On Tue, 22 Nov 2016 19:40:30 +0000, Benedikt Ritter wrote:
Gary Gregory <garydgreg...@gmail.com> schrieb am Sa., 19. Nov. 2016
um
19:09 Uhr:
On Nov 19, 2016 9:50 AM, "Gilles" <gil...@harfang.homelinux.org>
wrote:
>
> On Sat, 19 Nov 2016 08:59:50 -0800, Gary Gregory wrote:
>>
>> On Sat, Nov 19, 2016 at 3:33 AM, Benedikt Ritter
<brit...@apache.org>
wrote:
>>
>>> Hello Gray,
>>>
>>> Gary Gregory <garydgreg...@gmail.com> schrieb am Sa., 19. Nov.
2016 um
>>> 01:07 Uhr:
>>>
>>> > Just a thought:
>>> >
>>> > Does all the current (and future) string escaping code (XML,
HTML,
...)
>>> > really belong in [lang]? Would it be more natural to have it
in
[text]?
>>> >
>>>
>>> My view on the whole think currently is, that we put stuff that
is
related
>>> to strings in Lang. Code that works on texts should go to Text.
To me a
>>> text is more than just a string. A text contains works, that
make up
>>> sentences, which in turn build paragraphs.
>>>
>>> Using this description, I'd argue that escaping belongs into
lang and
not
>>> into text, because it works on individual characters rather than
on
texts.
>>>
>>> But this would also raise the question if the various edit
distance
>>> algorithms works on texts or on strings. So maybe my distinction
is not
>>> good at all.
>>>
>>> Do we need to better specify the scope of text?
>>>
>>
>> Great question of course.
>>
>> I'd like to think of [lang] as "What is missing from the JRE's
most
basic
>> classes and specifically from the java.lang package and some
java.util
>> classes".
>>
>> Quoting from our site:
>>
>> "The standard Java libraries fail to provide enough methods for
>> manipulation of its core classes. Apache Commons Lang provides
these
extra
>> methods.
>> Lang provides a host of helper utilities for the java.lang API,
notably
>> String manipulation methods, basic numerical methods, object
reflection,
>> concurrency, creation and serialization and System properties.
Additionally
>> it contains basic enhancements to java.util.Date
>
>
> How about "Date" becoming a nice standalone component? ;-)
> [Components should be concept-based.]
Joda-time covers more than we will ever do here IMO. And Java 8 has
new
time APIs... maybe when lang is Java 8 based we can look again. For
now I'd
rather leave dates as is.
Yes, let's get back to topic.
I think we need a clear distinction between string related stuff that
goes
into Lang and more complex stuff that goes into text.
IMHO "more complex" is key, not so much "string" vs "text".
Hence I suggest [text] is a better place for "RandomStringUtils"
than [lang], and the former should allow dependencies as a
foundation for that complexity; in that case, that would be
"Commons RNG".
Regards,
Gilles
Gary
>
> How about deprecating "RandomUtils"?
> [(About to be) superseded by "Commons RNG".]
>
> How about to
> * moving "RandomStringUtils" to [text] too and
> * implement it against a custom interface (as per Jochen's
remark)
> rather than "java.util.Random"
> ?
>
>
>> and a series of utilities
>> dedicated to help with building methods, such as hashCode,
toString and
>> equals."
>>
>> I do not think edit distances fit into this at all.
>
>
> +1
>
>
>> I am also questioning whether string escaping belongs in lang as
well
since
>> there are so many escaping domains XML, HTML, JSON, and so on.
>
>
> They don't belong.
>
>
>> IMO, anything that is word based does not belong in lang like
>> capitlization. The WordUtils class should be in [text] IMO. The
whole
lang
>> text package should be in [text] IMO.
>
>
> +1
> [To anything that imposes a strict diet on the humongous
"components".]
>
>
> Regards,
> Gilles
>
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org
For additional commands, e-mail: dev-h...@commons.apache.org
