Syntax tree definitions with change metadata.
A matched token (an atom, a delimiter, or a comment word).
Different types of strings. We want to diff these the same way, but highlight them differently.
Unlike atoms, tokens can be delimiters like
nodes and return a vec of all the changed positions.
Skip line spans at the beginning or end that have zero width.
set_content_id has already run.
Are there sufficient common words that we should only highlight individual changed words?
Initialise all the fields in
Set all the
SyntaxInfo values for all the
roots on a single
side (LHS or RHS).
For every syntax node in the tree, mark the previous node according to a preorder traversal.
Given the text
content from a comment or strings, split it into
MatchedPos values for the novel and unchanged words.
rhs, but repeat the last item from the shorter