Fix #363: Do not generate empty Text events

Author: Mingun

Changelog.md

@@ -29,6 +29,7 @@
 - [#393]: Now `event_namespace`, `attribute_namespace` and `read_event_namespaced`
   returns `ResolveResult::Unknown` if prefix was not registered in namespace buffer
 - [#393]: Fix breaking processing after encounter an attribute with a reserved name (started with "xmlns")
+- [#363]: Do not generate empty `Event::Text` events
 
 ### Misc Changes
 

src/reader.rs

@@ -266,9 +266,11 @@ impl<R: BufRead> Reader<R> {
 
         if self.trim_text_start {
             self.reader.skip_whitespace(&mut self.buf_position)?;
-            if self.reader.skip_one(b'<', &mut self.buf_position)? {
-                return self.read_event_buffered(buf);
-            }
+        }
+
+        // If we already at the `<` symbol, do not try to return an empty Text event
+        if self.reader.skip_one(b'<', &mut self.buf_position)? {
+            return self.read_event_buffered(buf);
         }
 
         match self

tests/unit_tests.rs

@@ -168,10 +168,7 @@ fn test_trim_test() {
 
     let mut r = Reader::from_str(txt);
     r.trim_text(false);
-    next_eq!(
-        r, Text, b"", Start, b"a", Text, b"", Start, b"b", Text, b"  ", End, b"b", Text, b"", End,
-        b"a"
-    );
+    next_eq!(r, Start, b"a", Start, b"b", Text, b"  ", End, b"b", End, b"a");
 }
 
 #[test]

tests/xmlrs_reader_tests.rs

@@ -257,9 +257,7 @@ fn issue_98_cdata_ending_with_right_bracket() {
         r#"<hello><![CDATA[Foo [Bar]]]></hello>"#,
         r#"
             |StartElement(hello)
-            |Characters()
             |CData(Foo [Bar])
-            |Characters()
             |EndElement(hello)
             |EndDocument
         "#,
@@ -306,9 +304,7 @@ fn issue_105_unexpected_double_dash() {
         r#"<hello><![CDATA[--]]></hello>"#,
         r#"
             |StartElement(hello)
-            |Characters()
             |CData(--)
-            |Characters()
             |EndElement(hello)
             |EndDocument
         "#,
@@ -359,10 +355,12 @@ fn default_namespace_applies_to_end_elem() {
     );
 }
 
+#[track_caller]
 fn test(input: &str, output: &str, is_short: bool) {
     test_bytes(input.as_bytes(), output.as_bytes(), is_short);
 }
 
+#[track_caller]
 fn test_bytes(input: &[u8], output: &[u8], is_short: bool) {
     // Normalize newlines on Windows to just \n, which is what the reader and
     // writer use.
@@ -380,11 +378,6 @@ fn test_bytes(input: &[u8], output: &[u8], is_short: bool) {
     let mut buf = Vec::new();
     let mut ns_buffer = Vec::new();
 
-    if !is_short {
-        // discard first whitespace
-        reader.read_event(&mut buf).unwrap();
-    }
-
     loop {
         buf.clear();
         let event = reader.read_namespaced_event(&mut buf, &mut ns_buffer);