[
Lists Home |
Date Index |
Thread Index
]
No, yesterday I decided to do what I would consider a stupid
Schematron trick. Someone suggested that validating a particular
constraint would be impossible because Schematron is not turing
complete (not ISO schematron in this case, in the case of ISO
schematron I think it depends as to what query language is bound to
it)
I figured it would be possible to do it since the things that
Schematron lacks in Turing completeness is basically looping, and in
this case the actual validation context was not unlimited so looping
was unnecessary.
However to make the assertions needed in Schematron I've figured out
it will require 86 assertions with increasingly longer xpaths (this is
basically splitting a string up by linefeeds and making sure that the
string in between each line feed cannot be longer than 35 characters
and also making sure that one does not have more than 43 lines. One
line feed between lines. 86 assertions because there is a slight
difference in xpaths if the last character in the string is a line
feed or not)
Rather than write the xpaths by hand I figured it made sense to
generate them. I have also encountered this situation before where it
would have made sense to generate xpaths rather than hand write them,
at least if I was going to be doing a lot of them.
For example the following two schematron report elements validate an
EAN location number as being correctly structured:
<sch:report test="@schemeID='EAN' and substring(.,13,1)!=0 and ((((10
- substring((substring(.,1,1) * 1 + substring(.,2,1) * 3) +
(substring(.,3,1) * 1 + substring(.,4,1) * 3) + (substring(.,5,1) * 1
+ substring(.,6,1) * 3) + (substring(.,7,1) * 1 + substring(.,8,1) *
3) + (substring(.,9,1) * 1 + substring(.,10,1) * 3) +
(substring(.,11,1) * 1 + substring(.,12,1) *
3),string-length((substring(.,1,1) * 1 + substring(.,2,1) * 3) +
(substring(.,3,1) * 1 + substring(.,4,1) * 3) + (substring(.,5,1) * 1
+ substring(.,6,1) * 3) + (substring(.,7,1) * 1 + substring(.,8,1) *
3) + (substring(.,9,1) * 1 + substring(.,10,1) * 3) +
(substring(.,11,1) * 1 + substring(.,12,1) * 3)),1)) +
((substring(.,1,1) * 1 + substring(.,2,1) * 3) + (substring(.,3,1) * 1
+ substring(.,4,1) * 3) + (substring(.,5,1) * 1 + substring(.,6,1) *
3) + (substring(.,7,1) * 1 + substring(.,8,1) * 3) + (substring(.,9,1)
* 1 + substring(.,10,1) * 3) + (substring(.,11,1) * 1 +
substring(.,12,1) * 3))) - ((substring(.,1,1) * 1 + substring(.,2,1) *
3) + (substring(.,3,1) * 1 + substring(.,4,1) * 3) + (substring(.,5,1)
* 1 + substring(.,6,1) * 3) + (substring(.,7,1) * 1 + substring(.,8,1)
* 3) + (substring(.,9,1) * 1 + substring(.,10,1) * 3) +
(substring(.,11,1) * 1 + substring(.,12,1) * 3))) != substring(.,13,1)
)">
there is an improperly formatted EAN number.
</sch:report>
<sch:report test="@schemeID='EAN' and substring(.,13,1) =0 and
substring((substring(.,1,1) * 1 + substring(.,2,1) * 3) +
(substring(.,3,1) * 1 + substring(.,4,1) * 3) + (substring(.,5,1) * 1
+ substring(.,6,1) * 3) + (substring(.,7,1) * 1 + substring(.,8,1) *
3) + (substring(.,9,1) * 1 + substring(.,10,1) * 3) +
(substring(.,11,1) * 1 + substring(.,12,1) *
3),string-length((substring(.,1,1) * 1 + substring(.,2,1) * 3) +
(substring(.,3,1) * 1 + substring(.,4,1) * 3) + (substring(.,5,1) * 1
+ substring(.,6,1) * 3) + (substring(.,7,1) * 1 + substring(.,8,1) *
3) + (substring(.,9,1) * 1 + substring(.,10,1) * 3) +
(substring(.,11,1) * 1 + substring(.,12,1) * 3)),1) != 0">
there is an improperly formatted EAN number.
</sch:report>
Somewhat mundane, but if doing a lot of it (for example modulus 10
functions might be useful - have not as yet thought out if it is
doable but I am of the opinion that it is) might make sense to
generate, transform, save xpath expressions in an easier to handle
format. Probably XML is not the optimal way to store this information
either, but for various reasons of being able to talk about it with
other people I will be using XML.
Also as noted these are something on the order of Stupid Schematron/
or stupid Xpath tricks, but what the hell, everybody needs a hobby.
Cheers,
Bryan Rasmussen
On 6/21/06, Tatu Saloranta <cowtowncoder@yahoo.com> wrote:
> --- bryan rasmussen <rasmussen.bryan@gmail.com> wrote:
>
> > Hi, there have been various xml based syntaxes for
> > Xpath over the
> > year. Anyone have some pointers to these, I am
> > basically hoping to
> > store extremely complicated/long xpath expressions
> > and don't see a
> > reason to roll my own (other than IIRC the xml based
> > syntaxes were not
> > pretty).
>
> ... but wont XML-defined representations be, almost by
> definition, even longer and more complicated? ;-D
>
> *duck*
>
> But seriously, are you planning to query these paths
> using xquery; for storage it'd seem that xpath syntax
> itself, stored as text (within xml or whatever) would
> not be that bad a storage choice?
>
> -+ Tatu +-
>
>
> __________________________________________________
> Do You Yahoo!?
> Tired of spam? Yahoo! Mail has the best spam protection around
> http://mail.yahoo.com
>
> -----------------------------------------------------------------
> The xml-dev list is sponsored by XML.org <http://www.xml.org>, an
> initiative of OASIS <http://www.oasis-open.org>
>
> The list archives are at http://lists.xml.org/archives/xml-dev/
>
> To subscribe or unsubscribe from this list use the subscription
> manager: <http://www.oasis-open.org/mlmanage/index.php>
>
>
|
