From 153acbee5650b9b006c7de215fb1421a64516fb4 Mon Sep 17 00:00:00 2001
From: Martin Fischer <martin@push-f.com>
Date: Fri, 1 Sep 2023 14:59:08 +0200
Subject: 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.
---
 CHANGELOG.md  |  2 ++
 src/offset.rs | 10 ++++++++--
 src/reader.rs | 11 +++++++++++
 3 files changed, 21 insertions(+), 2 deletions(-)

diff --git a/CHANGELOG.md b/CHANGELOG.md
index 28431e4..5025516 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -29,6 +29,8 @@
 
   * `AttributeOwned`: The `value_offset` field has been replaced with `value_span`.
 
+* Added required `len_of_char_in_current_encoding` method to `Reader` trait.
+
 * Added missing `R: Position<O>` bounds for `Tokenizer`/`NaiveParser` constructors.  
   (If you are able to construct a Tokenizer/NaiveParser,
   you should be able to iterate over it.)
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)
+    }
 }
diff --git a/src/reader.rs b/src/reader.rs
index 15b9224..3c39b16 100644
--- a/src/reader.rs
+++ b/src/reader.rs
@@ -30,6 +30,9 @@ pub trait Reader {
     /// the input stream and returns `true`. If not, it does nothing and returns `false`.
     // TODO: document a maximum s length that may be assumed (depends on longest named character reference ... which may change?)
     fn try_read_string(&mut self, s: &str, case_sensitive: bool) -> Result<bool, Self::Error>;
+
+    /// Returns the number of bytes that the given character takes up in the current character encoding.
+    fn len_of_char_in_current_encoding(&self, c: char) -> usize;
 }
 
 /// An object that can be converted into a [`Reader`].
@@ -96,6 +99,10 @@ impl<'a> Reader for StringReader<'a> {
 
         Ok(false)
     }
+
+    fn len_of_char_in_current_encoding(&self, c: char) -> usize {
+        c.len_utf8()
+    }
 }
 
 impl<'a> IntoReader<'a> for &'a str {
@@ -248,6 +255,10 @@ impl<R: Read> Reader for BufReadReader<R> {
 
         Ok(true)
     }
+
+    fn len_of_char_in_current_encoding(&self, c: char) -> usize {
+        c.len_utf8()
+    }
 }
 
 impl<'a, R: Read + 'a> IntoReader<'a> for BufReader<R> {
-- 
cgit v1.2.3