Hartmut,<div><br></div><div>thanks for the answer. I don&#39;t use semantic actions and my alternatives are only used in the case where I omit some tokens, so there should not be any copies than.</div><div><br></div><div>Really great lib! I enjoy using it.</div>
<div><br></div><div>Ovanes.</div><div><br><div class="gmail_quote">On Thu, Oct 7, 2010 at 3:29 PM, Hartmut Kaiser <span dir="ltr">&lt;<a href="mailto:hartmut.kaiser@gmail.com">hartmut.kaiser@gmail.com</a>&gt;</span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;"><div><div></div><div class="h5">...<br>
<br>
</div></div>The general idea is for all attributes to be directly used to receive the<br>
parsed data.<br>
<br>
Period.<br>
<br>
As you might imagine this is not always true. You have to be aware of a<br>
couple of pitfalls in order to avoid additional copies inside the parser<br>
generated from a Spirit.Qi expression.<br>
<br>
1)  Don&#39;t use semantic actions, these need to instantiate an instance of the<br>
attribute in order to pass it to the function (object) to be called.<br>
2) Alternatives currently are known to create a new instance of the<br>
attribute as well. This is a known issue and we need to come up with a<br>
better solution.<br>
3) Somebody posted examples on the Spirit ML a while ago, demonstrating that<br>
certain constructs cause additional copying, but I&#39;m not able to find that,<br>
currently. All I remember is that it seemed to be possible to overcome those<br>
problems as well. We just have not have the time to look into this more<br>
closely yet.<br>
<br>
HTH<br>
Regards Hartmut<br>
---------------<br>
<a href="http://boost-spirit.com" target="_blank">http://boost-spirit.com</a><br>
<br></blockquote></div></div>