epubcheck
epubcheck copied to clipboard
w3c
"rel" attribute disallowed in "a" element inside "svg" element (epubcheck 4.2.2)
The nesting in the title yields the following error:
error {
resource: "OEBPS/page_22.xhtml"
line: 11
column: 1431618
description: "Error while parsing file: attribute \"rel\" not allowed here; expected attribute \"alignment-baseline\", \"aria-activedescendant\", \"aria-atomic\", \"aria-autocomplete\", \"aria-busy\", \"aria-checked\", \"aria-colcount\", \"aria-colindex\", \"aria-colspan\", \"aria-controls\", \"aria-current\", \"aria-describedby\", \"aria-details\", \"aria-disabled\", \"aria-dropeffect\", \"aria-errormessage\", \"aria-expanded\", \"aria-flowto\", \"aria-grabbed\", \"aria-haspopup\", \"aria-hidden\", \"aria-invalid\", \"aria-keyshortcuts\", \"aria-label\", \"aria-labelledby\", \"aria-level\", \"aria-live\", \"aria-modal\", \"aria-multiline\", \"aria-multiselectable\", \"aria-orientation\", \"aria-owns\", \"aria-placeholder\", \"aria-posinset\", \"aria-pressed\", \"aria-readonly\", \"aria-relevant\", \"aria-required\", \"aria-roledescription\", \"aria-rowcount\", \"aria-rowindex\", \"aria-rowspan\", \"aria-selected\", \"aria-setsize\", \"aria-sort\", \"aria-valuemax\", \"aria-valuemin\", \"aria-valuenow\", \"aria-valuetext\", \"baseline-shift\", \"class\", \"clip\", \"clip-path\", \"clip-rule\", \"color\", \"color-interpolation\", \"color-interpolation-filters\", \"color-profile\", \"color-rendering\", \"cursor\", \"direction\", \"display\", \"dominant-baseline\", \"enable-background\", \"externalResourcesRequired\", \"fill\", \"fill-opacity\", \"fill-rule\", \"filter\", \"flood-color\", \"flood-opacity\", \"focusable\", \"font-family\", \"font-size\", \"font-size-adjust\", \"font-stretch\", \"font-style\", \"font-variant\", \"font-weight\", \"glyph-orientation-horizontal\", \"glyph-orientation-vertical\", \"href\", \"id\", \"image-rendering\", \"kerning\", \"letter-spacing\", \"lighting-color\", \"marker-end\", \"marker-mid\", \"marker-start\", \"mask\", \"ns:type\", \"onactivate\", \"onclick\", \"onfocusin\", \"onfocusout\", \"onload\", \"onmousedown\", \"onmousemove\", \"onmouseout\", \"onmouseover\", \"onmouseup\", \"opacity\", \"overflow\", \"pointer-events\", \"requiredExtensions\", \"requiredFeatures\", \"role\", \"shape-rendering\", \"stop-color\", \"stop-opacity\", \"stroke\", \"stroke-dasharray\", \"stroke-dashoffset\", \"stroke-linecap\", \"stroke-linejoin\", \"stroke-miterlimit\", \"stroke-opacity\", \"stroke-width\", \"style\", \"systemLanguage\", \"tabindex\", \"text-anchor\", \"text-decoration\", \"text-rendering\", \"transform\", \"unicode-bidi\", \"visibility\", \"word-spacing\", \"writing-mode\", \"xlink:actuate\", \"xlink:arcrole\", \"xlink:role\", \"xlink:show\", \"xlink:title\", \"xlink:type\", \"xml:base\", \"xml:lang\" or \"xml:space\" (with xmlns:ns=\"http://www.idpf.org/2007/ops\") or an attribute from another namespace"
message_id: "RSC-005"
}
An extract of markup would be (ellipses indicate clipped markup):
<body id="68c60dac_body">
<div>
<svg width="100%" height="100%" viewBox="0.0 0.0 960.0 720.0" fill="none" stroke="none" stroke-linecap="square" stroke-miterlimit="10" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns="http://www.w3.org/2000/svg">
<g ...>
...
<a xlink:href="https://..." target="_blank" rel="noreferrer">
...
</g>
</svg>
</div>
</body>
I think this should be okay. Does the SVG schema need updating?
Yes, it's valid, but the schemas don't fully support svg2.
I've opened an issue against the nu validator to add: https://github.com/validator/validator/issues/1008
I am not an svg expert, but I am concerned that it will be difficult to validate svg in the era of living standards. An svg 1.1 validator should reject the example since rel wasn't allowed on a elements in that version. While it is now allowed in svg 2, an svg 2 validator should reject the example because the svg element has a version attribute which is no longer allowed. Are we using (or proposing) some sort of hybrid validator? I wonder if the correct answer is to not worry about svg validation, but that is a larger discussion.
Are we using ... some sort of hybrid validator?
That's what I understand is done for the nu validator based on this discussion I found.
For epubcheck, my understanding is that we want to be in sync with web validation generally, except where epub specifically diverges, so that you don't get mixed signals. That isn't entirely possible at this stage, since we only use the schemas and not the additional java checks, but I believe fuller integration might come in a later refactor of epubcheck.
No opinion here on whether or not to check svg, though. That would be a discussion for the main group to pick up.
Note that I updated the initial description of this bug to remove version="1.1" which was technically wrong for including "rel" inside an SVG nested "a" element. No change in behavior... except now the bug report is now completely correct (I think!).
Another example of a similar issue is this error:
ERROR(RSC-005): target.epub/OEBPS/page_22.xhtml(11,81921): Error while parsing file: attribute "fill" not allowed here; expected attribute "aria-activedescendant", "aria-atomic", "aria-autocomplete", "aria-busy", "aria-checked", "aria-colcount", "aria-colindex", "aria-colspan", "aria-controls", "aria-current", "aria-describedby", "aria-details", "aria-disabled", "aria-dropeffect", "aria-errormessage", "aria-expanded", "aria-flowto", "aria-grabbed", "aria-haspopup", "aria-hidden", "aria-invalid", "aria-keyshortcuts", "aria-label", "aria-labelledby", "aria-level", "aria-live", "aria-modal", "aria-multiline", "aria-multiselectable", "aria-orientation", "aria-owns", "aria-placeholder", "aria-posinset", "aria-pressed", "aria-readonly", "aria-relevant", "aria-required", "aria-roledescription", "aria-rowcount", "aria-rowindex", "aria-rowspan", "aria-selected", "aria-setsize", "aria-sort", "aria-valuemax", "aria-valuemin", "aria-valuenow", "aria-valuetext", "class", "clip", "clip-rule", "color", "color-interpolation", "color-profile", "color-rendering", "cursor", "display", "externalResourcesRequired", "fill-opacity", "filter", "focusable", "height", "href", "id", "image-rendering", "mask", "ns:type", "onactivate", "onclick", "onfocusin", "onfocusout", "onload", "onmousedown", "onmousemove", "onmouseout", "onmouseover", "onmouseup", "opacity", "overflow", "pointer-events", "preserveAspectRatio", "requiredExtensions", "requiredFeatures", "role", "shape-rendering", "stroke-opacity", "style", "systemLanguage", "tabindex", "text-rendering", "transform", "vector-effect", "visibility", "width", "x", "xlink:actuate", "xlink:arcrole", "xlink:href", "xlink:role", "xlink:show", "xlink:title", "xlink:type", "xml:base", "xml:lang", "xml:space" or "y" (with xmlns:ns="http://www.idpf.org/2007/ops") or an attribute from another namespace
when presented with:
<body id="68c60dac_body">
<div>
<svg width="100%" height="100%" viewBox="0.0 0.0 960.0 720.0" fill="none" stroke="none" stroke-linecap="square" stroke-miterlimit="10" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns="http://www.w3.org/2000/svg">
<g ...>
...
<image clip-path="url(#g4d46da3d71_0_0.1)" fill="#000" width="719.0" height="483.0" x="0.0" y="0.0" preserveAspectRatio="none" xlink:href="data:image/png;base64,iVBO ... ">
...
</g>
</svg>
</div>
</body>
I think that SVG validation should not be lifted, since maintenance of EPUB assets containing invalid SVG fragments is likely to be annoying.
While I'm fine if the EPUB specs ultimately do not require validity for conformance, it would greatly reduce the usefulness of epubcheck if we stop checking the validity of SVG and XHTML files completely.
Perhaps we need to consider other options, like defining a separate message category for validity issues or only strictly validating if a flag is set.
