Use word-level diffs for text content too

We already do word-level diffs for comments, so use comment atoms in
the line-based parser.

Fixes #33

CHANGELOG.md

@@ -11,6 +11,8 @@ and `.joker`.
 Fixed an issue with parsing integer values in CSS with units,
 e.g. `123px`.
 
+The line-based parser for text files now uses word-level diffs.
+
 ## 0.8
 
 ### Git integration

src/line_parser.rs

@@ -8,7 +8,8 @@ use crate::{positions::SingleLineSpan, syntax::Syntax};
 pub fn parse<'a>(arena: &'a Arena<Syntax<'a>>, s: &str) -> Vec<&'a Syntax<'a>> {
     let mut res = vec![];
     for (i, line) in s.lines().enumerate() {
-        res.push(Syntax::new_atom(
+        // Mark each line as a comment atom, so we get word-level diffs.
+        res.push(Syntax::new_comment(
             arena,
             vec![SingleLineSpan {
                 line: i.into(),