Document.getLinkedPages() returns non-link entries

If there is JavaScript on a page, the getLinkedPages method often interprets javascript subscript expressions as XWiki document names and returns them.

This is irrespective of whether the

tag surrounds the script, since the search is a simple regex for anything inside a pair of brackets that lexically qualifies as an XWiki doc name on the unrendered content.

XWiki.XWikiCommonJavaScript (at least in 0.9.840) contains several examples.