An expansion's first set is simply the set of all tokens that can begin the expansion.
The most trivial example imaginable is any expansion that begins with a single token. Thus, the first set of:
"foo" "bar" "baz"
is simply a set with one element, “foo”.
The first set of
("foo" | "bar") "baz"
has two elements, “foo” and “bar”. The above expansion must necessarily begin with one of those two tokens.
For example, if the production
Foobar is defined as:
Foobar : "foo" | "bar";
then the first set of the expansion:
[ Foobar ] "baz"
contains the three tokens, “foo”, “bar” , and “baz. To say that a grammar is LL(1) is equivalent to saying that, at any choice point the expansion's first set is enough to decide whether or not to enter the expansion.