зеркало из https://github.com/mozilla/gecko-dev.git
servo: Parse HTML attributes and support images of different widths and heights
Source-Repo: https://github.com/servo/servo Source-Revision: d4ea1477a764f992ae8faae684287cc2917f9c63
This commit is contained in:
Patrick Walton
Родитель
23e895c155
Коммит
a566a859bf
|
|
@ -1,14 +1,22 @@
|
|||
import comm::{port, chan};
|
||||
|
||||
enum parse_state {
|
||||
ps_normal,
|
||||
ps_tag
|
||||
}
|
||||
|
||||
type parser = {
|
||||
mut lookahead: option<char_or_eof>,
|
||||
mut state: parse_state,
|
||||
reader: io::reader
|
||||
};
|
||||
|
||||
enum token {
|
||||
to_start_tag(str),
|
||||
to_start_opening_tag(str),
|
||||
to_end_opening_tag,
|
||||
to_end_tag(str),
|
||||
to_text(str),
|
||||
to_attr(str, str),
|
||||
to_doctype,
|
||||
to_eof
|
||||
}
|
||||
|
|
@ -18,6 +26,18 @@ enum char_or_eof {
|
|||
coe_eof
|
||||
}
|
||||
|
||||
impl u8_methods for u8 {
|
||||
fn is_alpha() -> bool {
|
||||
ret (self >= ('A' as u8) && self <= ('Z' as u8)) ||
|
||||
(self >= ('a' as u8) && self <= ('z' as u8));
|
||||
}
|
||||
}
|
||||
|
||||
impl u8_vec_methods for [u8] {
|
||||
fn to_str() -> str { ret str::from_bytes(self); }
|
||||
fn to_str_token() -> token { ret to_text(self.to_str()); }
|
||||
}
|
||||
|
||||
impl methods for parser {
|
||||
fn get() -> char_or_eof {
|
||||
alt self.lookahead {
|
||||
|
|
@ -62,8 +82,7 @@ impl methods for parser {
|
|||
loop {
|
||||
alt self.get() {
|
||||
coe_char(c) {
|
||||
if (c >= ('A' as u8) && c <= ('Z' as u8)) ||
|
||||
(c >= ('a' as u8) && c <= ('z' as u8)) {
|
||||
if (c.is_alpha()) {
|
||||
result += [c];
|
||||
} else if result.len() == 0u {
|
||||
self.parse_err("expected ident");
|
||||
|
|
@ -112,6 +131,14 @@ impl methods for parser {
|
|||
coe_eof { ret to_eof; }
|
||||
}
|
||||
|
||||
ret alt self.state {
|
||||
ps_normal { self.parse_in_normal_state(ch) }
|
||||
ps_tag { self.parse_in_tag_state(ch) }
|
||||
}
|
||||
}
|
||||
|
||||
fn parse_in_normal_state(c: u8) -> token {
|
||||
let mut ch = c;
|
||||
if ch == ('<' as u8) {
|
||||
alt self.get() {
|
||||
coe_char(c) { ch = c; }
|
||||
|
|
@ -139,8 +166,9 @@ impl methods for parser {
|
|||
self.eat_whitespace();
|
||||
let ident = self.parse_ident();
|
||||
self.eat_whitespace();
|
||||
self.expect('>' as u8);
|
||||
ret to_start_tag(ident);
|
||||
|
||||
self.state = ps_tag;
|
||||
ret to_start_opening_tag(ident);
|
||||
}
|
||||
|
||||
// Make a text node.
|
||||
|
|
@ -150,18 +178,66 @@ impl methods for parser {
|
|||
coe_char(c) {
|
||||
if c == ('<' as u8) {
|
||||
self.unget(c);
|
||||
ret to_text(str::from_bytes(s));
|
||||
ret s.to_str_token();
|
||||
}
|
||||
s += [c];
|
||||
}
|
||||
coe_eof { ret to_text(str::from_bytes(s)); }
|
||||
coe_eof { ret s.to_str_token(); }
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
fn parse_in_tag_state(c: u8) -> token {
|
||||
let mut ch = c;
|
||||
|
||||
if ch == ('>' as u8) {
|
||||
self.state = ps_normal;
|
||||
ret to_end_opening_tag;
|
||||
}
|
||||
|
||||
if !ch.is_alpha() {
|
||||
fail "expected alphabetical in tag";
|
||||
}
|
||||
|
||||
// Parse an attribute.
|
||||
let mut attribute_name = [ch];
|
||||
loop {
|
||||
alt self.get() {
|
||||
coe_char(c) {
|
||||
if c == ('=' as u8) { break; }
|
||||
attribute_name += [c];
|
||||
}
|
||||
coe_eof {
|
||||
ret to_attr(attribute_name.to_str(),
|
||||
attribute_name.to_str()); }
|
||||
}
|
||||
}
|
||||
|
||||
// Parse the attribute value.
|
||||
self.expect('"' as u8);
|
||||
let mut attribute_value = [];
|
||||
loop {
|
||||
alt self.get() {
|
||||
coe_char(c) {
|
||||
if c == ('"' as u8) { break; }
|
||||
attribute_value += [c];
|
||||
}
|
||||
coe_eof {
|
||||
ret to_attr(attribute_name.to_str(),
|
||||
attribute_value.to_str());
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
// Eat whitespace.
|
||||
self.eat_whitespace();
|
||||
|
||||
ret to_attr(attribute_name.to_str(), attribute_value.to_str());
|
||||
}
|
||||
}
|
||||
|
||||
fn parser(reader: io::reader) -> parser {
|
||||
ret { mut lookahead: none, reader: reader };
|
||||
ret { mut lookahead: none, mut state: ps_normal, reader: reader };
|
||||
}
|
||||
|
||||
fn spawn_parser_task(filename: str) -> port<token> {
|
||||
|
|
|
|||
|
|
@ -1,4 +1,4 @@
|
|||
// Constructs a DOM tree from an incoming token stream.
|
||||
#[doc="Constructs a DOM tree from an incoming token stream."]
|
||||
|
||||
import dom::rcu::writer_methods;
|
||||
import dom::base::{methods, rd_tree_ops, wr_tree_ops};
|
||||
|
|
@ -7,6 +7,35 @@ import parser = parser::html;
|
|||
import html::token;
|
||||
import gfx::geom;
|
||||
|
||||
fn link_up_attribute(scope: dom::node_scope, node: dom::node, key: str,
|
||||
value: str) {
|
||||
// TODO: Implement atoms so that we don't always perform string
|
||||
// comparisons.
|
||||
// FIXME: This is wrong... we should not have DIV and IMG be separate types
|
||||
// of nodes and instead have them inherit from Element, obviously.
|
||||
scope.rd(node) {
|
||||
|node_contents|
|
||||
alt node_contents.kind {
|
||||
dom::nk_img(dims) if key == "width" {
|
||||
alt int::from_str(value) {
|
||||
none { /* drop on the floor */ }
|
||||
some(s) { dims.width = geom::px_to_au(s); }
|
||||
}
|
||||
}
|
||||
dom::nk_img(dims) if key == "height" {
|
||||
alt int::from_str(value) {
|
||||
none { /* drop on the floor */ }
|
||||
some(s) { dims.height = geom::px_to_au(s); }
|
||||
}
|
||||
}
|
||||
dom::nk_div | dom::nk_img(*) { /* drop on the floor */ }
|
||||
dom::nk_text(*) {
|
||||
fail "attempt to link up an attribute to a text node"
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
fn build_dom(scope: dom::node_scope,
|
||||
stream: port<token>) -> dom::node {
|
||||
// The current reference node.
|
||||
|
|
@ -16,14 +45,14 @@ fn build_dom(scope: dom::node_scope,
|
|||
#debug["token=%?", token];
|
||||
alt token {
|
||||
parser::to_eof { break; }
|
||||
parser::to_start_tag("div") {
|
||||
parser::to_start_opening_tag("div") {
|
||||
#debug["DIV"];
|
||||
let new_node = scope.new_node(
|
||||
dom::nk_div);
|
||||
scope.add_child(cur, new_node);
|
||||
cur = new_node;
|
||||
}
|
||||
parser::to_start_tag("img") {
|
||||
parser::to_start_opening_tag("img") {
|
||||
#debug["IMG"];
|
||||
let new_node = scope.new_node(
|
||||
dom::nk_img({mut width: geom::px_to_au(100),
|
||||
|
|
@ -31,9 +60,16 @@ fn build_dom(scope: dom::node_scope,
|
|||
scope.add_child(cur, new_node);
|
||||
cur = new_node;
|
||||
}
|
||||
parser::to_start_tag(t) {
|
||||
parser::to_start_opening_tag(t) {
|
||||
fail ("Unrecognized tag: " + t);
|
||||
}
|
||||
parser::to_attr(key, value) {
|
||||
#debug["attr: %? = %?", key, value];
|
||||
link_up_attribute(scope, cur, key, value);
|
||||
}
|
||||
parser::to_end_opening_tag {
|
||||
#debug("end opening tag");
|
||||
}
|
||||
parser::to_end_tag(_) {
|
||||
// TODO: Assert that the closing tag has the right name.
|
||||
// TODO: Fail more gracefully (i.e. according to the HTML5
|
||||
|
|
@ -54,3 +90,4 @@ fn build_dom(scope: dom::node_scope,
|
|||
}
|
||||
ret cur;
|
||||
}
|
||||
|
||||
|
|
|
|||
Загрузка…
Ссылка в новой задаче