这是奇怪的...
我没有花时间去查看，但如果我在Firefox的JS控制台中复制粘贴：

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
 }

然后我输入：

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

=> 对我来说看起来是正确的结果

应用补丁并运行测试？这应该会重新创建问题(也许那个函数没有被调用？)

以下是我在使用上述补丁后遇到的测试失败：

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>

好的，我对 Bootstrap 代码不熟悉，但在我看来，scrollspy似乎使用了各种SelectorEngine函数(如find等),而这些函数似乎根本不调用parseSelector?

我遇到了类似的问题，但是链接到一个外部PDF文件，使用"file.pdf#page=10"的格式。

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

我得到了JS错误：

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

我也在Bootstrap v5.2.3上遇到了这个问题，运行在Windows 11上的Microsoft Edge v112。

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

导致了错误

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)

看起来 #34412 (本应由 #35566 关闭)实际上仍然没有解决。

可以在5.3.2上确认。
像这样修补SelectorEngine.findOne()可以解决(那部分的)scrollspy问题，对我和我的#2-foo-bar ID也是如此。(没有检查副作用！)

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