What is the view point of the JDK team about string interpolation

[Behrang Saeedzadeh]

Hi Neal,

We already have two expression languages (JSP's and JSF's) that finally got
integrated into the Unified EL. Even though they are used outside Java
classes, I am not sure what would be the benefit of yet another way for
String interpolation in the Java sphere. A subset of the Unified EL might
be a better option (e.g. leading to more consistency across the whole
platform).

To quote Jan L. A. van de Snepscheut (probably a misattribution):

*In theory there is no difference between theory and practice; in practice
there is.*


In practice, AFAIK, using the following syntaxes:

   - "foo #{...} bar"
   - "foo ${...} bar"
   - "foo \(...) bar"

Have not caused any real-world problems in languages like Ruby,
JavaScript/Groovy, and Swift. Also IMHO aesthetically they are more
beautiful. So, I personally prefer any of them to $"..." which reminds me
of Objective-C's @"..." that I never liked.

Anyhow, just my 2 cents.

*P.S:* IMHO, it would be nice if the Java language and its related projects
had uservoice accounts similar to Visual Studio:
https://visualstudio.uservoice.com/forums/121579-visual-studio-2015/category/30931-languages-c

Best regards,
Behrang Saeedzadeh






On 11 March 2016 at 12:31, Neal Gafter <neal at gafter.com> wrote:

> We like string interpolation in C#
> <https://msdn.microsoft.com/en-us/library/Dn961160.aspx>. I think it
> would be a nice addition to Java, presuming its benefit/cost compares
> favorably to other things under consideration. Java tends to be more
> conservative about adding complexity to the language than C#, so it is
> natural to find differences like this.
>
> We looked at two syntax options: first *"Fixed stuff \{expression} etc"*
> where you put an expression inside a string between curly braces prefixed
> by a backslash. That has the advantage that you can just start typing it
> into the middle of an existing string. It has the disadvantage that it
> silently turns a *constant* expression into a *non-constant* expression,
> and it is a little more work than the other option because of the backslash.
>
> Our second option was *$"Fixed stuff {expression} etc"*, and that is what
> we ended up going with.
>
> Either way it would be defined by translation into a call to
> *String.format*. You would probably want to add syntax to support format
> specifiers, and a way of deferring the formatting so that, for example, it
> can be localized.
>
> The nice thing about this is that common problems that would cause
> *IllegalFormatException* with *String.format* ether cannot occur or
> become compile-time errors.
>
>
> On Fri, Dec 19, 2014 at 8:40 AM, Brian Goetz <brian.goetz at oracle.com>
> wrote:
>
>> Have you met String.format()?
>>
>> No direct language support is being considered for the forseeable future.
>>
>>
>>
>> On 12/15/2014 9:44 PM, Behrang Saeedzadeh wrote:
>>
>>> Hi,
>>>
>>> Does the JDK team have any view points about adding string interpolation
>>> to
>>> the Java language?
>>>
>>> Best regards,
>>> Behrang
>>> http://www.behrang.org
>>>
>>>
>