Re: Help with Flex input matching

[wlestes]

On Wednesday, 13 April 2005,11:39 +0200, Rubén Béjar wrote:

> Thanks for answering. I was just trying to understand the input match 
> process in flex, so
> I don't really need to read such problematic strings. It's just that the 
> flex manual doesn't tell too much
> about the kind of lexical analyzer used, it just *suggests* it is using 
> a DFA in the bibliography section,
> and the manual neither is very clear, at least to me, in the input 
> matching process description. It seems
> most people are *assuming* certain behavior based mostly in 
> intuition/experience, and I would have
> expected the manual to be more comprehensive on that issue. But at least 
> now, I already know what
> to expect when matching some complex inputs.

A flex scanner will match all input or it will fail trying to do
so. The example you gave was clear in so far as your presumption
failed to match all input even though the rules you gave were
sufficient to do so. A flex scanner will match input in stream order.

--Will