Update the use of ECHAR and UCHAR in canonical N-Quads. #27
Conversation
gkellogg
added
needs discussion
Co-authored-by: Andy Seaborne <[email protected]>
gkellogg
removed
the
needs discussion
|
Should we have a formal review of this document? We could use the opportunity to run a process in the WG. Such as review needn't affect FPWD - it could be before or soon after. The aim is to (unofficially) get a stable point for RCH. |
|
Sounds like a good general topic for the WG to discuss. Maybe as part of the FPWD discussion. |
spec/index.html
Outdated
| @@ -266,35 +271,18 @@ <h2>A Canonical form of N-Quads</h2> | |||
| MUST NOT use the datatype IRI part of the <a href="#grammar-production-literal">literal</a>, | |||
| and are represented using only <a href="#grammar-production-STRING_LITERAL_QUOTE">STRING_LITERAL_QUOTE</a>. | |||
| </li> | |||
| <!--li><code><a href="#grammar-production-HEX">HEX</a></code> MUST use only uppercase letters (<code>[A-F]</code>).</li--> | |||
| <li>Characters MUST NOT be represented by <code><a href="#grammar-production-UCHAR">UCHAR</a></code>.</li> | |||
| <li><code><a href="#grammar-production-HEX">HEX</a></code> MUST use only uppercase letters (<code>[A-F]</code>).</li> | |||
| <li>Within <a href="#grammar-production-STRING_LITERAL_QUOTE">STRING_LITERAL_QUOTE</a>, | |||
| the characters | |||
| <code>U+0022</code>, <code>U+005C</code>, <code>U+000A</code>, <code>U+000D</code> | |||
There was a problem hiding this comment.
| <code>U+0022</code>, <code>U+005C</code>, <code>U+000A</code>, <code>U+000D</code> | |
| <code>U+0022</code>, <code>U+005C</code>, <code>U+000A</code>, <code>U+000D</code>, | |
| <code>U+0008</code>, <code>U+0009</code>, <code>U+000C</code>, <code>U+0027</code> |
IIRC, it might be a good idea to add here the other four "ECHAR"s, i.e., \b (U+0008), \t (U+0009), \f (U+000C), \' (U+0027), otherwise how to encode them seems currently ambiguous.
There was a problem hiding this comment.
While those characters aren’t included in the requirement for using ECHAR, there is no ambiguity, as the text explicitly says characters between U+0000 and U+001F along with U+007F that are not explicitly encoded using ECHAR must be encoded as UCHAR. All other characters must be represented natively in Unicode. There are requirements elsewhere about \\, but this could be made more explicit. I think \’ should not be represented using ECHAR.
There was a problem hiding this comment.
Thanks, I think I misunderstood the following text:
characters between U+0000 and U+001F along with U+007F that are not explicitly encoded using ECHAR must be encoded as UCHAR
From your comment, I now understand that these four ECHARs, \b (U+0008), \t (U+0009), \f (U+000C), \' (U+0027), should be encoded or represented as follows. Is that correct?
\b(U+0008),\t(U+0009),\f(U+000C) => encoded using UCHAR (i.e.,\u0008,\u0009,\u000C)\'(U+0027) => represented natively in Unicode (i.e.,')
There was a problem hiding this comment.
Probably requires tweaking for the characters you mention, but yes, in that lower range, characters are either ECHAR or UCHAR. Some characters outside that range also must use ECHAR and 007F must use UCHAR. I’ll tweak the wording some more tomorrow.
There was a problem hiding this comment.
Added a commit to clarify this, and use the relevant character abbreviations which were originally intended and in fact in my own implementation.
Co-authored-by: Dan Yamamoto <[email protected]>
Co-authored-by: Ted Thibodeau Jr <[email protected]>
Co-authored-by: Ted Thibodeau Jr <[email protected]>
* Updates canonical N-Triples to be consistent with N-Quads. Fixes #2. * Synchronize changes made in w3c/rdf-n-quads#27 for canonicalization. --------- Co-authored-by: Ted Thibodeau Jr <[email protected]>
Fixes #16.
Previously discussed in the RCH WG https://www.w3.org/2023/03/15-rch-minutes.html#r01.
cc/ @philarcher @yamdan
Preview | Diff