Writing a Lexer, regular expressions

Hi there,

I am tasked with writing a lexer, its a college assignment so I wont ask you to write it for me, dont worry :P But I could really use some help on clearing up the concept!... the notes are terrible and the lecturer loves his jargon

As I understand it, lexers parse a body of text and numbers etc, tokenise it and store it in symbol table, is this correct?

In terms of parsing, I should be following a finite state automaton model, perhaps employing a method for each state and and gererate a token in each? and calling the next method depending on the value of a state? Does this make sense? Or should I use a looping switch statement with methods defined in each for nums, chars, etc?

Finally, how are regular expressions involved in the parsing process, if I am coding in a procedural manner, like below, where are they involved, should they be? Are they part of the grammar, and is that involved in my code?

Heres a excerpt from my notes, entitled "Constucing a lexer". I am slightly confused by it as there is no context for th 'number' variable and im not sure how its related to a token. The assignments are also a bit strange to me (bold)

state:=0;
ch:=readchar(inputstream);
loop forever
    {switch(state):
        case 0: if ch=‘0’ or ch=‘1’ or … ch=‘9’
                then {[B]number:=ch-’0’;[/B] state:=1;
                      ch:=readchar(inputstream); break}
                else {unreadchar(inputstream);
                      return false}
        case 1: if ch=‘0’ or ch=‘1’ or … ch=‘9’
                then {[B]number:=number*10+ch-’0’;[/B] state:=1;
                      ch:=readchar(inputstream); break}
                else {unreadchar(inputstream);
                      return true}
    }
}

Sorry if im way off the mark, and am asking too many questions :P but if someone could answer any of them, or even point me to some useful resources regarding lexers (book is also ****e ) that would be great! Cheers for reading.