SyntaxError: Failed to execute 'querySelector' on 'Element': '#div-2.1' is not a valid selector

[larseggert]

Prerequisites

• I have searched for duplicate or closed issues
• I have validated any HTML to avoid common problems
• I have read the contributing guidelines

Describe the issue

#35566 didn't fix things for all weird IDs. The test case addition below causes the error in the issue title.

Reduced test cases

diff --git a/js/tests/unit/scrollspy.spec.js b/js/tests/unit/scrollspy.spec.js
index c7951e6..704b52a 100644
--- a/js/tests/unit/scrollspy.spec.js
+++ b/js/tests/unit/scrollspy.spec.js
@@ -341,11 +341,13 @@ describe('ScrollSpy', () => {
           '  <ul class="nav">',
           '    <li class="nav-item"><a class="nav-link" id="a-1" href="#div-1">div 1</a></li>',
           '    <li class="nav-item"><a class="nav-link" id="a-2" href="#div-2">div 2</a></li>',
+          '    <li class="nav-item"><a class="nav-link" id="a-2.1" href="#div-2.1">div 2</a></li>',
           '  </ul>',
           '</nav>',
           '<div class="content" style="overflow: auto; height: 50px">',
           '  <div id="div-1" style="height: 100px; padding: 0; margin: 0">div 1</div>',
           '  <div id="div-2" style="height: 200px; padding: 0; margin: 0">div 2</div>',
+          '  <div id="div-2.1" style="height: 200px; padding: 0; margin: 0">div 2.1</div>',
           '</div>'
         ].join('')

What operating system(s) are you seeing the problem on?

macOS

What browser(s) are you seeing the problem on?

Firefox

What version of Bootstrap are you using?

v5.2.3

[pierresouchay]

@larseggert this is weird...

did not take the time to look at, but If I copy-paste:

const parseSelector = selector => {
   if (selector && window.CSS && window.CSS.escape) {
     // document.querySelector needs escaping to handle IDs (html5+) containing for instance /
     selector = selector.replaceAll(/#([^\s"#']+)/g, (match, id) => '#' + CSS.escape(id))
   }

   return selector
 }

in my JS Console in Firefox, and then I type:

parseSelector("#div-2.1")
"#div-2\\.1" 

=> Looks like the correct result to me

[larseggert]

Hm. Apply the patch and run the tests? That should recreate the issue. (Maybe that function isn't called?)

[larseggert]

Here is the test failure I get with the patch above:

Chrome Headless 109.0.5414.87 (Mac OS 10.15.7) ScrollSpy constructor should add the active class to the correct element FAILED
	SyntaxError: Failed to execute 'querySelector' on 'Element': '#div-2.1' is not a valid selector.
	error properties: Object({ code: 12, INDEX_SIZE_ERR: 1, DOMSTRING_SIZE_ERR: 2, HIERARCHY_REQUEST_ERR: 3, WRONG_DOCUMENT_ERR: 4, INVALID_CHARACTER_ERR: 5, NO_DATA_ALLOWED_ERR: 6, NO_MODIFICATION_ALLOWED_ERR: 7, NOT_FOUND_ERR: 8, NOT_SUPPORTED_ERR: 9, INUSE_ATTRIBUTE_ERR: 10, INVALID_STATE_ERR: 11, SYNTAX_ERR: 12, INVALID_MODIFICATION_ERR: 13, NAMESPACE_ERR: 14, INVALID_ACCESS_ERR: 15, VALIDATION_ERR: 16, TYPE_MISMATCH_ERR: 17, SECURITY_ERR: 18, NETWORK_ERR: 19, ABORT_ERR: 20, URL_MISMATCH_ERR: 21, QUOTA_EXCEEDED_ERR: 22, TIMEOUT_ERR: 23, INVALID_NODE_TYPE_ERR: 24, DATA_CLONE_ERR: 25 })
	    at Object.findOne (js/src/dom/selector-engine.js:41:44 <- js/tests/unit/scrollspy.spec.js:9260:46)
	    at ScrollSpy._initializeTargetsAndObservables (js/src/scrollspy.js:211:48 <- js/tests/unit/scrollspy.spec.js:15411:76)
	    at ScrollSpy.refresh (js/src/scrollspy.js:91:10 <- js/tests/unit/scrollspy.spec.js:15227:12)
	    at new ScrollSpy (js/src/scrollspy.js:73:10 <- js/tests/unit/scrollspy.spec.js:15207:12)
	    at js/tests/unit/scrollspy.spec.js:355:27 <- js/tests/unit/scrollspy.spec.js:15743:29
	    at new Promise (<anonymous>)
	    at UserContext.<anonymous> (js/tests/unit/scrollspy.spec.js:338:14 <- js/tests/unit/scrollspy.spec.js:15740:16)
	    at <Jasmine>

[larseggert]

OK, I'm not familiar with the bootstrap code, but to me it looks like scrollspy uses various SelectorEngine functions (find, etc.) that don't seem to call parseSelector at all?

[neurotools]

I have a similar problem but with a tab linking to an external PDF file, using the "file.pdf#page=10" format.

<a href="http://example.com/document.pdf#page=10" target="_blank" class="nav-link">External link</a>

I got JS errors:

Uncaught DOMException: Document.querySelector: '#page=10' is not a valid selector

[chrisyeh96]

+1. I am also seeing this issue on Bootstrap v5.2.3, running Microsoft Edge v112 on Windows 11.

<h2 id="2-partition-normalized-cut">2-partition: Normalized Cut</h2>

causes the error

Uncaught DOMException: Failed to execute 'querySelector' on 'Element': '#2-partition-normalized-cut' is not a valid selector.
    at Object.findOne (https://cdn.jsdelivr.net/npm/bootstrap@5.2.3/dist/js/bootstrap.min.js:6:10283)
    at Be._initializeTargetsAndObservables (https://cdn.jsdelivr.net/npm/bootstrap@5.2.3/dist/js/bootstrap.min.js:6:53059)
    at Be.refresh (https://cdn.jsdelivr.net/npm/bootstrap@5.2.3/dist/js/bootstrap.min.js:6:51311)
    at new Be (https://cdn.jsdelivr.net/npm/bootstrap@5.2.3/dist/js/bootstrap.min.js:6:51184)
    at Be.getOrCreateInstance (https://cdn.jsdelivr.net/npm/bootstrap@5.2.3/dist/js/bootstrap.min.js:6:8552)
    at https://cdn.jsdelivr.net/npm/bootstrap@5.2.3/dist/js/bootstrap.min.js:6:54113
    at i (https://cdn.jsdelivr.net/npm/bootstrap@5.2.3/dist/js/bootstrap.min.js:6:4608)

It seems like #34412 (which was supposed to be closed by #35566) is actually still not solved.

[InsanePrawn]

Can confirm on 5.3.2.

Patching SelectorEngine.findOne() like so fixes (that part of...) scrollspy for me and my #2-foo-bar IDs. (not checked for side-effects!)

    findOne(selector, element = document.documentElement) {
-      return Element.prototype.querySelector.call(element, selector);
+      return Element.prototype.querySelector.call(element, parseSelector(selector));
    },

[Zetanova]

the SelectorEngine.findOne() expects a css selector as valid parameter.

Scrollspy, the caller simply does not escape it:

bootstrap/js/src/scrollspy.js

Line 213 in 0a005b3

const observableSection = SelectorEngine.findOne(decodeURI(anchor.hash), this._element)

It seams of but maybe someone want to find one element
#theId.myClass => SelectorEngine should not try to escape it.