d1f780df0b
also gets the tests running Changes to scripts: * pkg/pkg.gyp: add 'third_party' to list of folders to search * tools/publish_pkg.py: update copyright year * tools/publish_all_pkgs.py: also include pkg/third_party Changes to html5lib: * pubspec.yaml -- removed versions * README.md -- removed some historical notes * added html5lib.status * test/browser -- rename browser_tests to browser_test so test framework finds it * test/parser_test.dart, test/parser_feature_test.dart -- moved dart:io test into parser_test so parser_feature_test can work in browser * test/support.dart -- now finds the data folder relative to entry point script * test/support.dart -- rename "pathos" import to "path" Not changed: * ./test/run.sh still works for testing, in addition to ./tools/test.dart R=dgrove@google.com, ricow@google.com, sigmund@google.com Review URL: https://codereview.chromium.org//22375011 git-svn-id: https://dart.googlecode.com/svn/branches/bleeding_edge/dart@26152 260f80e4-7a28-3924-810f-c04153c831b5
401 lines
11 KiB
Dart
401 lines
11 KiB
Dart
/** Internals to the tree builders. */
|
|
library treebuilder;
|
|
|
|
import 'dart:collection';
|
|
import 'package:html5lib/dom.dart';
|
|
import 'package:source_maps/span.dart' show FileSpan;
|
|
import 'constants.dart';
|
|
import 'list_proxy.dart';
|
|
import 'token.dart';
|
|
import 'utils.dart';
|
|
|
|
// The scope markers are inserted when entering object elements,
|
|
// marquees, table cells, and table captions, and are used to prevent formatting
|
|
// from "leaking" into tables, object elements, and marquees.
|
|
final Node Marker = null;
|
|
|
|
// TODO(jmesserly): this should extend ListBase<Node>, but my simple attempt
|
|
// didn't work.
|
|
class ActiveFormattingElements extends ListProxy<Node> {
|
|
ActiveFormattingElements() : super();
|
|
|
|
// Override the "add" method.
|
|
// TODO(jmesserly): I'd rather not override this; can we do this in the
|
|
// calling code instead?
|
|
void add(Node node) {
|
|
int equalCount = 0;
|
|
if (node != Marker) {
|
|
for (Node element in reversed) {
|
|
if (element == Marker) {
|
|
break;
|
|
}
|
|
if (_nodesEqual(element, node)) {
|
|
equalCount += 1;
|
|
}
|
|
if (equalCount == 3) {
|
|
remove(element);
|
|
break;
|
|
}
|
|
}
|
|
}
|
|
super.add(node);
|
|
}
|
|
}
|
|
|
|
// TODO(jmesserly): this should exist in corelib...
|
|
bool _mapEquals(Map a, Map b) {
|
|
if (a.length != b.length) return false;
|
|
if (a.length == 0) return true;
|
|
|
|
for (var keyA in a.keys) {
|
|
var valB = b[keyA];
|
|
if (valB == null && !b.containsKey(keyA)) {
|
|
return false;
|
|
}
|
|
|
|
if (a[keyA] != valB) {
|
|
return false;
|
|
}
|
|
}
|
|
return true;
|
|
}
|
|
|
|
|
|
bool _nodesEqual(Node node1, Node node2) {
|
|
return node1.nameTuple == node2.nameTuple &&
|
|
_mapEquals(node1.attributes, node2.attributes);
|
|
}
|
|
|
|
/** Basic treebuilder implementation. */
|
|
class TreeBuilder {
|
|
final String defaultNamespace;
|
|
|
|
Document document;
|
|
|
|
final openElements = <Node>[];
|
|
|
|
final activeFormattingElements = new ActiveFormattingElements();
|
|
|
|
Node headPointer;
|
|
|
|
Node formPointer;
|
|
|
|
/**
|
|
* Switch the function used to insert an element from the
|
|
* normal one to the misnested table one and back again
|
|
*/
|
|
bool insertFromTable;
|
|
|
|
TreeBuilder(bool namespaceHTMLElements)
|
|
: defaultNamespace = namespaceHTMLElements ? Namespaces.html : null {
|
|
reset();
|
|
}
|
|
|
|
void reset() {
|
|
openElements.clear();
|
|
activeFormattingElements.clear();
|
|
|
|
//XXX - rename these to headElement, formElement
|
|
headPointer = null;
|
|
formPointer = null;
|
|
|
|
insertFromTable = false;
|
|
|
|
document = new Document();
|
|
}
|
|
|
|
bool elementInScope(target, {String variant}) {
|
|
//If we pass a node in we match that. if we pass a string
|
|
//match any node with that name
|
|
bool exactNode = target is Node && target.nameTuple != null;
|
|
|
|
List listElements1 = scopingElements;
|
|
List listElements2 = const [];
|
|
bool invert = false;
|
|
if (variant != null) {
|
|
switch (variant) {
|
|
case "button":
|
|
listElements2 = const [const Pair(Namespaces.html, "button")];
|
|
break;
|
|
case "list":
|
|
listElements2 = const [const Pair(Namespaces.html, "ol"),
|
|
const Pair(Namespaces.html, "ul")];
|
|
break;
|
|
case "table":
|
|
listElements1 = const [const Pair(Namespaces.html, "html"),
|
|
const Pair(Namespaces.html, "table")];
|
|
break;
|
|
case "select":
|
|
listElements1 = const [const Pair(Namespaces.html, "optgroup"),
|
|
const Pair(Namespaces.html, "option")];
|
|
invert = true;
|
|
break;
|
|
default: assert(false); break;
|
|
}
|
|
}
|
|
|
|
for (Node node in openElements.reversed) {
|
|
if (node.tagName == target && !exactNode ||
|
|
node == target && exactNode) {
|
|
return true;
|
|
} else if (invert !=
|
|
(listElements1.contains(node.nameTuple) ||
|
|
listElements2.contains(node.nameTuple))) {
|
|
return false;
|
|
}
|
|
}
|
|
|
|
assert(false); // We should never reach this point
|
|
}
|
|
|
|
void reconstructActiveFormattingElements() {
|
|
// Within this algorithm the order of steps described in the
|
|
// specification is not quite the same as the order of steps in the
|
|
// code. It should still do the same though.
|
|
|
|
// Step 1: stop the algorithm when there's nothing to do.
|
|
if (activeFormattingElements.length == 0) {
|
|
return;
|
|
}
|
|
|
|
// Step 2 and step 3: we start with the last element. So i is -1.
|
|
int i = activeFormattingElements.length - 1;
|
|
var entry = activeFormattingElements[i];
|
|
if (entry == Marker || openElements.contains(entry)) {
|
|
return;
|
|
}
|
|
|
|
// Step 6
|
|
while (entry != Marker && !openElements.contains(entry)) {
|
|
if (i == 0) {
|
|
//This will be reset to 0 below
|
|
i = -1;
|
|
break;
|
|
}
|
|
i -= 1;
|
|
// Step 5: let entry be one earlier in the list.
|
|
entry = activeFormattingElements[i];
|
|
}
|
|
|
|
while (true) {
|
|
// Step 7
|
|
i += 1;
|
|
|
|
// Step 8
|
|
entry = activeFormattingElements[i];
|
|
|
|
// TODO(jmesserly): optimize this. No need to create a token.
|
|
var cloneToken = new StartTagToken(
|
|
entry.tagName,
|
|
namespace: entry.namespace,
|
|
data: new LinkedHashMap.from(entry.attributes))
|
|
..span = entry.sourceSpan;
|
|
|
|
// Step 9
|
|
var element = insertElement(cloneToken);
|
|
|
|
// Step 10
|
|
activeFormattingElements[i] = element;
|
|
|
|
// Step 11
|
|
if (element == activeFormattingElements.last) {
|
|
break;
|
|
}
|
|
}
|
|
}
|
|
|
|
void clearActiveFormattingElements() {
|
|
var entry = activeFormattingElements.removeLast();
|
|
while (activeFormattingElements.length > 0 && entry != Marker) {
|
|
entry = activeFormattingElements.removeLast();
|
|
}
|
|
}
|
|
|
|
/**
|
|
* Check if an element exists between the end of the active
|
|
* formatting elements and the last marker. If it does, return it, else
|
|
* return null.
|
|
*/
|
|
Node elementInActiveFormattingElements(String name) {
|
|
for (Node item in activeFormattingElements.reversed) {
|
|
// Check for Marker first because if it's a Marker it doesn't have a
|
|
// name attribute.
|
|
if (item == Marker) {
|
|
break;
|
|
} else if (item.tagName == name) {
|
|
return item;
|
|
}
|
|
}
|
|
return null;
|
|
}
|
|
|
|
void insertRoot(Token token) {
|
|
var element = createElement(token);
|
|
openElements.add(element);
|
|
document.nodes.add(element);
|
|
}
|
|
|
|
void insertDoctype(DoctypeToken token) {
|
|
var doctype = new DocumentType(token.name, token.publicId, token.systemId)
|
|
..sourceSpan = token.span;
|
|
document.nodes.add(doctype);
|
|
}
|
|
|
|
void insertComment(StringToken token, [Node parent]) {
|
|
if (parent == null) {
|
|
parent = openElements.last;
|
|
}
|
|
parent.nodes.add(new Comment(token.data)..sourceSpan = token.span);
|
|
}
|
|
|
|
/** Create an element but don't insert it anywhere */
|
|
Element createElement(StartTagToken token) {
|
|
var name = token.name;
|
|
var namespace = token.namespace;
|
|
if (namespace == null) namespace = defaultNamespace;
|
|
var element = new Element(name, namespace)
|
|
..attributes = token.data
|
|
..sourceSpan = token.span;
|
|
return element;
|
|
}
|
|
|
|
Element insertElement(StartTagToken token) {
|
|
if (insertFromTable) return insertElementTable(token);
|
|
return insertElementNormal(token);
|
|
}
|
|
|
|
Element insertElementNormal(StartTagToken token) {
|
|
var name = token.name;
|
|
var namespace = token.namespace;
|
|
if (namespace == null) namespace = defaultNamespace;
|
|
var element = new Element(name, namespace)
|
|
..attributes = token.data
|
|
..sourceSpan = token.span;
|
|
openElements.last.nodes.add(element);
|
|
openElements.add(element);
|
|
return element;
|
|
}
|
|
|
|
Element insertElementTable(token) {
|
|
/** Create an element and insert it into the tree */
|
|
var element = createElement(token);
|
|
if (!tableInsertModeElements.contains(openElements.last.tagName)) {
|
|
return insertElementNormal(token);
|
|
} else {
|
|
// We should be in the InTable mode. This means we want to do
|
|
// special magic element rearranging
|
|
var nodePos = getTableMisnestedNodePosition();
|
|
if (nodePos[1] == null) {
|
|
// TODO(jmesserly): I don't think this is reachable. If insertFromTable
|
|
// is true, there will be a <table> element open, and it always has a
|
|
// parent pointer.
|
|
nodePos[0].nodes.add(element);
|
|
} else {
|
|
nodePos[0].insertBefore(element, nodePos[1]);
|
|
}
|
|
openElements.add(element);
|
|
}
|
|
return element;
|
|
}
|
|
|
|
/** Insert text data. */
|
|
void insertText(String data, FileSpan span) {
|
|
var parent = openElements.last;
|
|
|
|
if (!insertFromTable || insertFromTable &&
|
|
!tableInsertModeElements.contains(openElements.last.tagName)) {
|
|
_insertText(parent, data, span);
|
|
} else {
|
|
// We should be in the InTable mode. This means we want to do
|
|
// special magic element rearranging
|
|
var nodePos = getTableMisnestedNodePosition();
|
|
_insertText(nodePos[0], data, span, nodePos[1]);
|
|
}
|
|
}
|
|
|
|
/**
|
|
* Insert [data] as text in the current node, positioned before the
|
|
* start of node [refNode] or to the end of the node's text.
|
|
*/
|
|
static void _insertText(Node parent, String data, FileSpan span,
|
|
[Element refNode]) {
|
|
var nodes = parent.nodes;
|
|
if (refNode == null) {
|
|
if (nodes.length > 0 && nodes.last is Text) {
|
|
Text last = nodes.last;
|
|
last.value = '${last.value}$data';
|
|
|
|
if (span != null) {
|
|
last.sourceSpan = span.file.span(last.sourceSpan.start.offset,
|
|
span.end.offset);
|
|
}
|
|
} else {
|
|
nodes.add(new Text(data)..sourceSpan = span);
|
|
}
|
|
} else {
|
|
int index = nodes.indexOf(refNode);
|
|
if (index > 0 && nodes[index - 1] is Text) {
|
|
Text last = nodes[index - 1];
|
|
last.value = '${last.value}$data';
|
|
} else {
|
|
nodes.insert(index, new Text(data)..sourceSpan = span);
|
|
}
|
|
}
|
|
}
|
|
|
|
/**
|
|
* Get the foster parent element, and sibling to insert before
|
|
* (or null) when inserting a misnested table node
|
|
*/
|
|
List<Node> getTableMisnestedNodePosition() {
|
|
// The foster parent element is the one which comes before the most
|
|
// recently opened table element
|
|
// XXX - this is really inelegant
|
|
Node lastTable = null;
|
|
Node fosterParent = null;
|
|
var insertBefore = null;
|
|
for (Node elm in openElements.reversed) {
|
|
if (elm.tagName == "table") {
|
|
lastTable = elm;
|
|
break;
|
|
}
|
|
}
|
|
if (lastTable != null) {
|
|
// XXX - we should really check that this parent is actually a
|
|
// node here
|
|
if (lastTable.parent != null) {
|
|
fosterParent = lastTable.parent;
|
|
insertBefore = lastTable;
|
|
} else {
|
|
fosterParent = openElements[openElements.indexOf(lastTable) - 1];
|
|
}
|
|
} else {
|
|
fosterParent = openElements[0];
|
|
}
|
|
return [fosterParent, insertBefore];
|
|
}
|
|
|
|
void generateImpliedEndTags([String exclude]) {
|
|
var name = openElements.last.tagName;
|
|
// XXX td, th and tr are not actually needed
|
|
if (name != exclude && const ["dd", "dt", "li", "option", "optgroup", "p",
|
|
"rp", "rt"].contains(name)) {
|
|
openElements.removeLast();
|
|
// XXX This is not entirely what the specification says. We should
|
|
// investigate it more closely.
|
|
generateImpliedEndTags(exclude);
|
|
}
|
|
}
|
|
|
|
/** Return the final tree. */
|
|
Document getDocument() => document;
|
|
|
|
/** Return the final fragment. */
|
|
DocumentFragment getFragment() {
|
|
//XXX assert innerHTML
|
|
var fragment = new DocumentFragment();
|
|
openElements[0].reparentChildren(fragment);
|
|
return fragment;
|
|
}
|
|
}
|