fix!: make PosTrackingReader encoding-independent

While much of the span logic currently assumes UTF-8, we also
want to support other character encodings, such as e.g. UTF-16
where characters can take up more or less bytes than in UTF-8.

1 files changed, 8 insertions, 2 deletions

diff --git a/src/offset.rs b/src/offset.rs
index 3152c78..bdb069a 100644
--- a/src/offset.rs
+++ b/src/offset.rs
@@ -98,7 +98,7 @@ impl<R: Reader> Reader for PosTrackingReader<R> {
     fn read_char(&mut self) -> Result<Option<char>, Self::Error> {
         match self.reader.read_char()? {
             Some(char) => {
-                self.position += char.len_utf8();
+                self.position += self.reader.len_of_char_in_current_encoding(char);
                 Ok(Some(char))
             }
             None => Ok(None),
@@ -108,10 +108,16 @@ impl<R: Reader> Reader for PosTrackingReader<R> {
     fn try_read_string(&mut self, s: &str, case_sensitive: bool) -> Result<bool, Self::Error> {
         match self.reader.try_read_string(s, case_sensitive)? {
             true => {
-                self.position += s.len();
+                for c in s.chars() {
+                    self.position += self.reader.len_of_char_in_current_encoding(c);
+                }
                 Ok(true)
             }
             false => Ok(false),
         }
     }
+
+    fn len_of_char_in_current_encoding(&self, c: char) -> usize {
+        self.reader.len_of_char_in_current_encoding(c)
+    }
 }