"Warning: Duplicate attribute id" unnecessary for id attributes that are unique to an SVG tree node?

[ajeanmahoney]

Describe the issue

When xml2rfc is run on [1], it produces many, many warnings that have the form of:

Warning: Duplicate attribute id="E1-STIXWEBMAIN-3D" found after including svg 
from inline:b'<svg xmlns:xlink="http://www.w3' ....  This can cause problems with some browsers.

There are multiple SVG drawings in [1] that are reusing shapes that have been defined by <path> elements (for instance, E1-STIXWEBMAIN-3D is a thin, horizontal rectangle). These <path> elements have id attributes that are unique to each SVG node tree; however, xml2rfc complains. I think these warnings may be incorrect. All major browsers support the id attribute [2], and this document is not using svgViewElement. The output appears fine in Firefox.

[1] https://www.rfc-editor.org/v3test/draft-ietf-tcpm-rfc8312bis-15.xml [2] https://developer.mozilla.org/en-US/docs/Web/SVG/Attribute/id

Code of Conduct

• [X] I agree to follow the IETF's Code of Conduct

[kesara]

The warning is correct. In HTML specs, id must be unique for the document [1]. These SVG drawings are embedded in HTML and are part of the final document.

[1] https://developer.mozilla.org/en-US/docs/Web/HTML/Global_attributes/id

[kesara]

W3C HTML validator reports these as errors: https://validator.w3.org/nu/?doc=https%3A%2F%2Fwww.rfc-editor.org%2Fv3test%2Fdraft-ietf-tcpm-rfc8312bis-15.html

[ajeanmahoney]

Hrm, okay, we can share that with the authors. Is this something svgcheck should look for then? Because svgcheck thinks the SVG in the document is valid.

[kesara]

Is this something svgcheck should look for then? Because svgcheck thinks the SVG in the document is valid.

Maybe... Because each SVG is valid on its own. This becomes a problem when all of them are combined to gather.

[rjsparks]

I don't know if we need to ask authors to put a bunch of effort into making their ids across svg documents unique.

Backing out to the original ask in the issue : can we perhaps build a mode (make it default?) where xml2rfc reports on the number of types of warnings vs dumping all of them? Or make it where the invoker can selectively suppress warnings?

[alicerusso]

So, what action should the RPC take when we encounter this warning? Ask the authors to change the ids within the SVG so that they are unique across all SVG within a given document? Or, disregard this warning?

Example: https://www.ietf.org/archive/id/draft-ietf-elegy-rfc8989bis-05.xml

Gives warnings as follows (and many more):

(343): Warning: Duplicate attribute id="E1-STIXWEBMAIN-29" found after including svg from inline:b'<svg xmlns:xlink="http://www.w3' ....  This can cause problems with some browsers.
(343): Warning: Duplicate attribute id="E1-STIXWEBSIZE3-28" found after including svg from inline:b'<svg xmlns:xlink="http://www.w3' ....  This can cause problems with some browsers.
(343): Warning: Duplicate attribute id="E1-STIXWEBSIZE3-29" found after including svg from inline:b'<svg xmlns:xlink="http://www.w3' ....  This can cause problems with some browsers.

[cabo]

On 2023-03-08, at 21:05, Kesara Rathnayake @.***> wrote:

The warning is correct. In HTML specs, id must be unique for the document [1]. These SVG drawings are embedded in HTML and are part of the final document.

[1] https://developer.mozilla.org/en-US/docs/Web/HTML/Global_attributes/id

Doing an id= rename across all SVGs might be something that an authoring tool should do, so that xml2rfc actually gets what counts valid XML to start with.

Now https://github.com/cabo/kramdown-rfc/issues/193

Of course, other courses that multiple SVG files generated by the same tool will take will also have the same problem, so maybe I’ll extract the deduplicator as a separate tool that can be employed by these other paths.

Grüße, Carsten

[cabo]

Kramdown-rfc 1.6.28 now has an experimental addition:

$ xml2rfc --html draft-ietf-tcpm-rfc8312bis-15.xml sprint/svg-ids/draft-ietf-tcpm-rfc8312bis-15.xml(490): Warning: Duplicate attribute id="E1-STIXWEBNORMALI-1D461" found after including svg from inline:b'<svg xmlns:xlink="http://www.w3' .... This can cause problems with some browsers. … and tons more …

$ kramdown-rfc-clean-svg-ids draft-ietf-tcpm-rfc8312bis-15.xml >cleaned.xml *** warning: unused ID: #<Set: {"E1-STIXWEBMAIN-7B"}> *** warning: unused ID: #<Set: {"E1-STIXWEBMAIN-7B"}> *** warning: unused ID: #<Set: {"E1-STIXWEBMAIN-7B"}> *** warning: unused ID: #<Set: {"E1-STIXWEBMAIN-7D"}> *** warning: unused ID: #<Set: {"E1-STIXWEBMAIN-28", "E1-STIXWEBMAIN-29"}> *** warning: unused ID: #<Set: {"E1-STIXWEBMAIN-28", "E1-STIXWEBMAIN-29"}> $ xml2rfc --html cleaned.xml Created file cleaned.html

Please ignore the warnings from the tool; there seem to be some superfluous definitions in the SVG elements, which does not hurt (nobody would notice these had I not written some diagnostics into the tool).

As usual,

gem update kramdown-rfc

to get the addition.

Grüße, Carsten

[cabo]

To avoid the need for adding processing pipelines, kramdown-rfc 1.6.29 now has a YAML header option:

svg-id-cleanup: true

More details in: https://github.com/cabo/kramdown-rfc/wiki/SVG#svg-id-collisions

[ajeanmahoney]

@cabo Thanks for these updates, Carsten. I successfully ran kramdown-rfc-clean-svg-ids on a document that had duplicate IDs to clean them up.