Re: Most efficient way to recognize a line by its first (up to) three characters

[meino.cramer@...]

Enrico Colombini <erix@erix.it> [14-03-11 19:44]:
> On 11/03/2014 19.14, meino.cramer@gmx.de wrote:
> >There are either a one character or a two character or a three
> >character code at the beginning of each line, which tells me about
> >the kind of data to expect after that code.
> 
> P.S. In that case I adapt my previous suggestion: you could use 
> string.byte(s) to get the code of the first character, then (if needed) 
> string.byte(s, 2) for the second and string.byte (s, 3) for the third 
> one.
> 
> Also, using a local will save table accesses:
> 
>   local charcode = string.byte
> 
> I am currently squeezing out the last CPU cycles from an 8-bit MCU, so 
> I am in the right mood for optimization, even if probably unneeded ;-)
> 
> -- 
>   Enrico
> 

I know this...it is called 'the deep hack mode'...either
one knows it...or not. It is far beyond what can be explained.
Or in other words... 

"There is a difference between knowing the path & walking the path." - Morpheus

;)