Update data for when href (not xlink:href) can be used in SVG

[foolip]

Test case used for <image href>:

<svg width="400" height="200"
     xmlns="http://www.w3.org/2000/svg"
     xmlns:xlink="http://www.w3.org/1999/xlink">
  <image xlink:href="https://mdn.mozillademos.org/files/6457/mdn_logo_only_color.png" x="0" height="200" width="200"/>
  <image href="https://mdn.mozillademos.org/files/6457/mdn_logo_only_color.png" x="200" height="200" width="200"/>
</svg>

Test case used for <use href>:

<svg viewBox="0 0 30 10"
     xmlns="http://www.w3.org/2000/svg"
     xmlns:xlink="http://www.w3.org/1999/xlink">
  <circle id="myCircle" cx="5" cy="5" r="4" stroke="black"/>
  <use xlink:href="#myCircle" x="10" fill="green"/>
  <use href="#myCircle" x="20" fill="green"/>
</svg>

Based on these tests, the change was narrowed down to Chrome 50, and
this commit pinpointed:
https://chromium.googlesource.com/chromium/src/+/2f0b655515af2df03d41d4f894e0dd59d5cb6411

At the time, SVGURIReference was used in these C++ classes:

https://chromium.googlesource.com/chromium/src/+/2f0b655515af2df03d41d4f894e0dd59d5cb6411/third_party/WebKit/Source/core/svg/SVGAElement.h#32
https://chromium.googlesource.com/chromium/src/+/2f0b655515af2df03d41d4f894e0dd59d5cb6411/third_party/WebKit/Source/core/svg/SVGCursorElement.h#35
https://chromium.googlesource.com/chromium/src/+/2f0b655515af2df03d41d4f894e0dd59d5cb6411/third_party/WebKit/Source/core/svg/SVGFEImageElement.h#35
https://chromium.googlesource.com/chromium/src/+/2f0b655515af2df03d41d4f894e0dd59d5cb6411/third_party/WebKit/Source/core/svg/SVGPatternElement.h#40
https://chromium.googlesource.com/chromium/src/+/2f0b655515af2df03d41d4f894e0dd59d5cb6411/third_party/WebKit/Source/core/svg/SVGTextPathElement.h#46
https://chromium.googlesource.com/chromium/src/+/2f0b655515af2df03d41d4f894e0dd59d5cb6411/third_party/WebKit/Source/core/svg/SVGUseElement.h#37

The following classes are the same, but for these there's no BCD entry
for the namespaceless href attribute, so they're not updated:
https://chromium.googlesource.com/chromium/src/+/2f0b655515af2df03d41d4f894e0dd59d5cb6411/third_party/WebKit/Source/core/svg/SVGFilterElement.h#42
https://chromium.googlesource.com/chromium/src/+/2f0b655515af2df03d41d4f894e0dd59d5cb6411/third_party/WebKit/Source/core/svg/SVGGradientElement.h#44
https://chromium.googlesource.com/chromium/src/+/2f0b655515af2df03d41d4f894e0dd59d5cb6411/third_party/WebKit/Source/core/svg/SVGImageElement.h#36
https://chromium.googlesource.com/chromium/src/+/2f0b655515af2df03d41d4f894e0dd59d5cb6411/third_party/WebKit/Source/core/svg/SVGMPathElement.h#34
https://chromium.googlesource.com/chromium/src/+/2f0b655515af2df03d41d4f894e0dd59d5cb6411/third_party/WebKit/Source/core/svg/SVGScriptElement.h#38

Data is mirror for Chromium browsers expect Edge, which is mostly set to
"12" already and where it's null would require additional testing.

[myakura]

TODO Safari.

The basic support for null-namespace href was added in r234683 and shipped in Safari 12.1 (iOS 12.2).
(the changeset landed in STP 63 and the Safari 12.1 blog post says it includes the changes made in STP 59-76).

However, the initial implementation has a bug where it preferes xlink:href over bare href when both attributes are specified (bug 195802). That was fixed in r249593, landed in STP 92, made its release in Safari 13.1.

[ddbeck]

Hey @foolip, this PR has been in draft status for some time. What needs to happen to get this into a reviewable state? Any help needed here?

[foolip]

Thanks for the details, @myakura! Going to update the Safari data, I found that it was actually already there, added in #5255.

Since the changes I've made are correct, I'm just going to leave it like this instead of trying to track down every last detail. I'll edit the description to say what this does now.

[foolip]

@ddbeck this is now ready for review.

[ddbeck]

(I've temporarily relaxed the branch protection rules to merge this, since the older workflows satisfied.)