How to parse an email address
Suppose I want to model an email address: dfdl-wg@ogf.org It's a sequence of a string, an a character (@), and a string. In the MBTK it might look like this: [cid:image001.png@01CE10EC.6600AE50] The problem, of course, is that the parser will grab all of "dfdl-wg@ogf.org" instead of just "dfdl-wg". Setting validation facets doesn't help, b/c validation happens after parsing. Setting @ as a separator or terminator would work, but that doesn't seem right either, as the @ is part of the email address, and I would want to keep it in the infoset. How do I think about this problem? TIA
If you want the infoset to contain dfdl-wg@ogf.org, but also to contain
localPart and domain, then you are putting the same data into multiple
fields, which can only be accomplished using inputValueCalc.
I suggest parse the data into localPart and domain by using the @ as
terminator for localPart, and whatever the boundary is for the domain.
then have another element which you inputValueCalc which contains the
concatenation of those fields with an '@' in the middle.
I choose this way, parse the pieces, compute the concatenation of them,
versus the other way round because the expression for the inputValueCalc
will be dead simple in this case, and you will get an ordinary error of
'delimiter not found' if the @ is missing from the data.
Contrast if you had one field which contains the whole email address and
two calculated elements which take substrings of that. Now you have two
inputValueCalc's. And if the @ isn't there in the string, you'll get errors
evaluating expressions, not ordinary parsing, etc.
On Fri, Feb 22, 2013 at 11:11 AM, Garriss Jr., James P.
Suppose I want to model an email address:****
** **
dfdl-wg@ogf.org****
** **
It’s a sequence of a string, an a character (@), and a string.****
** **
In the MBTK it might look like this:****
** **
****
** **
The problem, of course, is that the parser will grab all of “ dfdl-wg@ogf.org” instead of just “dfdl-wg”. Setting validation facets doesn’t help, b/c validation happens after parsing. Setting @ as a separator or terminator would work, but that doesn’t seem right either, as the @ is part of the email address, and I would want to keep it in the infoset.****
** **
How do I think about this problem?****
** **
TIA****
-- dfdl-wg mailing list dfdl-wg@ogf.org https://www.ogf.org/mailman/listinfo/dfdl-wg
-- Mike Beckerle | OGF DFDL Workgroup Co-Chair | Tresys Technology | www.tresys.com
Or you can use dfdl:lengthKind 'pattern' and supply a dfdl:lengthPattern
for localPart that consumes all data up to but not including the @.
Note: Neither dfdl:lengthKind 'pattern' or dfdl:inputValueCalc are
supported by IBM DFDL yet - this will be addressed in the near future :)
Regards
Steve Hanson
Architect, Data Format Description Language (DFDL)
Co-Chair, OGF DFDL Working Group
IBM SWG, Hursley, UK
smh@uk.ibm.com
tel:+44-1962-815848
From: Mike Beckerle
Other option would be to put @ as separator on sequence.. the sequence
contains 2 elements of type string localPart and domain
Suman Kalia
IBM Canada Lab
WMB Toolkit Architect and Development Lead
Tel: 905-413-3923 T/L 313-3923
Email: kalia@ca.ibm.com
For info on Message broker
http://www.ibm.com/developerworks/websphere/zones/businessintegration/wmb.ht...
From: Steve Hanson
Right but still need dfdl:inputValueCalc to get the @ into the infoset, as
per Mike's suggestion.
Regards
Steve Hanson
Architect, Data Format Description Language (DFDL)
Co-Chair, OGF DFDL Working Group
IBM SWG, Hursley, UK
smh@uk.ibm.com
tel:+44-1962-815848
From: Suman Kalia
participants (4)
-
Garriss Jr., James P. -
Mike Beckerle -
Steve Hanson -
Suman Kalia