or similar AND have discovered
DateSerial(y,m, d).

Opera is 9.60.

--
(c) John Stockton, nr London, UK. ?@merlyn.demon. co.uk Turnpike v6.05 MIME.
Web <URL:http://www.merlyn.demo n.co.uk/- FAQish topics, acronyms, & links.

Usenet News services are currently unreliable; I may not see all articles here.