3 files changed, 113 insertions, 0 deletions
diff --git a/tests/files/test.html b/tests/files/test.html
new file mode 100644
index 0000000..0dcbdbf
--- /dev/null
+++ b/tests/files/test.html
@@ -0,0 +1,7 @@
+This is a file.
+
+Here is a tag: <strong >very cool</strong>
+
+Tags can have attributes: <div id = foo >...</div>
+
+Attribute values can be quoted: <input name = 'age' type = "number">
diff --git a/tests/files/test.out b/tests/files/test.out
new file mode 100644
index 0000000..7127ebc
--- /dev/null
+++ b/tests/files/test.out
@@ -0,0 +1,17 @@
+note:
+  ┌─ test.html:3:17
+  │
+3 │ Here is a tag: <strong >very cool</strong>
+  │                 ^^^^^^             ^^^^^^ EndTag
+  │                 │
+  │                 StartTag
+4 │
+5 │ Tags can have attributes: <div id = foo >...</div>
+  │                                ^^   ^^^ attribute value
+  │                                │
+  │                                attribute name
+6 │
+7 │ Attribute values can be quoted: <input name = 'age' type = "number">
+  │                                                ^^^          ^^^^^^ in double quotes
+  │                                                │
+  │                                                in single quotes
diff --git a/tests/spans.rs b/tests/spans.rs
new file mode 100644
index 0000000..bfa42f6
--- /dev/null
+++ b/tests/spans.rs
@@ -0,0 +1,89 @@
+#![cfg(feature = "spans")]
+use std::include_str;
+
+use codespan_reporting::{
+    self,
+    diagnostic::{Diagnostic, Label},
+    files::SimpleFiles,
+    term::{self, termcolor::Buffer},
+};
+use html5tokenizer::{
+    BufferQueue, Tag, Token, TokenSink, TokenSinkResult, Tokenizer, TokenizerOpts,
+};
+
+#[derive(Default)]
+struct TagSink {
+    tags: Vec<Tag>,
+}
+
+impl TokenSink for TagSink {
+    fn process_token(&mut self, token: Token, _line_number: u64) -> TokenSinkResult {
+        if let Token::TagToken(tag) = token {
+            self.tags.push(tag);
+        }
+        TokenSinkResult::Continue
+    }
+}
+
+#[test]
+fn test() {
+    let sink = TagSink::default();
+
+    let mut input = BufferQueue::new();
+    let text = include_str!("files/test.html");
+    input.push_back(text.to_string());
+
+    let mut tok = Tokenizer::new(sink, TokenizerOpts::default());
+    let _ = tok.feed(&mut input);
+
+    let mut files = SimpleFiles::new();
+    let file_id = files.add("test.html", text);
+    let mut labels = Vec::new();
+
+    let tags = tok.sink.tags;
+    for tag in &tags[..2] {
+        labels.push(
+            Label::primary(file_id, tag.name_span.clone()).with_message(format!("{:?}", tag.kind)),
+        );
+    }
+    labels.push(
+        Label::primary(file_id, tags[2].attrs[0].name_span.clone()).with_message("attribute name"),
+    );
+    labels.push(
+        Label::primary(file_id, tags[2].attrs[0].value_span.clone())
+            .with_message("attribute value"),
+    );
+    labels.push(
+        Label::primary(file_id, tags[4].attrs[0].value_span.clone())
+            .with_message("in single quotes"),
+    );
+    labels.push(
+        Label::primary(file_id, tags[4].attrs[1].value_span.clone())
+            .with_message("in double quotes"),
+    );
+    let diagnostic = Diagnostic::note().with_labels(labels);
+
+    let mut writer = Buffer::no_color();
+    let config = codespan_reporting::term::Config::default();
+    term::emit(&mut writer, &config, &files, &diagnostic).unwrap();
+
+    let actual = remove_trailing_spaces(std::str::from_utf8(writer.as_slice()).unwrap());
+    let expected = include_str!("files/test.out");
+
+    if actual != expected {
+        println!(
+            "EXPECTED:\n{banner}\n{expected}{banner}\n\nACTUAL OUTPUT:\n{banner}\n{actual}{banner}",
+            banner = "-".repeat(30),
+            expected = expected,
+            actual = actual
+        );
+        panic!("failed");
+    }
+}
+
+fn remove_trailing_spaces(text: &str) -> String {
+    text.lines()
+        .map(|l| l.trim_end())
+        .collect::<Vec<_>>()
+        .join("\n")
+}