Regular Expressions

[Scott Wolfington]

Thanks Florent! I really appreciate your help.

Regards,
Scott




------------------

[Florent Heyworth]

Hi Scott

PB regular expression don't support escaped characters in
character classes. The "-" sign is a range operator so that
placing it in the middle of a character class as in:

[a-z0-9.-~]

is interpreted as "all alphanumeric with characters ranging from
"." to "~"" which will give incorrect results.

As you found out, placing the "-" last means that the expression
is correctly evaluated. Character class meta-characters placed first
or last in a character class are usually evaluated as literals
since they otherwise make no sense.

So:

[a-z0-9.~-]

and

[-a-z0-9.~]

will give equivalent results.

Cheers

Florent




[This message has been edited by Florent Heyworth (edited September 07, 2000).]

[Scott Wolfington]

Hey all! I'm searching through text data (using regular expressions)
and searching for SMTP error codes. I've commented some code below
to highlight what I'm experiencing. I figure posting the code with
comments would be easier than explaining without the code. Any help
would be HIGHLY appreciated.
Thanks! -Scott

$COMPILE EXE


FUNCTION PBMAIN() AS LONG

    LOCAL strMessage     AS STRING
    LOCAL a              AS STRING
    LOCAL b              AS STRING

    
    
    'NOTE:  This a$ does not work.  Note position of - symbol.
    'I'm assuming this doesn't work because it's being interpreted
    'as a range between the _ and ~.
    'a$ = "[a-z0-9._-~][email protected][a-z0-9._-~]+"


    'NOTE: How come this doesn't work?  I'm escaping the - symbol.
    'Is it okay to escape characters within a character class?  Seems
    'like it *should* work.
    'a$ = "[a-z0-9._\-~][email protected][a-z0-9._\-~]+"
    
    'NOTE: Moved the minus sign to the end of each character class
    'and everything seems to work fine.
    a$ = "[a-z0-9._~-][email protected][a-z0-9._~-]+"
    
    'NOTE: Do I need to escape the . in a character class?  Doesn't
    'seem to make a difference.
    'a$ = "[a-z0-9\._~-][email protected][a-z0-9\._~-]+"


    
    
    b$ = "550 5\.1\.1 <" & a$ & "> is NOT a valid mailbox"

    strMessage$ = "Diagnostic-Code: smtp;550 5.1.1 <[email protected]> is not a valid mailbox"
    REGEXPR b$ IN strMessage$ TO lPosition&, llength&

    MSGBOX STR$(lPosition&) & ":" & strMessage$
   
END FUNCTION

------------------




[This message has been edited by Scott Wolfington (edited September 06, 2000).]