On 2017-05-18 05:22, Mark Nottingham wrote:
On 17 May 2017, at 8:41 pm, Julian Reschke <julian.reschke@xxxxxx> wrote:
On 2017-05-16 04:37, Mark Nottingham wrote:
...
This is what I currently have:
https://gist.github.com/mnot/2fb569e7303dbcdde8b27cb7a404a648
I need to get some more eyeballs onto it to make sure it's correct (or at least reasonable). Reviews appreciated.
...
I think this goes into the right direction; it's mostly a parser based on a relaxed ABNF transformed to prose (which makes me wonder whether there's a tool to be written here :-).
Other comments:
1) "... If it is not present, context_string is the identity of the representation carrying the Link header ..." - similar prose is elsewhere in the spec - what is "the identity of a representation"?????
See the linked reference. Would it be better to use "URL" instead of "identity"?
URL (or URI).
2) "Consume the contents up to but not including the first DQUOTE character that is not preceded by a "" character" - I assume there's a backslash missing here. That said: doesn't work, because it would fail for an input such as: "\\" - here, the DQUOTE is preceded by backslash, but it's not part of the quoted-pair.
I've updated to address this.
Goes into the right direction, but doesn't seem to be correct yet:
4. While input has content:
i. If the first character is a backslash ("\"):
1. Discard the first character.
2. If there is no more input, return output.
3. Else, consume the first character and append it to output.
ii. Else, if the first character is DQUOTE, discard it and return output.
There seems to be missing a step between (i) and (ii) for non-quoted-pairs.
Best regards, Julian