Changes to synchronize with N-Triples. #107
Conversation
gkellogg
added
the
spec:substantive
…l`, which is the correct term to reference.
spec/index.html
Outdated
| </thead> | ||
| <tbody> | ||
| <tr id="handle-IRIREF" ><td style="text-align:left;" ><a class="type IRI" href="#grammar-production-IRIREF" >IRIREF </a></td><td><a data-cite="RDF12-CONCEPTS#dfn-iri" >IRI </a></td><td>The characters between <a href="#cp-less-than"><code title="less-than sign"><</code></a> and <a href="#cp-greater-than"><code title="greater-than sign">></code></a> are taken, with the <a data-lt="numeric escape sequence">numeric escape sequences</a> unescaped, to form the IRI. <a data-cite="RDF12-CONCEPTS#dfn-relative-iri">Relative IRI reference</a> resolution is performed per <a href="#sec-iri-references" class="sectionRef">Section 6.3</a>.</td></tr> | ||
| <tr id="handle-IRIREF" ><td style="text-align:left;" ><a class="type IRI" href="#grammar-production-IRIREF" >IRIREF </a></td><td><a data-cite="RDF12-CONCEPTS#dfn-iri" >IRI </a></td><td>The characters between <a href="#cp-less-than"><code title="less-than sign"><</code></a> and <a href="#cp-greater-than"><code title="greater-than sign">></code></a> are taken, with the <a data-lt="numeric escape sequence">numeric escape sequences</a> unescaped, to form the IRI. <a data-cite="RDF12-CONCEPTS#dfn-relative-iri">Relative IRI reference</a> resolution is performed per <a href="#sec-iri-references" class="sectionRef">Section 6.3</a>. The resulting IRI MUST comply with the <a data-cite="RFC3987#section-2.2">syntactic restrictions</a> of generic IRI syntax, and SHOULD comply with any narrower restrictions imposed by the corresponding IRI scheme specification.</td></tr> |
There was a problem hiding this comment.
and SHOULD conform to section 3.3 of RFC3986
"""
The path segments "." and "..", also known as dot-segments, are
defined for relative reference within the path name hierarchy. They
are intended for use at the beginning of a relative-path reference
"""
which is advice at the level of a specific scheme except it applies to all URIs with apath.
There was a problem hiding this comment.
In Turtle, do you think that wording would be repeated anywhere an IRI can be formed? (E.g., PNAME_NS/LN).
There was a problem hiding this comment.
If the NS is valid, can "NS:LN" be undesirable?
It might need PN_LOCAL_ESC to do that.
There was a problem hiding this comment.
I don't quite get how PN_LOCAL_ESC might play into this. Either PNAME_NS or PNAME_LN can be used in the PrefixedName production, which results in an IRI that should have the same provisions as IRIREF.
We could potentially add a new "iri" term contsructor and put the requirements for the form of the resulting IRI there.
You might have a legitimate PNAME_NS IRI but when combined with PNAME_LN still not result in a valid IRI.
Co-authored-by: Andy Seaborne <andy@apache.org>
…ange how whitespace is introduced between `--`.
Fixes #37. Fixes #108.
Preview | Diff