Path: csiph.com!fu-berlin.de!uni-berlin.de!individual.net!not-for-mail
From: Arno Welzel <usenet@arnowelzel.de>
Newsgroups: comp.lang.php
Subject: Re: Check for military time
Date: Mon, 2 Nov 2015 18:57:23 +0100
Lines: 98
Message-ID: <5637A403.5030200@arnowelzel.de>
References: <d8nh4uFgdr2U1@mid.individual.net> <h5tmfc-678.ln1@none.inv> <1874827.QbCUPLQBcH@PointedEars.de> <n0ard5$937$1@dont-email.me> <562F77B7.1050001@arnowelzel.de> <n0ntrp$kpq$1@dont-email.me> <ut1v2b5bkvm5378ldg5ucojj4r1k8o8vnf@4ax.com> <n0o2as$69i$1@dont-email.me> <8737wvd78w.fsf@ahungry.com> <n0pfa8$ckm$1@dont-email.me> <5631e20e$0$23743$e4fe514c@news.xs4all.nl> <87y4elb05f.fsf@ahungry.com> <56375c5b$0$23788$e4fe514c@news.xs4all.nl>
Mime-Version: 1.0
Content-Type: text/plain; charset=windows-1252
Content-Transfer-Encoding: 7bit
X-Trace: individual.net InMuhDWSumXLjyKbx6/lZQ5wJ4H8D/kOVTg+2sjSL1v5V1Ihhm
Cancel-Lock: sha1:/KSSqZzyqlxYsZ33rstiex77+Lk=
In-Reply-To: <56375c5b$0$23788$e4fe514c@news.xs4all.nl>
Xref: csiph.com comp.lang.php:15839

Erwin Moller schrieb am 2015-11-02 um 13:51:

> On 10/30/2015 2:44 AM, Matthew Carter wrote:
[...]
>> Tell me, in these two sample files, which is easier to debug if the
>> aforementioned use cases occurred that cause an error when the file is
>> included by another file making use of header responses:
>>
>> ------------------File One (no closing tag, misplaced HTML)-----------
>> <?php
>>
>> function add($a, $b) {
>>      return $a + $b;
>> }
>>
>> <h1>This is my great page of addition!</h1>
>>
>> ------------------File Two (closing tag, extra space)-----------------
>> <?php
>>
>> function add($a, $b) {
>>      return $a + $b;
>> }
>>
>> ?>
> 
> 
> People who include files and put whitespaces after the closing tag will 
> quickly see the infamous "Output created before headers bla bla, etc 
> etc." error.

This will just happen, if a scripts wants to add headers which is not
always the case.


> Hmmm, HUGE PROBLEM! Now what?
> 
> We have a few choices:
> 1) Remove the offending whitespace, and tell the programmer NOT to do 
> that again.

And hope that he will obey this rule. More likely this will happen again
many times.

> 2) Change the language, violate the logical assumption that each opening 
> tag should be accompanied by a closing tag.
> 
> 
> It is beyond my understanding how anyone can opt for the second option.

You will do this, when you first had to spend hours in finding bugs in a
system which consist of hundreds of files and some problem shows up
caused by unintentional whitespaces.

> What is even more flabberghasting is the fact many people are willing to 
> defend that crazy decision.

As already mentioned by T. Lahn:

<http://www.php-fig.org/psr/psr-2/>

Cite:

"The closing ?> tag MUST be omitted from files containing only PHP."


[...]
> And if some programmer REALLY has trouble avoiding whitespaces after the 
> closing tag, write some simple bash script, that checks for them.
> (I never needed that, but it is easy to write: just match the last ?> 
> and see if anything follows)

Before I create a script to check other scripts I would just omit the
closing tag - then I don't need any "check scripts" at all ;-).


>> There have been 0 good reasons for including a closing tag (it's just too
>> bad we can't actually have a '.phpdef' file type or something that
>> simulates a 100% PHP only file, without dropping in and out of tags and
>> just avoids the use of the <?php ?> nonsense entirely).
>>
> 
> I fail, and keep failing, to see the importance of this.
> We programmers have puzzles and problems to be solved daily.
> Finding an unintended whitespace isn't really a big part of my daily 
> business......
> 
> This solves a problem that didn't have to be solved, and it made the 
> language uglier.

PHP is already ugly in many places.


-- 
Arno Welzel
http://arnowelzel.de
http://de-rec-fahrrad.de
http://fahrradzukunft.de