зеркало из https://github.com/mozilla/gecko-dev.git
9b8908c88e
It's a bit strange for the editor to distrust the parser service in this way. The double-checking seems unnecessary, especially given that it is *buggy*: it incorrectly includes `col`, `colgroup`, and `legend` as block elements, but excludes `pre`. (It must only be called in situations where these incorrectly-classified elements are not passed in, otherwise it would have triggered by now.) So this patch removes it. The patch also removes `li` and `pre` from the IsAnyOfHTMLElements() test in HTMLEditor::NodeIsBlockStatic. Contrary to what the comment in NodeIsBlockStatic() says, they *are* identified as block elements by the parser service. --HG-- extra : rebase_source : c9bb3c9e07320fafed17942229c316f41ac2a0a6 |
||
---|---|---|
.. | ||
composer | ||
libeditor | ||
reftests | ||
txmgr | ||
txtsvc | ||
AsyncSpellCheckTestHelper.jsm | ||
moz.build | ||
nsEditorCID.h | ||
nsIContentFilter.idl | ||
nsIDocumentStateListener.idl | ||
nsIEditActionListener.idl | ||
nsIEditor.idl | ||
nsIEditorMailSupport.idl | ||
nsIEditorObserver.idl | ||
nsIEditorSpellCheck.idl | ||
nsIEditorStyleSheets.idl | ||
nsIEditorUtils.idl | ||
nsIHTMLAbsPosEditor.idl | ||
nsIHTMLEditor.idl | ||
nsIHTMLInlineTableEditor.idl | ||
nsIHTMLObjectResizer.idl | ||
nsIPlaintextEditor.idl | ||
nsITableEditor.idl | ||
nsIURIRefObject.idl | ||
nsPIEditorTransaction.idl |