problem

If a segment contains the following del structure (the whitespaces are important too!):

"This <del>is</del><ins>was</ins> the <ins>castle</ins><del>house</del> of St. Nicholas."

The expected result would be after export:

"This was the castle of St. Nicholas."

but the result is indeed:

"This of St. Nicholas."

reason

Currently we have the following regular expression to remove double whitespaces resulting from deleting a <del></del> where a whitespace was before and one was after:

/ <del[^>]*>.*?<\/del> /

The problem is now, that the regex behaves completely correctly: it searches for the first <del> with leading whitespace, and then for the first </del> with trailing whitespace. The fact that the first del is closed with a </del> without a trailing whitespace is not checked in the regex so far. Then the whole content in between is replaced.