Automatic entity conversion screwing up search and replacerch and replace
The older versions of Sigil didn't automatically convert entities. I'm not saying I'm against the idea of automatically converting entities, but the way it's currently implemented makes things very confusing when using search and replace.
I had a document that started as markdown, and converted -- to —. The em-dash was directly encoded, not an entity in the actual ePub.
I just went into Sigil to give all the paragraphs starting with em-dash negative indents. In Sigil code view the em-dash was automatically displayed as an entity. Search and Replace for the entity and surrounding pattern worked fine for the displayed html flow, but when I expanded the search to all html files, the search failed to find any more occurrences.
It was only when I changed the search expression to an actual em-dash that it then found all the instances in the document and replaced them - interestingly enough the actual em-dash search expression also found/replaced the entity em-dash text in the actively open xhtml file.
I did like that searching for the actual unicode character caught both encoded and non-encoded characters. However it's extremely unintuitive that searching for the entity didn't catch both versions, seeing as Sigil will never show you the unicode version.
Is this a bug or a known issue?
Edit - appreciate if a forum mod can fix the typo in my title, not sure how the extra text got there