rustdoc: make langstring parsing more robust

This changes the parsing of the language string
in code examples so that unrecognized examples
are not considered Rust code. This was, for example,
the case when a code example was marked `sh` for shell
code.

This relieves authors of having to mark those samples
as `notrust`.

Also adds recognition of the positive marker `rust`.

By default, unmarked examples are still considered rust.

If any rust-specific tags are seen, code is considered
rust unless marked as "notrust".

Adds test cases for the detection logic.

src/librustdoc/html/markdown.rs

@@ -287,10 +287,7 @@ pub fn find_testable_code(doc: &str, tests: &mut ::test::Collector) {
                 slice::raw::buf_as_slice((*lang).data,
                                        (*lang).size as uint, |lang| {
                     let s = str::from_utf8(lang).unwrap();
-                    (s.contains("should_fail"),
-                     s.contains("no_run"),
-                     s.contains("ignore"),
-                     s.contains("notrust"))
+                    parse_lang_string(s)
                 })
             };
             if notrust { return }
@@ -340,6 +337,35 @@ pub fn find_testable_code(doc: &str, tests: &mut ::test::Collector) {
     }
 }
 
+fn parse_lang_string(string: &str) -> (bool,bool,bool,bool) {
+    let mut seen_rust_tags = false;
+    let mut seen_other_tags = false;
+    let mut should_fail = false;
+    let mut no_run = false;
+    let mut ignore = false;
+    let mut notrust = false;
+
+    let mut tokens = string.as_slice().split(|c: char|
+      !(c == '_' || c == '-' || c.is_alphanumeric())
+    );
+
+    for token in tokens {
+        match token {
+            "" => {},
+            "should_fail" => { should_fail = true; seen_rust_tags = true; },
+            "no_run" => { no_run = true; seen_rust_tags = true; },
+            "ignore" => { ignore = true; seen_rust_tags = true; },
+            "notrust" => { notrust = true; seen_rust_tags = true; },
+            "rust" => { notrust = false; seen_rust_tags = true; },
+            _ => { seen_other_tags = true }
+        }
+    }
+
+    let notrust = notrust || (seen_other_tags && !seen_rust_tags);
+
+    (should_fail, no_run, ignore, notrust)
+}
+
 /// By default this markdown renderer generates anchors for each header in the
 /// rendered document. The anchor name is the contents of the header spearated
 /// by hyphens, and a task-local map is used to disambiguate among duplicate
@@ -367,3 +393,22 @@ impl<'a> fmt::Show for MarkdownWithToc<'a> {
         render(fmt, md.as_slice(), true)
     }
 }
+
+#[cfg(test)]
+mod tests {
+    use super::parse_lang_string;
+
+    #[test]
+    fn test_parse_lang_string() {
+        assert_eq!(parse_lang_string(""), (false,false,false,false))
+        assert_eq!(parse_lang_string("rust"), (false,false,false,false))
+        assert_eq!(parse_lang_string("sh"), (false,false,false,true))
+        assert_eq!(parse_lang_string("notrust"), (false,false,false,true))
+        assert_eq!(parse_lang_string("ignore"), (false,false,true,false))
+        assert_eq!(parse_lang_string("should_fail"), (true,false,false,false))
+        assert_eq!(parse_lang_string("no_run"), (false,true,false,false))
+        assert_eq!(parse_lang_string("{.no_run .example}"), (false,true,false,false))
+        assert_eq!(parse_lang_string("{.sh .should_fail}"), (true,false,false,false))
+        assert_eq!(parse_lang_string("{.example .rust}"), (false,false,false,false))
+    }
+}