RE: [Freemarker-devel] Hexadecimal formatting

[Mark v. de V. <mar...@se...>]

Hi,

> [snip] recently there was a guy who wanted a 
> "remove_tags" built-in (to remove HTML tags from text coming 
> from the DB), but then people here in the list voted against 
> it, because it's too special. So maybe this will be the case 
> with the "hex" built-in as well. But personally I would 
> accept it (as well as "remove_tags"). Now, of course I would 
> vote against a fibonacci number calculator built-in... 
Imho my "hex" build-in is really just a formatting thing and formatting
(view generation in my case) is what Freemarker is all about. It does
not calculate anything (like a fibonacci would), doesn't change data or
affect anything else. If you acknowledge that Freemarker can be used in
more technical applications then a "hex" build-in seems pretty generic.
But that's just me. :)

> There is no such documentation... but what info do you miss 
> regarding that?
You're right, probably not much. I thought I saw other (big) open source
projects have some pointers on what a wanne-be-developer should do.
Things like: please look at this or that mailing list, learn about this
or that before thinking of contributing or mail this person when you
want to contribute. Maybe just a single line somewhere in the manual
would do fine if it says something like "If you want to contribute
please join the fre...@li... mailing list".

> > below. It's not a complicated build-in. I'm sure you guys(?)
> (Really... are there women?)
Just covering my *ss. ;)

In response to Attila:
> Well - it's mostly okay, except for two things :-)
> 1. If it's a BigInteger, then BigInteger.toString(int radix) 
> is the API to use
> 2. Otherwise, num.longValue() is a better choice to support 64-bit 
> integers, too.
You're right. I would go for the second option. I always use
Integer.toHexString() and forgot that Long.toHexString() exists too...
I'm not a very experienced Java developer but I read some place
(http://mindprod.com/jgloss/gotchas.html) that BigInteger isn't a very
nice class. Is it used much? It sounds to me that having 'limited'
support for BigInteger isn't a big deal.

> Also, I'm not sure if we want to include support for padding 
> zeroes...
I understand. It's just that in more technical environments hexadecimal
values very often printed as 2, 4 or 8 digit numbers. Why not
incorporate it if the feature will be used most of the time? If only for
performance (I have a embedded software back-ground). It really is kind
of like ${12345?string("0.####E0")} and that is supported. Come and
think if it, if it is possible to extend the ?string build-in to support
hexadecimal formatting of integers then that could be a good solution
too.


Mark