Fix links, remove as many anchor definitions as possible #1270
Conversation
|
Oh hm, it's failing due to a badly-encoded URL, which I think is a Bikeshed bug I noticed before. I'll go investigate. |
|
Nah not a Bikeshed bug, it's a data source bug. Reported at w3c/webref#889 |
index.bs
Outdated
| spec: ecma-262; type: dfn; for: /; text: internal method | ||
| spec: ecma-262; type: dfn; for: /; text: internal slot | ||
| spec: ecma-262; type: dfn; for: /; text:realm | ||
| spec:wasm-js-api-2; |
There was a problem hiding this comment.
No variant without a version? cc @Ms2ger
This is something that needs to be fixed as this gives us two nearly identical references.
There was a problem hiding this comment.
Happy to do whatever you tell me here :)
Note WebAssembly/spec#1611 , btw, which is why this xref is listed explicitly here.
There was a problem hiding this comment.
Yeah, it'll be fixed separately, but in the meantime this link-default fixes the error locally.
| [=ECMA-262 Ordinary object internal methods and internal slots=], and if the | ||
| object is a [=function object=], [=ECMA-262 Built-in function objects=]. | ||
| [[ecmascript#sec-ordinary-object-internal-methods-and-internal-slots]], and if the | ||
| object is a [=function object=], [[ecmascript#sec-built-in-function-objects]]. |
index.bs
Outdated
| @@ -7655,7 +7587,7 @@ values are represented by ECMAScript Object values. | |||
| An ECMAScript value |V| is [=converted to an IDL value|converted=] | |||
| to an IDL {{object}} value by running the following algorithm: | |||
|
|
|||
| 1. If <a abstract-op>Type</a>(|V|) is not Object, then [=ECMAScript/throw=] a {{ECMAScript/TypeError}}. | |||
| 1. If <a abstract-op>Type</a>(|V|) is not Object, then [=ECMAScript/throw=] a {{TypeError}}. | |||
There was a problem hiding this comment.
This is another name clash between ES and Web IDL. Previously it linked to ECMAScript, but now it goes to Web IDL. It's mostly fine since Web IDL's {{TypeError}} is basically a redirect to ES. But since these algorithm steps are written in the style of ECMAScript (including the throw reference), I'd still prefer to use the ECMAScript link if possible.
There was a problem hiding this comment.
Ohhh, I missed that distinction, indeed. That can be fixed. (Using <l spec=ecmascript>{{TypeError}}</l>.)
index.bs
Outdated
| Note: Don't confuse the "{{SyntaxError!!exception}}" {{DOMException}} defined here | ||
| with ECMAScript's {{ECMAScript/SyntaxError}}. | ||
| with ECMAScript's {{SyntaxError}}. | ||
| "{{SyntaxError!!exception}}" {{DOMException}} is used to report parsing errors in web APIs, | ||
| for example when parsing selectors, | ||
| while the ECMAScript {{ECMAScript/SyntaxError}} is reserved for the ECMAScript parser. | ||
| while the ECMAScript {{SyntaxError}} is reserved for the ECMAScript parser. | ||
| To help disambiguate this further, | ||
| always favor the "{{SyntaxError!!exception}}" {{DOMException}} notation | ||
| over just using {{SyntaxError!!exception}} to refer to the {{DOMException}}. [[DOM]] |
There was a problem hiding this comment.
The new {{SyntaxError}} references are incorrect: it goes to the unrelated "SyntaxError" DOMException in Web IDL rather than ES. In fact, this paragraph explains exactly how they are different.
|
Looks like the webref data source bug is fixed, so this is ready for @tabatkins to fix up the exception references and then we should be good to go. |
|
All right, the TypeError and SyntaxError refs are fixed, and now that WebRef has fixed a lot of the missing dfns, I've done another pass thru the anchors and removed every single one that we no longer need. I specifically confirmed for all of these new removals that they didn't change any behavior; two or three now link more directly to a term in the first paragraph of a section, rather than the section itself, but I figured that was fine. The rest all link to the same anchors as before (just using the multipage version of the spec instead). |
PR whatwg#1270's substitutions ended up turning some occurrences of {{ECMAScript/TypeError}} into <l spec=ecmascript>{{TypeError}}</l>} which a trailing "}" character at the end.
This is purely an editorial change. It fixes the link errors the current spec has (around Memory and Memory/buffer), and additionally removes a large number of the explicit anchors defined in the spec. For the most part these removals were simple; in a few places I fixed the linking syntax down in the spec body as well.
The abstract-op anchors left behind just aren't being detected by WebRef; I opened an issue about them. When that's fixed we should be able to remove that entire block.
A number of the remaining anchors are things that should probably be fixed in ECMAScript, but I don't have bandwidth to do that right now. I haven't done a full accounting of the remaining anchor values; some of them might also be removable right now.
Preview | Diff