rspec/rules/S1952/flex/rule.adoc

== Why is this an issue?

It can be expensive to instantiate a new object, and doing so inside a loop is typically an error. Instead, create the object once, before the loop.


=== Noncompliant code example

[source,flex]
----
for (var i:int = 0; i < 10; i++) {
  var temp:MyObj = new MyObject();  // Noncompliant
  //...  
}
----

=== Compliant solution

[source,flex]
----
var temp:MyObj = new MyObject();
for (var i:int = 0; i < 10; i++) {
  //...  
}
----

ifdef::env-github,rspecator-view[]

'''
== Implementation Specification
(visible only on this page)

=== Message

Move the instantiation of this "xxx" outside the loop.


'''
== Comments And Links
(visible only on this page)

=== on 7 Apr 2015, 23:24:31 Evgeny Mandrikov wrote:
I'm wondering if this rule is really relevant for C# , Java , C/{cpp}? 

\[~dinesh.bolkensteyn], [~nicolas.peru], [~samuel.mercier], [~massimo.paladin], [~pierre-yves.nicolas], [~linda.martin] any thoughts?

=== on 8 Apr 2015, 05:14:57 Dinesh Bolkensteyn wrote:
I don't think it is [~evgeny.mandrikov]

=== on 9 Apr 2015, 13:00:35 Nicolas Peru wrote:
Neither for java. this one is a deprecation of a pmd rule which was already controversial : \http://stackoverflow.com/questions/17340421/pmd-avoid-instantiating-new-objects-inside-loops 

So, [~evgeny.mandrikov] I agree with [~dinesh.bolkensteyn] I don't think it is worth to implement in Java.

=== on 29 Mar 2016, 23:48:20 Evgeny Mandrikov wrote:
This case is mentioned in "Google {cpp} Style Guide", so CPP-1448.


endif::env-github,rspecator-view[]