Update to zero-copy* HTML parsing

html5ever now uses the Tendril string type to minimize copying internally,
but Servo still converts from/to `String` at the boundary
(which involves copying).

components/script/Cargo.toml

@@ -76,7 +76,8 @@ num = "0.1.24"
 websocket = "0.12"
 uuid = "0.1.16"
 smallvec = "0.1"
-html5ever = "0.1"
+html5ever = "0.2"
 string_cache = "0.1"
 string_cache_plugin = "0.1"
 euclid = "0.1"
+tendril = "0.1.1"

components/script/dom/servohtmlparser.rs

@@ -312,7 +312,7 @@ impl<'a> PrivateServoHTMLParserHelpers for &'a ServoHTMLParser {
             let mut pending_input = self.pending_input.borrow_mut();
             if !pending_input.is_empty() {
                 let chunk = pending_input.remove(0);
-                self.tokenizer.borrow_mut().feed(chunk);
+                self.tokenizer.borrow_mut().feed(chunk.into());
             } else {
                 self.tokenizer.borrow_mut().run();
             }

components/script/lib.rs

@@ -67,6 +67,7 @@ extern crate url;
 extern crate uuid;
 extern crate string_cache;
 extern crate offscreen_gl_context;
+extern crate tendril;
 
 pub mod cors;
 pub mod document_loader;

components/script/parse/html.rs

@@ -44,6 +44,7 @@ use html5ever::serialize::TraversalScope;
 use html5ever::serialize::TraversalScope::{IncludeNode, ChildrenOnly};
 use html5ever::tree_builder::{TreeSink, QuirksMode, NodeOrText, AppendNode, AppendText, NextParserState};
 use string_cache::QualName;
+use tendril::StrTendril;
 
 trait SinkHelpers {
     fn get_or_create(&self, child: NodeOrText<JS<Node>>) -> Root<Node>;
@@ -55,7 +56,7 @@ impl SinkHelpers for servohtmlparser::Sink {
             AppendNode(n) => n.root(),
             AppendText(t) => {
                 let doc = self.document.root();
-                let text = Text::new(t, doc.r());
+                let text = Text::new(t.into(), doc.r());
                 NodeCast::from_root(text)
             }
         }
@@ -91,16 +92,16 @@ impl<'a> TreeSink for servohtmlparser::Sink {
                                    ElementCreator::ParserCreated);
 
         for attr in attrs.into_iter() {
-            elem.r().set_attribute_from_parser(attr.name, attr.value, None);
+            elem.r().set_attribute_from_parser(attr.name, attr.value.into(), None);
         }
 
         let node = NodeCast::from_ref(elem.r());
         JS::from_ref(node)
     }
 
-    fn create_comment(&mut self, text: String) -> JS<Node> {
+    fn create_comment(&mut self, text: StrTendril) -> JS<Node> {
         let doc = self.document.root();
-        let comment = Comment::new(text, doc.r());
+        let comment = Comment::new(text.into(), doc.r());
         let node = NodeCast::from_root(comment);
         JS::from_rooted(&node)
     }
@@ -137,10 +138,12 @@ impl<'a> TreeSink for servohtmlparser::Sink {
         assert!(parent.r().AppendChild(child.r()).is_ok());
     }
 
-    fn append_doctype_to_document(&mut self, name: String, public_id: String, system_id: String) {
+    fn append_doctype_to_document(&mut self, name: StrTendril, public_id: StrTendril,
+                                  system_id: StrTendril) {
         let doc = self.document.root();
         let doc_node = NodeCast::from_ref(doc.r());
-        let doctype = DocumentType::new(name, Some(public_id), Some(system_id), doc.r());
+        let doctype = DocumentType::new(
+            name.into(), Some(public_id.into()), Some(system_id.into()), doc.r());
         let node: Root<Node> = NodeCast::from_root(doctype);
 
         assert!(doc_node.AppendChild(node.r()).is_ok());
@@ -151,7 +154,7 @@ impl<'a> TreeSink for servohtmlparser::Sink {
         let elem = ElementCast::to_ref(node.r())
             .expect("tried to set attrs on non-Element in HTML parsing");
         for attr in attrs.into_iter() {
-            elem.set_attribute_from_parser(attr.name, attr.value, None);
+            elem.set_attribute_from_parser(attr.name, attr.value.into(), None);
         }
     }
 
@@ -275,7 +278,7 @@ pub fn parse_html(document: &Document,
         ParseContext::Fragment(fc) =>
             ServoHTMLParser::new_for_fragment(Some(url.clone()), document, fc),
     };
-    parser.r().parse_chunk(input);
+    parser.r().parse_chunk(input.into());
 }
 
 // https://html.spec.whatwg.org/multipage/#parsing-html-fragments

components/servo/Cargo.lock

@@ -388,6 +388,14 @@ name = "freetype-sys"
 version = "2.4.11"
 source = "git+https://github.com/servo/libfreetype2#3f22b9dd3be53cdea2a46a0d8eadf72eaeafe2b3"
 
+[[package]]
+name = "futf"
+version = "0.1.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+dependencies = [
+ "mac 0.0.2 (registry+https://github.com/rust-lang/crates.io-index)",
+]
+
 [[package]]
 name = "gcc"
 version = "0.3.8"
@@ -547,7 +555,7 @@ dependencies = [
 
 [[package]]
 name = "html5ever"
-version = "0.1.1"
+version = "0.2.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 dependencies = [
  "html5ever_macros 0.1.0 (registry+https://github.com/rust-lang/crates.io-index)",
@@ -557,6 +565,7 @@ dependencies = [
  "phf_macros 0.7.3 (registry+https://github.com/rust-lang/crates.io-index)",
  "string_cache 0.1.4 (registry+https://github.com/rust-lang/crates.io-index)",
  "string_cache_plugin 0.1.2 (registry+https://github.com/rust-lang/crates.io-index)",
+ "tendril 0.1.1 (registry+https://github.com/rust-lang/crates.io-index)",
  "time 0.1.26 (registry+https://github.com/rust-lang/crates.io-index)",
 ]
 
@@ -1055,7 +1064,7 @@ dependencies = [
  "euclid 0.1.1 (registry+https://github.com/rust-lang/crates.io-index)",
  "fnv 1.0.0 (registry+https://github.com/rust-lang/crates.io-index)",
  "gfx 0.0.1",
- "html5ever 0.1.1 (registry+https://github.com/rust-lang/crates.io-index)",
+ "html5ever 0.2.0 (registry+https://github.com/rust-lang/crates.io-index)",
  "hyper 0.5.2 (registry+https://github.com/rust-lang/crates.io-index)",
  "js 0.1.0 (git+https://github.com/servo/rust-mozjs)",
  "libc 0.1.8 (registry+https://github.com/rust-lang/crates.io-index)",
@@ -1074,6 +1083,7 @@ dependencies = [
  "string_cache 0.1.4 (registry+https://github.com/rust-lang/crates.io-index)",
  "string_cache_plugin 0.1.2 (registry+https://github.com/rust-lang/crates.io-index)",
  "style 0.0.1",
+ "tendril 0.1.1 (registry+https://github.com/rust-lang/crates.io-index)",
  "time 0.1.26 (registry+https://github.com/rust-lang/crates.io-index)",
  "unicase 0.1.0 (registry+https://github.com/rust-lang/crates.io-index)",
  "url 0.2.35 (registry+https://github.com/rust-lang/crates.io-index)",
@@ -1234,6 +1244,16 @@ name = "tenacious"
 version = "0.0.5"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 
+[[package]]
+name = "tendril"
+version = "0.1.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+dependencies = [
+ "encoding 0.2.32 (registry+https://github.com/rust-lang/crates.io-index)",
+ "futf 0.1.0 (registry+https://github.com/rust-lang/crates.io-index)",
+ "mac 0.0.2 (registry+https://github.com/rust-lang/crates.io-index)",
+]
+
 [[package]]
 name = "time"
 version = "0.1.26"

ports/cef/Cargo.lock

@@ -387,6 +387,14 @@ name = "freetype-sys"
 version = "2.4.11"
 source = "git+https://github.com/servo/libfreetype2#3f22b9dd3be53cdea2a46a0d8eadf72eaeafe2b3"
 
+[[package]]
+name = "futf"
+version = "0.1.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+dependencies = [
+ "mac 0.0.2 (registry+https://github.com/rust-lang/crates.io-index)",
+]
+
 [[package]]
 name = "gcc"
 version = "0.3.8"
@@ -539,7 +547,7 @@ dependencies = [
 
 [[package]]
 name = "html5ever"
-version = "0.1.1"
+version = "0.2.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 dependencies = [
  "html5ever_macros 0.1.0 (registry+https://github.com/rust-lang/crates.io-index)",
@@ -549,6 +557,7 @@ dependencies = [
  "phf_macros 0.7.3 (registry+https://github.com/rust-lang/crates.io-index)",
  "string_cache 0.1.4 (registry+https://github.com/rust-lang/crates.io-index)",
  "string_cache_plugin 0.1.2 (registry+https://github.com/rust-lang/crates.io-index)",
+ "tendril 0.1.1 (registry+https://github.com/rust-lang/crates.io-index)",
  "time 0.1.26 (registry+https://github.com/rust-lang/crates.io-index)",
 ]
 
@@ -1035,7 +1044,7 @@ dependencies = [
  "euclid 0.1.1 (registry+https://github.com/rust-lang/crates.io-index)",
  "fnv 1.0.0 (registry+https://github.com/rust-lang/crates.io-index)",
  "gfx 0.0.1",
- "html5ever 0.1.1 (registry+https://github.com/rust-lang/crates.io-index)",
+ "html5ever 0.2.0 (registry+https://github.com/rust-lang/crates.io-index)",
  "hyper 0.5.2 (registry+https://github.com/rust-lang/crates.io-index)",
  "js 0.1.0 (git+https://github.com/servo/rust-mozjs)",
  "libc 0.1.8 (registry+https://github.com/rust-lang/crates.io-index)",
@@ -1054,6 +1063,7 @@ dependencies = [
  "string_cache 0.1.4 (registry+https://github.com/rust-lang/crates.io-index)",
  "string_cache_plugin 0.1.2 (registry+https://github.com/rust-lang/crates.io-index)",
  "style 0.0.1",
+ "tendril 0.1.1 (registry+https://github.com/rust-lang/crates.io-index)",
  "time 0.1.26 (registry+https://github.com/rust-lang/crates.io-index)",
  "unicase 0.1.0 (registry+https://github.com/rust-lang/crates.io-index)",
  "url 0.2.35 (registry+https://github.com/rust-lang/crates.io-index)",
@@ -1218,6 +1228,16 @@ name = "tenacious"
 version = "0.0.5"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 
+[[package]]
+name = "tendril"
+version = "0.1.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+dependencies = [
+ "encoding 0.2.32 (registry+https://github.com/rust-lang/crates.io-index)",
+ "futf 0.1.0 (registry+https://github.com/rust-lang/crates.io-index)",
+ "mac 0.0.2 (registry+https://github.com/rust-lang/crates.io-index)",
+]
+
 [[package]]
 name = "time"
 version = "0.1.26"

ports/gonk/Cargo.lock

@@ -374,6 +374,14 @@ name = "freetype-sys"
 version = "2.4.11"
 source = "git+https://github.com/servo/libfreetype2#3f22b9dd3be53cdea2a46a0d8eadf72eaeafe2b3"
 
+[[package]]
+name = "futf"
+version = "0.1.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+dependencies = [
+ "mac 0.0.2 (registry+https://github.com/rust-lang/crates.io-index)",
+]
+
 [[package]]
 name = "gcc"
 version = "0.3.8"
@@ -473,7 +481,7 @@ dependencies = [
 
 [[package]]
 name = "html5ever"
-version = "0.1.1"
+version = "0.2.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 dependencies = [
  "html5ever_macros 0.1.0 (registry+https://github.com/rust-lang/crates.io-index)",
@@ -483,6 +491,7 @@ dependencies = [
  "phf_macros 0.7.3 (registry+https://github.com/rust-lang/crates.io-index)",
  "string_cache 0.1.4 (registry+https://github.com/rust-lang/crates.io-index)",
  "string_cache_plugin 0.1.2 (registry+https://github.com/rust-lang/crates.io-index)",
+ "tendril 0.1.1 (registry+https://github.com/rust-lang/crates.io-index)",
  "time 0.1.26 (registry+https://github.com/rust-lang/crates.io-index)",
 ]
 
@@ -943,7 +952,7 @@ dependencies = [
  "euclid 0.1.1 (registry+https://github.com/rust-lang/crates.io-index)",
  "fnv 1.0.0 (registry+https://github.com/rust-lang/crates.io-index)",
  "gfx 0.0.1",
- "html5ever 0.1.1 (registry+https://github.com/rust-lang/crates.io-index)",
+ "html5ever 0.2.0 (registry+https://github.com/rust-lang/crates.io-index)",
  "hyper 0.5.2 (registry+https://github.com/rust-lang/crates.io-index)",
  "js 0.1.0 (git+https://github.com/servo/rust-mozjs)",
  "libc 0.1.8 (registry+https://github.com/rust-lang/crates.io-index)",
@@ -962,6 +971,7 @@ dependencies = [
  "string_cache 0.1.4 (registry+https://github.com/rust-lang/crates.io-index)",
  "string_cache_plugin 0.1.2 (registry+https://github.com/rust-lang/crates.io-index)",
  "style 0.0.1",
+ "tendril 0.1.1 (registry+https://github.com/rust-lang/crates.io-index)",
  "time 0.1.26 (registry+https://github.com/rust-lang/crates.io-index)",
  "unicase 0.1.0 (registry+https://github.com/rust-lang/crates.io-index)",
  "url 0.2.35 (registry+https://github.com/rust-lang/crates.io-index)",
@@ -1116,6 +1126,16 @@ name = "tenacious"
 version = "0.0.5"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 
+[[package]]
+name = "tendril"
+version = "0.1.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+dependencies = [
+ "encoding 0.2.32 (registry+https://github.com/rust-lang/crates.io-index)",
+ "futf 0.1.0 (registry+https://github.com/rust-lang/crates.io-index)",
+ "mac 0.0.2 (registry+https://github.com/rust-lang/crates.io-index)",
+]
+
 [[package]]
 name = "time"
 version = "0.1.26"