| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
| Code-Review | +1 |
invalidate: ["layout", "paint"],arguably only layout?
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
class CSSMarginTrimValue : public CSSValue {We typically use CSSValuePair instead of creating new CSSValue classes and do the mapping to e.g. a mask for ApplyValue().
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
class CSSMarginTrimValue : public CSSValue {We typically use CSSValuePair instead of creating new CSSValue classes and do the mapping to e.g. a mask for ApplyValue().
In patch set 6 I created a CSSValueList.
https://chromium-review.googlesource.com/c/chromium/src/+/7544283/6
... which worked fine except for serialization.
So I ended up with this, so that both getPropertyValue() and getComputedStyle() would return "block-start block-end" as "block". There are tests for this.
Why would CSSValuePair be any better than CSSValueList? This property takes between one and four values.
invalidate: ["layout", "paint"],Morten Stenshornearguably only layout?
Done
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
class CSSMarginTrimValue : public CSSValue {Morten StenshorneWe typically use CSSValuePair instead of creating new CSSValue classes and do the mapping to e.g. a mask for ApplyValue().
In patch set 6 I created a CSSValueList.
https://chromium-review.googlesource.com/c/chromium/src/+/7544283/6... which worked fine except for serialization.
So I ended up with this, so that both getPropertyValue() and getComputedStyle() would return "block-start block-end" as "block". There are tests for this.
Why would CSSValuePair be any better than CSSValueList? This property takes between one and four values.
Should I rather go back to patch set 6, and simplify the values in MarginTrim::ParseSingleValue()?
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
class CSSMarginTrimValue : public CSSValue {Morten StenshorneWe typically use CSSValuePair instead of creating new CSSValue classes and do the mapping to e.g. a mask for ApplyValue().
Morten StenshorneIn patch set 6 I created a CSSValueList.
https://chromium-review.googlesource.com/c/chromium/src/+/7544283/6... which worked fine except for serialization.
So I ended up with this, so that both getPropertyValue() and getComputedStyle() would return "block-start block-end" as "block". There are tests for this.
Why would CSSValuePair be any better than CSSValueList? This property takes between one and four values.
Should I rather go back to patch set 6, and simplify the values in MarginTrim::ParseSingleValue()?
Why would CSSValuePair be any better than CSSValueList? This property takes between one and four values.
I thought it took two, so yeah, a list would be the right choice then.
Should I rather go back to patch set 6, and simplify the values in MarginTrim::ParseSingleValue()?
That's what would match how we typically do it, at least for re-ordering.
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
class CSSMarginTrimValue : public CSSValue {Morten StenshorneWe typically use CSSValuePair instead of creating new CSSValue classes and do the mapping to e.g. a mask for ApplyValue().
Morten StenshorneIn patch set 6 I created a CSSValueList.
https://chromium-review.googlesource.com/c/chromium/src/+/7544283/6... which worked fine except for serialization.
So I ended up with this, so that both getPropertyValue() and getComputedStyle() would return "block-start block-end" as "block". There are tests for this.
Why would CSSValuePair be any better than CSSValueList? This property takes between one and four values.
Rune LillesveenShould I rather go back to patch set 6, and simplify the values in MarginTrim::ParseSingleValue()?
Why would CSSValuePair be any better than CSSValueList? This property takes between one and four values.
I thought it took two, so yeah, a list would be the right choice then.
Should I rather go back to patch set 6, and simplify the values in MarginTrim::ParseSingleValue()?
That's what would match how we typically do it, at least for re-ordering.
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
typedom_types: ["Keyword"],How did this in combination with ConvertFlags and all the keywords above not crash StylePropertyMapTest_CSSKeywordValuesTest?
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
How did this in combination with ConvertFlags and all the keywords above not crash StylePropertyMapTest_CSSKeywordValuesTest?
Oh, wait. It appears that I re-added `keywords`. This will cause a crash. Removed again.
computable: false,Morten StenshorneWhy is this false?
Removed. Doesn't have any effect?
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
computable: false,Morten StenshorneWhy is this false?
Removed. Doesn't have any effect?
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
computable: false,Morten StenshorneWhy is this false?
Rune LillesveenRemoved. Doesn't have any effect?
Seems to have had an effect given the latest run?
It sure did.
There was also a parser mistake (assuming that everything is keywords) that I've now fixed.
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Support margin-trim on block containers.
This adds parsing support for the CSS margin-trim property, and layout
support for block containers. Still no support for flex, grid or
multicol containers.
See https://drafts.csswg.org/css-box-4/#margin-trim
Also note that in our implementation margin-trim doesn't affect the
resolved value of margin properties, at least for now. There are
existing tests that expect this, but not clear whether it's correct.
See https://github.com/w3c/csswg-drafts/issues/11506
There are still some failures in css-box/parsing/margin-trim.html , but
not sure if they are valid. Canonical order should match the order in
the spec, which is "block" before "inline", and "start" before "end".
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
The WPT PR for this CL has been merged upstream! https://github.com/web-platform-tests/wpt/pull/57608
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |