Path: csiph.com!eternal-september.org!feeder.eternal-september.org!mx02.eternal-september.org!.POSTED!not-for-mail
From: Steven Simpson <ss@my.domain.invalid>
Newsgroups: comp.lang.java.programmer
Subject: Re: hashtables syntax
Date: Mon, 16 May 2016 15:51:59 +0100
Organization: A noiseless patient Spider
Lines: 46
Message-ID: <fd3p0d-s27.ln1@s.simpson148.btinternet.com>
References: <nh954m$bu8$1@reader1.panix.com> <nh9n7r$ira$1@dont-email.me> <338d0b8c-9a5e-ccc0-9a5c-792c870f84bf@mrbrklyn.com> <nhcbrv$7qk$1@dont-email.me>
Mime-Version: 1.0
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Transfer-Encoding: 7bit
Injection-Info: mx02.eternal-september.org; posting-host="0f928dddd1117028fbe6efb26c856d65"; logging-data="18257"; mail-complaints-to="abuse@eternal-september.org";	posting-account="U2FsdGVkX1/Bfjmyr7oTwSqBtuCCEW73lURlA+IS1Pc="
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:38.0) Gecko/20100101 Thunderbird/38.7.2
In-Reply-To: <nhcbrv$7qk$1@dont-email.me>
Cancel-Lock: sha1:qPF2Yl/prdtc6wLIaS3aCuCbv+g=
Xref: csiph.com comp.lang.java.programmer:34304

On 16/05/16 12:48, Eric Sosman wrote:
> On 5/16/2016 1:26 AM, ruben safir wrote:
>> Gotchya, but it seems inconsistent.
>>
>>
>> why not like
>>
>> HashMap<> function = new HashMap<String, Type[]>
>
>     With what is this "inconsistent?"  I'm not seeing your point.
>
>     The types must be specified on one side or the other; Java
> decided to insist that they appear on the left.  Java *could*
> have chosen to have them on the right (I suppose, but there may
> be considerations I'm unaware of), but it chose the left.  Shrug.

It would arguably be inappropriate to put the type information 'on the 
right' in the the OP's case, where the declaration of the reference is 
separate from its assignment.  The generic parameters would be visible 
only within the static initializer, while the declaration is visible 
throughout.  Imagine:

    private static HashMap<> functions;

     static { // initialize function types
         functions = new HashMap<String, Type[]>();

Compile-time checking of calls like functions.get(20) would have to 
inspect the static initializer, rather than the declaration.  Also, I 
expect it would be more complex to inspect situations like this:

    private static HashMap<> functions;

    // at some point
    functions = new HashMap<String, Type[]>();

    // later
    functions = new HashMap<Integer, String>();

The declaration acts as a kind of rendezvous point between users and 
providers of 'function', so it makes sense to propagate type information 
out from that point.

-- 
ss at comp dot lancs dot ac dot uk