> > > Then they're simply wrong. The fundamental operation is
> > > to check if a string is in a regular set of strings.
> > > Plainly 'aa' is not in the set { 'a' }.
> >
> > This is a strange notion of right and wrong.  The function behaves exactly
> > as it is specified in the documentation.
> 
> Huh? I'm really confused. Two people think the documentation
> is correct for the behaviour.
> 
> I do not -- Saying
> 
>        string x is matched by regexp r
> 
> means to me the same as
> 
>        x is a member of the regular set of strings r denotes
> 
> and with that interpretation there is no possible doubt that
> the documentation is wrong and should say
> 
>        matches a prefix of the argument string
> 
> in order to describe the behaviour.
> 

Going by PCRE's documentation the regex /a/ matches: 'a', 'aa',
'baaaaa', 'nota bene', and any other string with 'a' anywhere inside.

>From OCaml's description of string_match:

val string_match: regexp -> string -> int -> bool
[string_match r s start] tests whether the characters in s starting at
position start match the regular expression r. The first character of
a string has position 0, as usual.

This sounds to me like you're doing the PCRE: /a/

if you say:

Str.string_match (Str.regexp "a") "ab" 0 ;;

And in PCRE /a/ matches "ab" or "aa" or "ba".

/^a/ matches "aa", "ab", but not "ba", and /a$/ matches "aa", "ba",
but not "ab". So as far as I can tell, OCaml's string_match is correct
as the manual describes.

-- 
Christopher A. Watford
christopher.watford@gmail.com
http://dorm.tunkeymicket.com