1816628 - Issues with text-shadow in SVG text (fill vs stroke)

Status: RESOLVED FIXED

(Core :: SVG, defect)

Description

[Robert Longson [:longsonr]]

See testcase: the SVG text renders with colored stroke and fill, as specified in the attributes of the <svg> element, and inherits its font family/weight/size from the body.

• The shadow is always filled even if the text is not filled.

• The shadow does not respect stroke at all.

• What should happen if the text has opacity? What if there is fill-opacity and stroke-opacity, should they be observed in the shadow? What if the fill and stroke opacity are different?

Comment 1

[Robert Longson [:longsonr]]

Attachment: testcase

Comment 2

[Jonathan Kew [:jfkthame]]

I notice that these issues aren't actually limited to SVG text; similar considerations apply to HTML text with shadows in combination with -webkit-text-stroke. Example: https://codepen.io/jfkthame/pen/LYJYRPd

So this may be a more general nsTextFrame painting issue.

Comment 3

[Jonathan Kew [:jfkthame]]

I've filed https://github.com/w3c/csswg-drafts/issues/8449 regarding the spec for text-shadow; this is primarily discussing HTML text, but we probably want to consider both the SVG and HTML behaviors together, as it's pretty weird for authors if text-shadow behaves quite differently in the two contexts.

Comment 4

[Robert Longson [:longsonr]]

Also, should the text decoration be shadowed too?

Comment 5

[Jonathan Kew [:jfkthame]]

The CSS text-decoration spec for text-shadow says that:

Each layer shadows the element’s text and all its text decorations (composited together).

So yes, decorations get shadowed. (I think we do that already, don't we?)

Comment 6

[Jonathan Kew [:jfkthame]]

Marking as S3 for now, we may want to re-evaluate this once the CSSWG takes a look at the spec issue and we have confirmation of the expected results.

Comment 7

[Robert Longson [:longsonr]]

Attachment: testcase with stroke-dasharray

Comment 8

[Robert Longson [:longsonr]]

Attachment: testcase with vector-effect: non-scaling-stroke

Comment 9

[Robert Longson [:longsonr]]

Attachment: Bug 1816628 Part 1 - Account for horizontal overflow caused by text-shadow r=jfkthame

Comment 10

[Robert Longson [:longsonr]]

Attachment: Bug 1816628 Part 2 - Make text-shadow react to stroke properties and fill opacity r=jfkthame

Depends on D200435

Comment 11

[Robert Longson [:longsonr]]

https://treeherder.mozilla.org/jobs?repo=try&revision=1b934ee10d2c196f024df50bce926c7c3cd4f71d

Comment 12

[Robert Longson [:longsonr]]

https://treeherder.mozilla.org/jobs?repo=try&revision=fbc53a9ed8bac0fe7d46833b2df2c241b30cecdd

Comment 13

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/autoland/rev/ddaf27e9243d
Part 1 - Account for horizontal overflow caused by text-shadow r=jfkthame
https://hg.mozilla.org/integration/autoland/rev/d58d91582a84
Part 2 - Make text-shadow react to stroke properties and fill opacity r=jfkthame

Comment 14

[Web Platform Test Sync Bot [:wpt-sync] (Matrix: #interop:mozilla.org)]

Created web-platform-tests PR https://github.com/web-platform-tests/wpt/pull/44750 for changes under testing/web-platform/tests

Comment 15

[Serban Stanca [:SerbanS]]

Backed out for causing linux related wr failures in SVGTextFrame.cpp.

• Backout link
• Push with failures
• Failure Log
• Failure line: SUMMARY: UndefinedBehaviorSanitizer: undefined-behavior /builds/worker/checkouts/gecko/layout/svg/SVGTextFrame.cpp:2666:7 in

L.E. there are also some OS X mochitests-plain failures that are related to this.

• Push with failures
• Failure Log
• Failure line: TEST-UNEXPECTED-FAIL | dom/svg/test/test_bug1426594.html | undefined assertion name - got 69.10000610351562, expected 63.100006103515625 epsilon: +/- 5

Comment 16

[Web Platform Test Sync Bot [:wpt-sync] (Matrix: #interop:mozilla.org)]

Upstream PR was closed without merging

Comment 17

[Robert Longson [:longsonr]]

https://treeherder.mozilla.org/jobs?repo=try&revision=5203480757a80ed85d80201cc9058cb2c41f87bd

Comment 18

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/autoland/rev/3ee753e740b9
Part 1 - Account for horizontal overflow caused by text-shadow r=jfkthame
https://hg.mozilla.org/integration/autoland/rev/e32deae0b0b4
Part 2 - Make text-shadow react to stroke properties and fill opacity r=jfkthame

Comment 19

[Serban Stanca [:SerbanS]]

Backed out for causing OS X mochitests-plain failures in test_bug1426594.html.

• Backout link
• Push with failures
• Failure Log
• Failure line: TEST-UNEXPECTED-FAIL | dom/svg/test/test_bug1426594.html | undefined assertion name - got 69.10000610351562, expected 63.100006103515625 epsilon: +/- 5

Comment 20

[Web Platform Test Sync Bot [:wpt-sync] (Matrix: #interop:mozilla.org)]

Upstream PR was closed without merging

Comment 21

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/autoland/rev/f637c422003f
Part 1 - Account for horizontal overflow caused by text-shadow r=jfkthame
https://hg.mozilla.org/integration/autoland/rev/27783bb8f6b7
Part 2 - Make text-shadow react to stroke properties and fill opacity r=jfkthame

Comment 22

[Serban Stanca [:SerbanS]]

https://hg.mozilla.org/mozilla-central/rev/f637c422003f
https://hg.mozilla.org/mozilla-central/rev/27783bb8f6b7

Comment 23

[Web Platform Test Sync Bot [:wpt-sync] (Matrix: #interop:mozilla.org)]

Upstream PR merged by moz-wptsync-bot