Trusted types attributes #1268

lukewarlow · 2024-03-27T19:22:16Z

This calls the get Trusted Types-compliant attribute value algorithm from Trusted Types (w3c/trusted-types#418) from attribute's change, append, and replace.

Changed the signature of setAttribute and setAttributeNS to accept Trusted Types as values. The underlying Attr node's values continue to be DOMString, so moving nodes across elements or adding standalone attributes to elements can cause TT violations. This matches WPT tests and the Chromium's implementation.

See and #789. Supercedes #809 and #1247

At least two implementers are interested (and none opposed):
- Chromium
- Gecko
Tests are written and can be reviewed and commented upon at:
- https://github.com/web-platform-tests/wpt/tree/master/trusted-types
Implementation bugs are filed:
- Chromium:https://bugs.chromium.org/p/chromium/issues/detail?id=739170
- Gecko: https://bugzilla.mozilla.org/show_bug.cgi?id=1508286
- WebKit: https://bugs.webkit.org/show_bug.cgi?id=266630
MDN issue is filed: …
The top of this comment includes a clear commit message to use.

(See WHATWG Working Mode: Changes for more details.)

Preview | Diff

lukewarlow · 2024-03-27T19:22:41Z

This is a clone of #1247 where I will finish any outstanding work to get this across the line

lukewarlow · 2024-04-11T12:21:53Z

See also #1258 which is another integration point with the DOM spec that we need for TT.

dom.bs

otherdaniel · 2024-04-15T17:13:57Z

dom.bs

+ <li><p><a>Validate and set attribute value</a> <var>attr</var>'s <a for="Attr">value</a> for
+ <var>attr</var> with <var>element</var>.
+
+ <li><p>If <var>element</var> <a lt="has an attribute">has</a> an <a>attribute</a>


I still find it difficult to wrap my head around this logic. I think conceptually, we want to have the old value -- from before the call, and thus before a default policy might have mucked with it. That's what should go into the change attribute logic. But once we have that, I'm not sure why we'd need to throw an exception here. I'm not sure why we'd have to care whether the default policy does anything funny with the attribute in the mean time.

I've read through this spec path and I believe that, if the default policy removes the existing attribute node, then this will call replace an attribute, which in turn calls replace a list item, which results in a no-op because the old value no longer exists to be replaced.

So I think in this situation you're probably right the spec doesn't need to do anything, will make that change.

Having said that I'm slightly uneasy about stuff like attribute change steps still firing and how that all works spec wise.

Both Chromium and WebKit don't actually follow the spec 100% here and so I think would actually result in a different behaviour to this spec (the index lookup happens after the TT call) so that's also not ideal.

lukewarlow · 2024-04-22T13:18:23Z

I think I've addressed all the comments from #1247.

I do want to point out Chrome and WebKit don't (or at least not in a way obvious to me) 100% follow the flow of the spec and as a result this may result in differences specifically in weird cases with attribute mutation.

So that bit especially it would be good to get feedback on.

It's also worth being aware that like Chromium's implementation this spec means that certain ways to update a nodes value don't work with a trusted type object as a user might expect. (e.g. iframe.getAttributeNode('srcdoc').value = trustedHTMLObj; will throw unless allowed by a default policy). I think in pratice this will be fine, and in some cases is the only real option we have without nodes keeping track of whether they're trusted or not (which would add lots of complexity)

annevk

This change is either incomplete or makes many cosmetic changes that would be best proposed separately as they confuse me quite a bit.

dom.bs

lukewarlow · 2024-04-22T15:47:56Z

This change is either incomplete or makes many cosmetic changes that would be best proposed separately as they confuse me quite a bit.

Apologies I thought I'd reverted all of these changes but I missed a few, it's because I changed stuff and then changed it back and this led to some wonky diffs. Have hopefully reverted all of these unnecessary changes

annevk · 2024-04-22T16:07:37Z

dom.bs

+ <li><p>If <var>attribute</var>'s <a for=Attr>element</a> <a lt="has an attribute">has</a>
+ an <a>attribute</a> <var>attribute</var>, then <a>handle attribute changes</a> for
+ <var>attribute</var> with <var>attribute</var>'s <a for=Attr>element</a>, <var>oldValue</var>, and
+ <var>value</var>.


Should this be value or attribute's value? They can be different, right?

dom.bs

lukewarlow · 2024-05-07T15:54:52Z

@otherdaniel, @annevk , and @smaug---- regarding the case where the default policy changes assumptions about the existence of an attribute mid-way through what would you prefer the spec say to do? Currently I've specced to throw, but Chromium currently re-looks up the index (spec doesn't explicitly work on an index basis but Chromium and WebKit do)

annevk · 2024-05-13T14:58:26Z

Attributes are stored in a list and those do have indices per Infra. What am I missing?

lukewarlow · 2024-05-13T15:03:24Z

Sorry I mean algorithmically the spec and implementations don't follow the same flow. So it's trickier to reason between the spec and implementation. This might just be my lack of familiarity with these APIs too.

annevk · 2024-05-13T15:11:38Z

The path of least resistance is prolly matching Chromium. Introducing new paths that throw is always risky. If you are looking for guidance as to how, I'd need quite a bit more context to provide helpful suggestions.

See w3c/trusted-types#418 and whatwg#789. Supercedes PR whatwg#809.

lukewarlow · 2024-05-16T15:44:07Z

The path of least resistance is prolly matching Chromium. Introducing new paths that throw is always risky. If you are looking for guidance as to how, I'd need quite a bit more context to provide helpful suggestions.

I've partially matched Chromium but gone a bit further with the error correction as Chromium still has some bugs wrt mutating attributes within the default policy.

lukewarlow · 2024-05-16T15:44:40Z

@otherdaniel if you've got time could you read through this to ensure it makes sense to you?

lukewarlow · 2024-05-16T15:49:29Z

dom.bs

@@ -6530,23 +6545,39 @@ string <var>namespace</var> (default null):</p>

 <div algorithm>
 <p>To <dfn export id=concept-element-attributes-set-value>set an attribute value</dfn> given an


This algorithm covers setAttributeNS

lukewarlow · 2024-05-16T15:49:54Z

dom.bs

- <a for=Attr>local name</a> is <var>qualifiedName</var>, <a for=Attr>value</a> is
- <var>value</var>, and <a for=Node>node document</a> is <a>this</a>'s <a for=Node>node document</a>,
- then <a lt="append an attribute">append</a> this <a>attribute</a> to <a>this</a>, and then return.
+ <li><p>Let <var>attributeExists</var> be false if <var>attribute</var> is null, and true otherwise.


This algorithm covers setAttribute

lukewarlow · 2024-05-16T15:51:10Z

dom.bs

@@ -7382,7 +7430,13 @@ string <var>value</var>, run these steps:
 <li><p>If <var>attribute</var>'s <a for=Attr>element</a> is null, then set <var>attribute</var>'s
 <a for=Attr>value</a> to <var>value</var>.

- <li><p>Otherwise, <a lt="change an attribute">change</a> <var>attribute</var> to <var>value</var>.
+ <li><p>Otherwise:


This covers attr.nodeValue, attr.value, and attr.textContent.

lukewarlow · 2024-05-16T15:53:00Z

dom.bs

@@ -6519,6 +6529,11 @@ string <var>namespace</var> (default null):</p>

 <li><p>If <var>oldAttr</var> is <var>attr</var>, return <var>attr</var>.

+ <li><p>Let <var>verifiedValue</var> be the result of calling <a>verify attribute value</a>


This covers, setAttributeNode, setAttributeNodeNS, setNamedItem and setNamedItemNS

dom.bs

… Chrome.

…erved in Chrome." This reverts commit 2f00466.

This reverts commit 6e40964.

https://bugs.webkit.org/show_bug.cgi?id=275352 Reviewed by NOBODY (OOPS!). The DOM spec PR no longer enforced Trusted Types within toggleAttribute so this removes that from the implementation. See whatwg/dom#1268 * LayoutTests/imported/w3c/web-platform-tests/trusted-types/Element-toggleAttribute-expected.txt: Added. * LayoutTests/imported/w3c/web-platform-tests/trusted-types/Element-toggleAttribute.html: Added. * Source/WebCore/dom/Element.cpp: (WebCore::Element::toggleAttribute):

annevk · 2024-06-12T11:44:20Z

This now seems to invoke TT even when we have specifications manipulate attributes, or when element.id's setter is used. Is that intentional?

For review it really helps to have an up-to-date commit message that clearly describes the intent. That's why we made it one of the PR checkboxes.

It also seems that it would cause a change in ordering between enforcing this in IDL and in the algorithms as now certain exceptions are thrown first. That will need test coverage.

https://bugs.webkit.org/show_bug.cgi?id=275352 Reviewed by Darin Adler. The DOM spec PR no longer enforced Trusted Types within toggleAttribute so this removes that from the implementation. See whatwg/dom#1268 * LayoutTests/imported/w3c/web-platform-tests/trusted-types/Element-toggleAttribute-expected.txt: Added. * LayoutTests/imported/w3c/web-platform-tests/trusted-types/Element-toggleAttribute.html: Added. * Source/WebCore/dom/Element.cpp: (WebCore::Element::toggleAttribute): Canonical link: https://commits.webkit.org/279950@main

lukewarlow · 2024-06-13T15:07:52Z

Trusted Types integration with DOM

The DOM spec already contains partial integration with Trusted Types this was included in #1258

This PR updates the DOM functions related to attributes.

The following functions' IDL are adjusted, to accept a TrustedType (TrustedHTML or TrustedScript or TrustedScriptURL) or a string.

Element.setAttribute
Element.setAttributeNS

The above two functions are the "sanctioned" methods for updating the value of an attribute, any other way relies on default policy else they'll fail TT checks.

The algorithm for setAttribute function (and the "set an attribute value" algorithm only when used by setAttributeNS) is updated to call into the Trusted Types such that if the attribute being set is one of those covered by Trusted Types,
the value being set is verified (potentially being replaced with a different value).

These algorithms include some handling for if the trusted types default policy changes whether the attribute being set exists or not.

Element.setAttributeNode and Element.setAttributeNodeNS both, along with NamedNodeMap.setNamedItem and NamedNodeMap.setNamedItemNS make use of the "setting an attribute" algorithm, which is similarly updated to call into Trusted Types.

Attr.value, Node.textContent, and Node.nodeValue setters call "set an existing attribute value", this algorithm is updated such that for attributes attached to an element, trusted types is called into and verifies the value.

In all these cases the changes only have any impact for a select few attributes (e.g. event handlers, iframe srcdoc attribute)

lukewarlow · 2024-06-13T15:08:30Z

The change to set an attribute value breaks iframe.srcdoc property setting currently (it'll do a double check). It's possible that there's other reflection properties that have the same issue (script.src comes to mind).

I'll take a look into how best to sort that out. I'm thinking adding an optional boolean to control this check, that setAttributeNS sets to true?

Edit: this should be fixed with the new verify flag.

…rue from setAttributeNS to fix issue with current spec

lukewarlow · 2024-06-13T15:28:52Z

This now seems to invoke TT even when we have specifications manipulate attributes,

Should I add a verify flag to these other algorithms so they can't be accidentally invoked doing TT checks?

I don't think element.id is related because it's not one of these specific attributes. But you raise a good point which I think the verify flag I just added addresses

lukewarlow force-pushed the trusted-types-attributes branch 2 times, most recently from 3562fcb to ac5b4aa Compare April 10, 2024 15:52

lukewarlow mentioned this pull request Apr 11, 2024

Draft integration with Trusted Types, take 2. #1247

Closed

5 tasks

lukewarlow force-pushed the trusted-types-attributes branch from 524d8cd to f8877b6 Compare April 11, 2024 12:18

lukewarlow marked this pull request as ready for review April 11, 2024 12:21

lukewarlow marked this pull request as draft April 15, 2024 10:22

lukewarlow force-pushed the trusted-types-attributes branch from f8877b6 to ee9915e Compare April 15, 2024 11:45

otherdaniel reviewed Apr 15, 2024

View reviewed changes

dom.bs Outdated Show resolved Hide resolved

otherdaniel reviewed Apr 15, 2024

View reviewed changes

lukewarlow marked this pull request as ready for review April 22, 2024 12:53

lukewarlow requested review from annevk and smaug---- April 22, 2024 13:13

lukewarlow force-pushed the trusted-types-attributes branch from ba80870 to 36476b5 Compare April 22, 2024 13:57

lukewarlow mentioned this pull request Apr 22, 2024

[Meta] Upstream changes w3c/trusted-types#476

Open

annevk reviewed Apr 22, 2024

View reviewed changes

dom.bs Outdated Show resolved Hide resolved

dom.bs Outdated Show resolved Hide resolved

dom.bs Outdated Show resolved Hide resolved

dom.bs Outdated Show resolved Hide resolved

dom.bs Outdated Show resolved Hide resolved

lukewarlow force-pushed the trusted-types-attributes branch from 370f5c9 to 4421d50 Compare April 22, 2024 15:46

lukewarlow force-pushed the trusted-types-attributes branch 2 times, most recently from 1d42460 to 1eeaf00 Compare April 22, 2024 15:52

annevk reviewed Apr 22, 2024

View reviewed changes

koto added 3 commits May 16, 2024 16:39

Draft integration with Trusted Types, take 2.

6d4b27d

See w3c/trusted-types#418 and whatwg#789. Supercedes PR whatwg#809.

Added integration in 'set an attribute value'.

98ba2fe

Removed stringification.

ce190dd

lukewarlow force-pushed the trusted-types-attributes branch 2 times, most recently from 4089eaf to 02db8c7 Compare May 16, 2024 15:41

lukewarlow requested a review from otherdaniel May 16, 2024 15:43

lukewarlow requested a review from annevk May 16, 2024 15:44

Update how the enforcement is done

55af76e

lukewarlow force-pushed the trusted-types-attributes branch from 02db8c7 to 55af76e Compare May 16, 2024 15:47

lukewarlow commented May 16, 2024

View reviewed changes

Add tentative enforcement for toggleAttribute

6e40964

lukewarlow commented May 16, 2024

View reviewed changes

dom.bs Outdated Show resolved Hide resolved

lukewarlow added 3 commits May 16, 2024 17:57

Update enforcement for toggleAttribute to match behaviour observed in…

2f00466

… Chrome.

Revert "Update enforcement for toggleAttribute to match behaviour obs…

08f47c8

…erved in Chrome." This reverts commit 2f00466.

Revert "Add tentative enforcement for toggleAttribute"

03384b2

This reverts commit 6e40964.

lukewarlow mentioned this pull request Jun 11, 2024

Remove Trusted Types enforcement from toggleAttribute WebKit/WebKit#29712

Merged

Add verify flag to set an attribute value that is default false and t…

1ec907b

…rue from setAttributeNS to fix issue with current spec

lukewarlow assigned domfarolino and unassigned domfarolino Jun 13, 2024

lukewarlow requested review from domfarolino and removed request for otherdaniel June 13, 2024 19:31

charliewolfe mentioned this pull request Jun 14, 2024

[Site Isolation] Fix file open panels WebKit/WebKit#29802

Merged

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Trusted types attributes #1268

Trusted types attributes #1268

lukewarlow commented Mar 27, 2024 •

edited by pr-preview bot

lukewarlow commented Mar 27, 2024

lukewarlow commented Apr 11, 2024

otherdaniel Apr 15, 2024

lukewarlow Apr 22, 2024

lukewarlow Apr 22, 2024

lukewarlow commented Apr 22, 2024

annevk left a comment

lukewarlow commented Apr 22, 2024

annevk Apr 22, 2024

lukewarlow commented May 7, 2024 •

edited

annevk commented May 13, 2024

lukewarlow commented May 13, 2024

annevk commented May 13, 2024

lukewarlow commented May 16, 2024

lukewarlow commented May 16, 2024

lukewarlow May 16, 2024

lukewarlow May 16, 2024

lukewarlow May 16, 2024

lukewarlow May 16, 2024

annevk commented Jun 12, 2024

lukewarlow commented Jun 13, 2024 •

edited

lukewarlow commented Jun 13, 2024 •

edited

lukewarlow commented Jun 13, 2024

		@@ -6530,23 +6545,39 @@ string <var>namespace</var> (default null):</p>

		<div algorithm>
		<p>To <dfn export id=concept-element-attributes-set-value>set an attribute value</dfn> given an

		@@ -6519,6 +6529,11 @@ string <var>namespace</var> (default null):</p>

		<li><p>If <var>oldAttr</var> is <var>attr</var>, return <var>attr</var>.

		<li><p>Let <var>verifiedValue</var> be the result of calling <a>verify attribute value</a>

Trusted types attributes #1268

Are you sure you want to change the base?

Trusted types attributes #1268

Conversation

lukewarlow commented Mar 27, 2024 • edited by pr-preview bot

lukewarlow commented Mar 27, 2024

lukewarlow commented Apr 11, 2024

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

lukewarlow commented Apr 22, 2024

annevk left a comment

Choose a reason for hiding this comment

lukewarlow commented Apr 22, 2024

Choose a reason for hiding this comment

lukewarlow commented May 7, 2024 • edited

annevk commented May 13, 2024

lukewarlow commented May 13, 2024

annevk commented May 13, 2024

lukewarlow commented May 16, 2024

lukewarlow commented May 16, 2024

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

annevk commented Jun 12, 2024

lukewarlow commented Jun 13, 2024 • edited

Trusted Types integration with DOM

lukewarlow commented Jun 13, 2024 • edited

lukewarlow commented Jun 13, 2024

lukewarlow commented Mar 27, 2024 •

edited by pr-preview bot

lukewarlow commented May 7, 2024 •

edited

lukewarlow commented Jun 13, 2024 •

edited

lukewarlow commented Jun 13, 2024 •

edited