Re: Recursion in RDF Data Shape Languages


Thanks for the feedback. Yes, it's dangerous to over-generalize. My
point is that if a language lets you name things and refer to those
things by name then you can end up with circular references. This
might lead to ill-defined semantics. I'll revise the text.

-- Arthur

On Wed, May 20, 2015 at 4:00 PM, Peter F. Patel-Schneider
<> wrote:
> Hash: SHA256
> I'm rather surprised at the statement there that "[r]ecursion depends on the
> ability to name things and to refer to those things by name".   I don't
> think that this is true in general - if you have loops in program constructs
> then you can get recursion and you don't need to have names for bits of a
> data structure to turn it into a loop.
> peter
> On 05/20/2015 05:35 AM, Arthur Ryman wrote:
>> I just posted an article that formalizes the kind of recursion that
>> Resource Shape 2.0 supports. [1] It therefore also covers a subset of the
>> kind of recursion that SHACL(Holger) and SHACL(ShEx) supports. I believe
>> this kind of recursion could be added to SHACL-SPARQL(Peter).
>> The article is language-neutral in the sense that it does restrict how
>> constraints are specified. It uses Z Notation, but I've attempted to keep
>> it as simple as possible, and I've illustrated every definition with a
>> running example.
>> In the interest of having a stable version, I've posted it on arXiv. I'd
>> appreciate feedback and will include corrections or improvements in the
>> next version. Thanks.
>> [1]
> Version: GnuPG v2
> iElKPb8KvXjiWdXgbT5TMoSH4VtG8ItRj8JeT2wawb38SWsnYvzh9Leeoqw5LxGv
> MQvZQHJFY5kHtWr7H/cD5y7zNtV2F3lsmyQQJzTXukyK7JBNP/LHuMytzwD+YoqW
> OFHQyQ71YI0KNWuYSiZjrwLI2zh1aTwL09vp7Mn+VbVqAG3Lm6Nnw9zJjoJVlIcl
> bHUau6x21QpVrZdTQs7pG8iCMcxd5mpHtpk8Sac/D6CdHGPVt6QpZtq2chiwMLxr
> zeY01uwSJ4XT3lfPXCG3xbmF6jxueUBhN3BW0zaa8Pdm+9Ldve+fgMCDeKxLilo=
> =H/si

Received on Wednesday, 20 May 2015 20:09:57 UTC