blob: 73868dcdce8ce74b63ddcbb5b897cee313d4a2c0 [file] [log] [blame]
/**
* Copyright (C) 1999 Lars Knoll (knoll@kde.org)
* (C) 1999 Antti Koivisto (koivisto@kde.org)
* Copyright (C) 2003, 2004, 2005, 2006, 2007 Apple Inc. All rights reserved.
*
* This library is free software; you can redistribute it and/or
* modify it under the terms of the GNU Library General Public
* License as published by the Free Software Foundation; either
* version 2 of the License, or (at your option) any later version.
*
* This library is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* Library General Public License for more details.
*
* You should have received a copy of the GNU Library General Public License
* along with this library; see the file COPYING.LIB. If not, write to
* the Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor,
* Boston, MA 02110-1301, USA.
*
* Portions are Copyright (C) 2002 Netscape Communications Corporation.
* Other contributors: David Baron <dbaron@fas.harvard.edu>
*
* This library is free software; you can redistribute it and/or
* modify it under the terms of the GNU Lesser General Public
* License as published by the Free Software Foundation; either
* version 2.1 of the License, or (at your option) any later version.
*
* This library is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public
* License along with this library; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
*
* Alternatively, the document type parsing portions of this file may be used
* under the terms of either the Mozilla Public License Version 1.1, found at
* http://www.mozilla.org/MPL/ (the "MPL") or the GNU General Public
* License Version 2.0, found at http://www.fsf.org/copyleft/gpl.html
* (the "GPL"), in which case the provisions of the MPL or the GPL are
* applicable instead of those above. If you wish to allow use of your
* version of this file only under the terms of one of those two
* licenses (the MPL or the GPL) and not to allow others to use your
* version of this file under the LGPL, indicate your decision by
* deleting the provisions above and replace them with the notice and
* other provisions required by the MPL or the GPL, as the case may be.
* If you do not delete the provisions above, a recipient may use your
* version of this file under any of the LGPL, the MPL or the GPL.
*/
#include "config.h"
#include "HTMLDocument.h"
#include "CSSPropertyNames.h"
#include "CSSStyleSelector.h"
#include "CString.h"
#include "CookieJar.h"
#include "DocumentLoader.h"
#include "DocumentType.h"
#include "ExceptionCode.h"
#include "Frame.h"
#include "FrameLoader.h"
#include "FrameView.h"
#include "HTMLBodyElement.h"
#include "HTMLElement.h"
#include "HTMLElementFactory.h"
#include "HTMLNames.h"
#include "HTMLTokenizer.h"
#include "InspectorController.h"
#include "KURL.h"
#include "Page.h"
#include "DocTypeStrings.cpp"
namespace WebCore {
using namespace HTMLNames;
HTMLDocument::HTMLDocument(DOMImplementation* implementation, Frame* frame)
: Document(implementation, frame)
{
clearXMLVersion();
}
HTMLDocument::~HTMLDocument()
{
}
int HTMLDocument::width()
{
updateLayoutIgnorePendingStylesheets();
FrameView* frameView = view();
return frameView ? frameView->contentsWidth() : 0;
}
int HTMLDocument::height()
{
updateLayoutIgnorePendingStylesheets();
FrameView* frameView = view();
return frameView ? frameView->contentsHeight() : 0;
}
String HTMLDocument::dir()
{
HTMLElement* b = body();
if (!b)
return String();
return b->dir();
}
void HTMLDocument::setDir(const String& value)
{
HTMLElement* b = body();
if (b)
b->setDir(value);
}
String HTMLDocument::designMode() const
{
return inDesignMode() ? "on" : "off";
}
void HTMLDocument::setDesignMode(const String& value)
{
InheritedBool mode;
if (equalIgnoringCase(value, "on"))
mode = on;
else if (equalIgnoringCase(value, "off"))
mode = off;
else
mode = inherit;
Document::setDesignMode(mode);
}
String HTMLDocument::compatMode() const
{
return inCompatMode() ? "BackCompat" : "CSS1Compat";
}
String HTMLDocument::bgColor()
{
HTMLElement* b = body();
HTMLBodyElement* bodyElement = (b && b->hasTagName(bodyTag)) ? static_cast<HTMLBodyElement*>(b) : 0;
if (!bodyElement)
return String();
return bodyElement->bgColor();
}
void HTMLDocument::setBgColor(const String& value)
{
HTMLElement* b = body();
HTMLBodyElement* bodyElement = (b && b->hasTagName(bodyTag)) ? static_cast<HTMLBodyElement*>(b) : 0;
if (bodyElement)
bodyElement->setBgColor(value);
}
String HTMLDocument::fgColor()
{
HTMLElement* b = body();
HTMLBodyElement* bodyElement = (b && b->hasTagName(bodyTag)) ? static_cast<HTMLBodyElement*>(b) : 0;
if (!bodyElement)
return String();
return bodyElement->text();
}
void HTMLDocument::setFgColor(const String& value)
{
HTMLElement* b = body();
HTMLBodyElement* bodyElement = (b && b->hasTagName(bodyTag)) ? static_cast<HTMLBodyElement*>(b) : 0;
if (bodyElement)
bodyElement->setText(value);
}
String HTMLDocument::alinkColor()
{
HTMLElement* b = body();
HTMLBodyElement* bodyElement = (b && b->hasTagName(bodyTag)) ? static_cast<HTMLBodyElement*>(b) : 0;
if (!bodyElement)
return String();
return bodyElement->aLink();
}
void HTMLDocument::setAlinkColor(const String& value)
{
HTMLElement* b = body();
HTMLBodyElement* bodyElement = (b && b->hasTagName(bodyTag)) ? static_cast<HTMLBodyElement*>(b) : 0;
if (bodyElement) {
// This check is a bit silly, but some benchmarks like to set the
// document's link colors over and over to the same value and we
// don't want to incur a style update each time.
if (bodyElement->aLink() != value)
bodyElement->setALink(value);
}
}
String HTMLDocument::linkColor()
{
HTMLElement* b = body();
HTMLBodyElement* bodyElement = (b && b->hasTagName(bodyTag)) ? static_cast<HTMLBodyElement*>(b) : 0;
if (!bodyElement)
return String();
return bodyElement->link();
}
void HTMLDocument::setLinkColor(const String& value)
{
HTMLElement* b = body();
HTMLBodyElement* bodyElement = (b && b->hasTagName(bodyTag)) ? static_cast<HTMLBodyElement*>(b) : 0;
if (bodyElement) {
// This check is a bit silly, but some benchmarks like to set the
// document's link colors over and over to the same value and we
// don't want to incur a style update each time.
if (bodyElement->link() != value)
bodyElement->setLink(value);
}
}
String HTMLDocument::vlinkColor()
{
HTMLElement* b = body();
HTMLBodyElement* bodyElement = (b && b->hasTagName(bodyTag)) ? static_cast<HTMLBodyElement*>(b) : 0;
if (!bodyElement)
return String();
return bodyElement->vLink();
}
void HTMLDocument::setVlinkColor(const String& value)
{
HTMLElement* b = body();
HTMLBodyElement* bodyElement = (b && b->hasTagName(bodyTag)) ? static_cast<HTMLBodyElement*>(b) : 0;
if (bodyElement) {
// This check is a bit silly, but some benchmarks like to set the
// document's link colors over and over to the same value and we
// don't want to incur a style update each time.
if (bodyElement->vLink() != value)
bodyElement->setVLink(value);
}
}
void HTMLDocument::captureEvents()
{
}
void HTMLDocument::releaseEvents()
{
}
Tokenizer *HTMLDocument::createTokenizer()
{
bool reportErrors = false;
if (frame())
if (Page* page = frame()->page())
reportErrors = page->inspectorController()->windowVisible();
return new HTMLTokenizer(this, reportErrors);
}
// --------------------------------------------------------------------------
// not part of the DOM
// --------------------------------------------------------------------------
bool HTMLDocument::childAllowed(Node *newChild)
{
return newChild->hasTagName(htmlTag) || newChild->isCommentNode();
}
PassRefPtr<Element> HTMLDocument::createElement(const String &name, ExceptionCode& ec)
{
String lowerName(name.lower());
if (!isValidName(lowerName)) {
ec = INVALID_CHARACTER_ERR;
return 0;
}
return HTMLElementFactory::createHTMLElement(AtomicString(lowerName), this, 0, false);
}
static void addItemToMap(HTMLDocument::NameCountMap& map, const String& name)
{
if (name.length() == 0)
return;
HTMLDocument::NameCountMap::iterator it = map.find(name.impl());
if (it == map.end())
map.set(name.impl(), 1);
else
++(it->second);
}
static void removeItemFromMap(HTMLDocument::NameCountMap& map, const String& name)
{
if (name.length() == 0)
return;
HTMLDocument::NameCountMap::iterator it = map.find(name.impl());
if (it == map.end())
return;
int oldVal = it->second;
ASSERT(oldVal != 0);
int newVal = oldVal - 1;
if (newVal == 0)
map.remove(it);
else
it->second = newVal;
}
void HTMLDocument::addNamedItem(const String& name)
{
addItemToMap(namedItemCounts, name);
}
void HTMLDocument::removeNamedItem(const String &name)
{
removeItemFromMap(namedItemCounts, name);
}
bool HTMLDocument::hasNamedItem(const String& name)
{
return namedItemCounts.get(name.impl()) != 0;
}
void HTMLDocument::addDocExtraNamedItem(const String& name)
{
addItemToMap(docExtraNamedItemCounts, name);
}
void HTMLDocument::removeDocExtraNamedItem(const String& name)
{
removeItemFromMap(docExtraNamedItemCounts, name);
}
bool HTMLDocument::hasDocExtraNamedItem(const String& name)
{
return docExtraNamedItemCounts.get(name.impl()) != 0;
}
const int PARSEMODE_HAVE_DOCTYPE = (1<<0);
const int PARSEMODE_HAVE_PUBLIC_ID = (1<<1);
const int PARSEMODE_HAVE_SYSTEM_ID = (1<<2);
const int PARSEMODE_HAVE_INTERNAL = (1<<3);
static int parseDocTypePart(const String& buffer, int index)
{
while (true) {
UChar ch = buffer[index];
if (ch == ' ' || ch == '\t' || ch == '\n' || ch == '\r')
++index;
else if (ch == '-') {
int tmpIndex=index;
if (buffer[index+1] == '-' &&
((tmpIndex=buffer.find("--", index+2)) != -1))
index = tmpIndex+2;
else
return index;
}
else
return index;
}
}
static bool containsString(const char* str, const String& buffer, int offset)
{
String startString(str);
if (offset + startString.length() > buffer.length())
return false;
String bufferString = buffer.substring(offset, startString.length()).lower();
String lowerStart = startString.lower();
return bufferString.startsWith(lowerStart);
}
static bool parseDocTypeDeclaration(const String& buffer,
int* resultFlags,
String& name,
String& publicID,
String& systemID)
{
bool haveDocType = false;
*resultFlags = 0;
// Skip through any comments and processing instructions.
int index = 0;
do {
index = buffer.find('<', index);
if (index == -1) break;
UChar nextChar = buffer[index+1];
if (nextChar == '!') {
if (containsString("doctype", buffer, index+2)) {
haveDocType = true;
index += 9; // Skip "<!DOCTYPE"
break;
}
index = parseDocTypePart(buffer,index);
index = buffer.find('>', index);
}
else if (nextChar == '?')
index = buffer.find('>', index);
else
break;
} while (index != -1);
if (!haveDocType)
return true;
*resultFlags |= PARSEMODE_HAVE_DOCTYPE;
index = parseDocTypePart(buffer, index);
if (!containsString("html", buffer, index))
return false;
name = buffer.substring(index, 4);
index = parseDocTypePart(buffer, index+4);
bool hasPublic = containsString("public", buffer, index);
if (hasPublic) {
index = parseDocTypePart(buffer, index+6);
// We've read <!DOCTYPE HTML PUBLIC (not case sensitive).
// Now we find the beginning and end of the public identifers
// and system identifiers (assuming they're even present).
UChar theChar = buffer[index];
if (theChar != '\"' && theChar != '\'')
return false;
// |start| is the first character (after the quote) and |end|
// is the final quote, so there are |end|-|start| characters.
int publicIDStart = index+1;
int publicIDEnd = buffer.find(theChar, publicIDStart);
if (publicIDEnd == -1)
return false;
index = parseDocTypePart(buffer, publicIDEnd+1);
UChar next = buffer[index];
if (next == '>') {
// Public identifier present, but no system identifier.
// Do nothing. Note that this is the most common
// case.
}
else if (next == '\"' || next == '\'') {
// We have a system identifier.
*resultFlags |= PARSEMODE_HAVE_SYSTEM_ID;
int systemIDStart = index+1;
int systemIDEnd = buffer.find(next, systemIDStart);
if (systemIDEnd == -1)
return false;
systemID = buffer.substring(systemIDStart, systemIDEnd - systemIDStart);
}
else if (next == '[') {
// We found an internal subset.
*resultFlags |= PARSEMODE_HAVE_INTERNAL;
}
else
return false; // Something's wrong.
// We need to trim whitespace off the public identifier.
publicID = buffer.substring(publicIDStart, publicIDEnd - publicIDStart);
publicID = publicID.stripWhiteSpace();
*resultFlags |= PARSEMODE_HAVE_PUBLIC_ID;
} else {
if (containsString("system", buffer, index)) {
// Doctype has a system ID but no public ID
*resultFlags |= PARSEMODE_HAVE_SYSTEM_ID;
index = parseDocTypePart(buffer, index+6);
UChar next = buffer[index];
if (next != '\"' && next != '\'')
return false;
int systemIDStart = index+1;
int systemIDEnd = buffer.find(next, systemIDStart);
if (systemIDEnd == -1)
return false;
systemID = buffer.substring(systemIDStart, systemIDEnd - systemIDStart);
index = parseDocTypePart(buffer, systemIDEnd+1);
}
UChar nextChar = buffer[index];
if (nextChar == '[')
*resultFlags |= PARSEMODE_HAVE_INTERNAL;
else if (nextChar != '>')
return false;
}
return true;
}
void HTMLDocument::determineParseMode(const String& str)
{
// This code more or less mimics Mozilla's implementation (specifically the
// doctype parsing implemented by David Baron in Mozilla's nsParser.cpp).
//
// There are three possible parse modes:
// COMPAT - quirks mode emulates WinIE
// and NS4. CSS parsing is also relaxed in this mode, e.g., unit types can
// be omitted from numbers.
// ALMOST STRICT - This mode is identical to strict mode
// except for its treatment of line-height in the inline box model. For
// now (until the inline box model is re-written), this mode is identical
// to STANDARDS mode.
// STRICT - no quirks apply. Web pages will obey the specifications to
// the letter.
String name, systemID, publicID;
int resultFlags = 0;
if (parseDocTypeDeclaration(str, &resultFlags, name, publicID, systemID)) {
if (resultFlags & PARSEMODE_HAVE_DOCTYPE)
setDocType(new DocumentType(this, name, publicID, systemID));
if (!(resultFlags & PARSEMODE_HAVE_DOCTYPE)) {
// No doctype found at all. Default to quirks mode and Html4.
setParseMode(Compat);
setHTMLMode(Html4);
}
else if ((resultFlags & PARSEMODE_HAVE_INTERNAL) ||
!(resultFlags & PARSEMODE_HAVE_PUBLIC_ID)) {
// Internal subsets always denote full standards, as does
// a doctype without a public ID.
setParseMode(Strict);
setHTMLMode(Html4);
}
else {
// We have to check a list of public IDs to see what we
// should do.
String lowerPubID = publicID.lower();
CString pubIDStr = lowerPubID.latin1();
// Look up the entry in our gperf-generated table.
const PubIDInfo* doctypeEntry = findDoctypeEntry(pubIDStr.data(), pubIDStr.length());
if (!doctypeEntry) {
// The DOCTYPE is not in the list. Assume strict mode.
setParseMode(Strict);
setHTMLMode(Html4);
return;
}
switch ((resultFlags & PARSEMODE_HAVE_SYSTEM_ID) ?
doctypeEntry->mode_if_sysid :
doctypeEntry->mode_if_no_sysid)
{
case PubIDInfo::eQuirks3:
setParseMode(Compat);
setHTMLMode(Html3);
break;
case PubIDInfo::eQuirks:
setParseMode(Compat);
setHTMLMode(Html4);
break;
case PubIDInfo::eAlmostStandards:
setParseMode(AlmostStrict);
setHTMLMode(Html4);
break;
default:
ASSERT(false);
}
}
}
else {
// Malformed doctype implies quirks mode.
setParseMode(Compat);
setHTMLMode(Html3);
}
styleSelector()->strictParsing = !inCompatMode();
}
}