auto merge of #1077 : kmcallister/servo/jsstr, r=jdm

This doesn't resolve the big questions of how Servo will represent strings; it's just about doing the conversion correctly for our existing types.

src/components/script/dom/bindings/utils.rs

@@ -14,17 +14,17 @@ use std::libc;
 use std::ptr;
 use std::ptr::{null, to_unsafe_ptr};
 use std::str;
+use std::vec;
 use std::unstable::raw::Box;
 use js::glue::*;
 use js::glue::{DefineFunctionWithReserved, GetObjectJSClass, RUST_OBJECT_TO_JSVAL};
 use js::glue::{js_IsObjectProxyClass, js_IsFunctionProxyClass, IsProxyHandlerFamily};
 use js::jsapi::{JS_AlreadyHasOwnProperty, JS_NewObject, JS_NewFunction, JS_GetGlobalObject};
 use js::jsapi::{JS_DefineProperties, JS_WrapValue, JS_ForwardGetPropertyTo};
-use js::jsapi::{JS_EncodeString, JS_free, JS_GetStringCharsAndLength};
-use js::jsapi::{JS_GetClass, JS_LinkConstructorAndPrototype};
+use js::jsapi::{JS_GetClass, JS_LinkConstructorAndPrototype, JS_GetStringCharsAndLength};
 use js::jsapi::{JS_GetFunctionPrototype, JS_InternString, JS_GetFunctionObject};
 use js::jsapi::{JS_HasPropertyById, JS_GetPrototype, JS_GetGlobalForObject};
-use js::jsapi::{JS_NewStringCopyN, JS_DefineFunctions, JS_DefineProperty};
+use js::jsapi::{JS_NewUCStringCopyN, JS_DefineFunctions, JS_DefineProperty};
 use js::jsapi::{JS_ValueToString, JS_GetReservedSlot, JS_SetReservedSlot};
 use js::jsapi::{JSContext, JSObject, JSBool, jsid, JSClass, JSNative, JSTracer};
 use js::jsapi::{JSFunctionSpec, JSPropertySpec, JSVal, JSPropertyDescriptor};
@@ -205,25 +205,27 @@ pub fn jsval_to_str(cx: *JSContext, v: JSVal) -> Result<~str, ()> {
             }
         }
 
-        let strbuf = JS_EncodeString(cx, jsstr);
-        let buf = str::raw::from_c_str(strbuf);
-        JS_free(cx, strbuf as *libc::c_void);
-        Ok(buf)
+        let length = 0;
+        let chars = JS_GetStringCharsAndLength(cx, jsstr, &length);
+        do vec::raw::buf_as_slice(chars, length as uint) |char_vec| {
+            Ok(str::from_utf16(char_vec))
+        }
     }
 }
 
 #[fixed_stack_segment]
 pub unsafe fn domstring_to_jsval(cx: *JSContext, string: &DOMString) -> JSVal {
     match string {
-      &None => {
-        JSVAL_NULL
-      }
-      &Some(ref s) => {
-        do s.as_imm_buf |buf, len| {
-            let cbuf = cast::transmute(buf);
-            RUST_STRING_TO_JSVAL(JS_NewStringCopyN(cx, cbuf, len as libc::size_t))
+        &None => JSVAL_NULL,
+        &Some(ref s) => do s.to_utf16().as_imm_buf |buf, len| {
+            let jsstr = JS_NewUCStringCopyN(cx, buf, len as libc::size_t);
+            if jsstr.is_null() {
+                // FIXME: is there something else we should do on failure?
+                JSVAL_NULL
+            } else {
+                RUST_STRING_TO_JSVAL(jsstr)
+            }
         }
-      }
     }
 }
 

src/components/script/html/hubbub_html_parser.rs

@@ -19,6 +19,7 @@ use std::cast;
 use std::cell::Cell;
 use std::comm;
 use std::comm::{Port, SharedChan};
+use std::str;
 use std::str::eq_slice;
 use std::task;
 use std::from_str::FromStr;
@@ -102,7 +103,7 @@ macro_rules! handle_element_base(
 
 
 pub struct JSFile {
-    data: ~[u8],
+    data: ~str,
     url: Url
 }
 
@@ -222,11 +223,11 @@ fn js_script_listener(to_parent: SharedChan<HtmlDiscoveryMessage>,
 
                 let bytes = result_port.recv();
                 if bytes.is_some() {
-                    result_vec.push(JSFile { data: bytes.unwrap(), url: url_clone });
+                    result_vec.push(JSFile { data: str::from_utf8(bytes.unwrap()), url: url_clone });
                 }
             }
             JSTaskNewInlineScript(data, url) => {
-                result_vec.push(JSFile { data: data.into_bytes(), url: url });
+                result_vec.push(JSFile { data: data, url: url });
             }
             JSTaskExit => {
                 break;

src/components/script/script_task.rs

@@ -30,6 +30,7 @@ use std::comm;
 use std::comm::{Port, SharedChan};
 use std::io::read_whole_file;
 use std::ptr;
+use std::str;
 use std::task::{spawn_sched, SingleThreaded};
 use std::util::replace;
 use dom::window::TimerData;
@@ -578,7 +579,7 @@ impl ScriptTask {
             Ok(bytes) => {
                 compartment.define_functions(debug_fns);
                 cx.evaluate_script(compartment.global_obj,
-                                   bytes,
+                                   str::from_utf8(bytes),
                                    url.path.clone(),
                                    1);
             }

src/support/spidermonkey/rust-mozjs

@@ -1 +1 @@
-Subproject commit 1a9a5c3c1462bd3b3e7905ee2834a5ba906f0214
+Subproject commit dc50fb7958312e8ee5b4a03db516bcafd6df3d51

src/test/html/content/test_collections.html

@@ -24,8 +24,10 @@ function check_collection(obj, num, classes, name) {
     }
 }
 
-function check_tag(tagname, num, classes) {
-    check_collection(document.getElementsByTagName(tagname), num, classes, tagname.toUpperCase());
+function check_tag(tagname, num, classes, tagname_upper) {
+    if (tagname_upper === undefined)
+        tagname_upper = tagname.toUpperCase();
+    check_collection(document.getElementsByTagName(tagname), num, classes, tagname_upper);
 }
 
 check_collection(document.links,   1, [HTMLAnchorElement], "A");
@@ -73,8 +75,9 @@ check_tag("track",    1, [HTMLTrackElement]);
 check_tag("audio",    1, [HTMLMediaElement, HTMLAudioElement]);
 check_tag("video",    1, [HTMLMediaElement, HTMLVideoElement]);
 
-// FIXME: Test non-ASCII tag names
-check_tag("foo", 1, [HTMLUnknownElement]);
+// Test non-ASCII tag names.  Uppercasing is ASCII-only per spec:
+// http://dom.spec.whatwg.org/#dom-element-tagname
+check_tag("foo-á", 1, [HTMLUnknownElement], "FOO-á");
 
 finish();
 </script>
@@ -134,7 +137,7 @@ finish();
   <track></track>
 </video>
 
-<foo>hi</foo>
+<foo-á>hi</foo-á>
 
 </body>
 </html>