[Expat-discuss] Internal operation of expat

[Karl Waclawek]

Mark wrote:
> Hi All,
>
> I wounder if anyone could clarify something about the internal operation of expat?
>
> When I was trying to speed up the operation of a program that parses  large files I noticed that, in my CharData callback, the address of the data pointer is always one of two values, and does not appear to point into the buffer allocated by XML_GetBuffer().
>
> This suggests to me that the data must be copied internally.
>
> Can someone tell me whether this is correct and, if so, how expat performs so well?
>
>   
Expat has to convert encodings, normalize line-breaks, expand entities, 
etc. So copying is normally required.
Why it performs so well? Because James Clark is a hell of a coder.

Karl