mozilla
/
pjs
зеркало из https://github.com/mozilla/pjs.git
1
0
Форкнуть 0
Код Задачи Пакеты Проекты Релизы Вики Активность

huge improvements to parser

This commit is contained in:
rickg 1998-04-30 05:56:39 +00:00
Родитель fee218f0dc
Коммит 211dce8234
34 изменённых файлов: 598 добавлений и 380 удалений
Показать статистику Скачать Patch файл Скачать Diff файл Показать все файлы Свернуть все файлы

106
htmlparser/src/CNavDTD.cpp
Просмотреть файл

@ -27,6 +27,7 @@
#include "nsHTMLTokens.h"
#include "nsCRT.h"
#include "nsParserTypes.h"
#include "nsIParser.h"
static NS_DEFINE_IID(kISupportsIID, NS_ISUPPORTS_IID);
static NS_DEFINE_IID(kIDTDIID, NS_IDTD_IID);
@ -98,6 +99,7 @@ NS_IMPL_RELEASE(CNavDTD)
* @return
*/
CNavDTD::CNavDTD() : nsIDTD() {
mParser=0;
}
/**
@ -110,6 +112,36 @@ CNavDTD::CNavDTD() : nsIDTD() {
CNavDTD::~CNavDTD(){
}
/**
*
*
* @update gess 3/25/98
* @param
* @return
*/
void CNavDTD::SetParser(nsIParser* aParser) {
mParser=aParser;
}
static char formElementTags[]= {
eHTMLTag_button, eHTMLTag_fieldset, eHTMLTag_input,
eHTMLTag_isindex, eHTMLTag_label, eHTMLTag_legend,
eHTMLTag_select, eHTMLTag_textarea,0};
/**
* This method is called to determine whether or not a tag
* of one type can contain a tag of another type.
*
* @update gess 4/8/98
* @param aParent -- tag enum of parent container
* @param aChild -- tag enum of child container
* @return PR_TRUE if parent can contain child
*/
PRBool CNavDTD::CanContainFormElement(PRInt32 aParent,PRInt32 aChild) const {
PRBool result=(mParser) ? mParser->HasOpenForm() : PR_FALSE;
return result;
}
/**
* This method is called to determine whether or not a tag
@ -121,6 +153,7 @@ CNavDTD::~CNavDTD(){
* @return PR_TRUE if parent can contain child
*/
PRBool CNavDTD::CanContain(PRInt32 aParent,PRInt32 aChild) const {
PRBool result=PR_FALSE;
//tagset1 has 64 members...
@ -194,6 +227,11 @@ PRBool CNavDTD::CanContain(PRInt32 aParent,PRInt32 aChild) const {
if(eHTMLTag_userdefined==aChild) // XXX Hack: For now...
result=PR_TRUE;
//handle form elements (this is very much a WIP!!!)
if(0!=strchr(formElementTags,aChild)){
return CanContainFormElement(aParent,aChild);
}
switch(aParent) {
case eHTMLTag_a:
result=PRBool(0!=strchr(gTagSet2,aChild)); break;
@ -302,7 +340,15 @@ PRBool CNavDTD::CanContain(PRInt32 aParent,PRInt32 aChild) const {
case eHTMLTag_h1: case eHTMLTag_h2:
case eHTMLTag_h3: case eHTMLTag_h4:
case eHTMLTag_h5: case eHTMLTag_h6:
result=PRBool(0!=strchr(gTagSet1,aChild)); break;
{
static char badTags[]={
eHTMLTag_h1, eHTMLTag_h2, eHTMLTag_h3,
eHTMLTag_h4, eHTMLTag_h5, eHTMLTag_h6, 0};
if(0!=strchr(badTags,aChild))
result=PR_FALSE;
else result=PRBool(0!=strchr(gTagSet1,aChild));
}
break;
case eHTMLTag_head:
{
@ -418,7 +464,7 @@ PRBool CNavDTD::CanContain(PRInt32 aParent,PRInt32 aChild) const {
break; //unadorned script text...
case eHTMLTag_select:
result=PRBool(eHTMLTag_option==aChild); break;
result=PR_TRUE; break; //for now, allow select to contain anything...
case eHTMLTag_small:
result=PRBool(0!=strchr(gTagSet2,aChild)); break;
@ -483,7 +529,7 @@ PRBool CNavDTD::CanContain(PRInt32 aParent,PRInt32 aChild) const {
case eHTMLTag_userdefined:
result=PR_TRUE; break; //XXX for now...
case eHTMLTag_xmp:
case eHTMLTag_xmp:
default:
break;
} //switch
@ -529,6 +575,8 @@ PRBool CNavDTD::CanContainIndirect(PRInt32 aParent,PRInt32 aChild) const {
}
break;
result=PR_TRUE; break;
default:
break;
}
@ -546,29 +594,43 @@ PRBool CNavDTD::CanContainIndirect(PRInt32 aParent,PRInt32 aChild) const {
PRBool CNavDTD::CanOmit(PRInt32 aParent,PRInt32 aChild) const {
PRBool result=PR_FALSE;
switch((eHTMLTags)aParent) {
case eHTMLTag_html:
case eHTMLTag_body:
case eHTMLTag_head:
case eHTMLTag_title:
case eHTMLTag_tr:
// case eHTMLTag_td:
case eHTMLTag_table:
case eHTMLTag_thead:
case eHTMLTag_tfoot:
case eHTMLTag_tbody:
case eHTMLTag_col:
case eHTMLTag_colgroup:
if((aChild==eHTMLTag_newline) ||
(aChild==eHTMLTag_whitespace))
result=PR_TRUE;
//begin with some simple (and obvious) cases...
switch((eHTMLTags)aChild) {
case eHTMLTag_userdefined:
case eHTMLTag_comment:
result=PR_TRUE;
break;
case eHTMLTag_button: case eHTMLTag_fieldset:
case eHTMLTag_input: case eHTMLTag_isindex:
case eHTMLTag_label: case eHTMLTag_legend:
case eHTMLTag_select: case eHTMLTag_textarea:
if(PR_FALSE==mParser->HasOpenForm())
result=PR_TRUE;
break;
case eHTMLTag_newline:
case eHTMLTag_whitespace:
switch((eHTMLTags)aParent) {
case eHTMLTag_html: case eHTMLTag_head:
case eHTMLTag_title: case eHTMLTag_map:
case eHTMLTag_tr: case eHTMLTag_table:
case eHTMLTag_thead: case eHTMLTag_tfoot:
case eHTMLTag_tbody: case eHTMLTag_col:
case eHTMLTag_colgroup: case eHTMLTag_unknown:
result=PR_TRUE;
default:
break;
} //switch
break;
default:
result=PR_FALSE;
if(eHTMLTag_unknown==aParent)
result=PRBool(eHTMLTag_html!=aChild);
break;
}
} //switch
return result;
}

15
htmlparser/src/CNavDTD.h
Просмотреть файл

@ -36,6 +36,7 @@
{0xaa, 0xda, 0x00, 0x80, 0x5f, 0x8a, 0x3e, 0x14}}
class nsIParser;
class CNavDTD : public nsIDTD {
@ -73,6 +74,15 @@ class CNavDTD : public nsIDTD {
*/
virtual PRBool CanContain(PRInt32 aParent,PRInt32 aChild) const;
/**
*
*
* @update gess 3/25/98
* @param
* @return
*/
virtual void SetParser(nsIParser* aParser);
/**
* This method is called to determine whether or not a tag
* of one type can contain a tag of another type.
@ -150,6 +160,11 @@ class CNavDTD : public nsIDTD {
*/
virtual PRBool BackwardPropagate(nsString& aVector,PRInt32 aParentTag,PRInt32 aChildTag) const;
protected:
PRBool CanContainFormElement(PRInt32 aParent,PRInt32 aChild) const;
nsIParser* mParser;
};

4
htmlparser/src/CNavDelegate.cpp
Просмотреть файл

@ -61,7 +61,7 @@ CNavDelegate::CNavDelegate(CNavDelegate& aDelegate) :
* @param
* @return
*/
eParseMode CNavDelegate::GetParseMode() const {
eParseMode CNavDelegate::GetParseMode(void) const {
return eParseMode_unknown;
}
@ -104,7 +104,7 @@ CToken* CNavDelegate::ConsumeTag(PRUnichar aChar,CScanner& aScanner,PRInt32& anE
default:
if(nsString::IsAlpha(aChar))
return ConsumeStartTag(aChar,aScanner,anErrorCode);
else if(kNotFound!=aChar) {
else if(kEOF!=aChar) {
nsAutoString temp("<");
return ConsumeText(temp,aScanner,anErrorCode);
}

11
htmlparser/src/CNavDelegate.h
Просмотреть файл

@ -49,15 +49,15 @@ class CNavDelegate : public ITokenizerDelegate {
CNavDelegate();
CNavDelegate(CNavDelegate& aDelegate);
virtual CToken* GetToken(CScanner& aScanner,PRInt32& anErrorCode);
virtual CToken* GetToken(CScanner& aScanner,PRInt32& anErrorCode);
virtual PRBool WillAddToken(CToken& aToken);
virtual PRBool WillTokenize();
virtual PRBool DidTokenize();
virtual PRBool WillTokenize();
virtual PRBool DidTokenize();
virtual eParseMode GetParseMode() const;
virtual eParseMode GetParseMode(void) const;
virtual nsIDTD* GetDTD(void) const;
static void SelfTest();
static void SelfTest();
protected:
@ -82,4 +82,3 @@ class CNavDelegate : public ITokenizerDelegate {
#endif

16
htmlparser/src/COtherDTD.cpp
Просмотреть файл

@ -27,6 +27,7 @@
#include "nsHTMLTokens.h"
#include "nsCRT.h"
#include "nsParserTypes.h"
#include "nsIParser.h"
static NS_DEFINE_IID(kISupportsIID, NS_ISUPPORTS_IID);
static NS_DEFINE_IID(kIDTDIID, NS_IDTD_IID);
@ -98,6 +99,7 @@ NS_IMPL_RELEASE(COtherDTD)
* @return
*/
COtherDTD::COtherDTD() : nsIDTD() {
mParser=0;
}
/**
@ -110,6 +112,18 @@ COtherDTD::COtherDTD() : nsIDTD() {
COtherDTD::~COtherDTD(){
}
/**
*
*
* @update gess 3/25/98
* @param
* @return
*/
void COtherDTD::SetParser(nsIParser* aParser) {
mParser=aParser;
}
/**
* This method is called to determine whether or not a tag
* of one type can contain a tag of another type.
@ -503,6 +517,8 @@ PRBool COtherDTD::CanContainIndirect(PRInt32 aParent,PRInt32 aChild) const {
}
break;
result=PR_TRUE; break;
default:
break;
}

15
htmlparser/src/COtherDTD.h
Просмотреть файл

@ -37,6 +37,7 @@
class nsIParser;
class COtherDTD : public nsIDTD {
@ -52,7 +53,7 @@ class COtherDTD : public nsIDTD {
* @param
* @return
*/
COtherDTD();
COtherDTD();
/**
*
@ -63,6 +64,16 @@ class COtherDTD : public nsIDTD {
*/
virtual ~COtherDTD();
/**
*
*
* @update gess 3/25/98
* @param
* @return
*/
virtual void SetParser(nsIParser* aParser);
/**
* This method is called to determine whether or not a tag
* of one type can contain a tag of another type.
@ -153,6 +164,8 @@ class COtherDTD : public nsIDTD {
*/
virtual PRBool BackwardPropagate(nsString& aVector,PRInt32 aParentTag,PRInt32 aChildTag) const;
protected:
nsIParser* mParser;
};

4
htmlparser/src/COtherDelegate.cpp
Просмотреть файл

@ -62,7 +62,7 @@ COtherDelegate::COtherDelegate(COtherDelegate& aDelegate) :
* @param
* @return
*/
eParseMode COtherDelegate::GetParseMode() const {
eParseMode COtherDelegate::GetParseMode(void) const {
return eParseMode_unknown;
}
@ -104,7 +104,7 @@ CToken* COtherDelegate::ConsumeTag(PRUnichar aChar,CScanner& aScanner,PRInt32& a
default:
if(nsString::IsAlpha(aChar))
return ConsumeStartTag(aChar,aScanner,anErrorCode);
else if(kNotFound!=aChar) {
else if(kEOF!=aChar) {
nsAutoString temp("<");
return ConsumeText(temp,aScanner,anErrorCode);
}

10
htmlparser/src/COtherDelegate.h
Просмотреть файл

@ -49,15 +49,15 @@ class COtherDelegate : public ITokenizerDelegate {
COtherDelegate();
COtherDelegate(COtherDelegate& aDelegate);
virtual CToken* GetToken(CScanner& aScanner,PRInt32& anErrorCode);
virtual CToken* GetToken(CScanner& aScanner,PRInt32& anErrorCode);
virtual PRBool WillAddToken(CToken& aToken);
virtual PRBool WillTokenize();
virtual PRBool DidTokenize();
virtual PRBool WillTokenize();
virtual PRBool DidTokenize();
virtual eParseMode GetParseMode() const;
virtual eParseMode GetParseMode(void) const;
virtual nsIDTD* GetDTD(void) const;
static void SelfTest();
static void SelfTest();
protected:

202
htmlparser/src/nsHTMLParser.cpp
Просмотреть файл

@ -100,6 +100,7 @@ nsHTMLParser::nsHTMLParser() {
nsCRT::zero(mContextStack,sizeof(mContextStack));
nsCRT::zero(mTokenHandlers,sizeof(mTokenHandlers));
mDTD=0;
mHasOpenForm=PR_FALSE;
mTokenHandlerCount=0;
InitializeDefaultTokenHandlers();
gVerificationOutputDir = PR_GetEnv("VERIFY_PARSER");
@ -176,8 +177,20 @@ nsresult nsHTMLParser::QueryInterface(const nsIID& aIID, void** aInstancePtr)
eHTMLTags nsHTMLParser::NodeAt(PRInt32 aPos) const {
NS_PRECONDITION(0 <= aPos, "bad nodeAt");
if((aPos>-1) && (aPos<mContextStackPos))
return mContextStack[aPos];
return (eHTMLTags)kNotFound;
return (eHTMLTags)mContextStack[aPos];
return eHTMLTag_unknown;
}
/**
* This method allows the caller to determine if a form
* element is currently open.
*
* @update gess 4/2/98
* @param
* @return
*/
PRBool nsHTMLParser::HasOpenForm() const {
return mHasOpenForm;
}
/**
@ -190,10 +203,11 @@ eHTMLTags nsHTMLParser::NodeAt(PRInt32 aPos) const {
*/
eHTMLTags nsHTMLParser::GetTopNode() const {
if(mContextStackPos)
return mContextStack[mContextStackPos-1];
return (eHTMLTags)kNotFound;
return (eHTMLTags)mContextStack[mContextStackPos-1];
return eHTMLTag_unknown;
}
/**
* Determine whether the given tag is open anywhere
* in our context stack.
@ -326,14 +340,13 @@ nsIContentSink* nsHTMLParser::SetContentSink(nsIContentSink* aSink) {
* @param aDTD is the DTD we plan to ask for verification
* @return TRUE if we know how to handle it, else false
*/
PRBool VerifyContextVector(eHTMLTags tags[],PRInt32 count,nsIDTD* aDTD) {
PRBool VerifyContextVector(PRInt32 aTags[],PRInt32 count,nsIDTD* aDTD) {
PRBool result=PR_TRUE;
if(0!=gVerificationOutputDir) {
if(aDTD){
PRInt32 theArray[50];
#ifdef XP_PC
char path[_MAX_PATH+1];
@ -341,17 +354,16 @@ PRBool VerifyContextVector(eHTMLTags tags[],PRInt32 count,nsIDTD* aDTD) {
#endif
for(int i=0;i<count;i++){
theArray[i]=tags[i];
#ifdef NS_WIN32
strcat(path,"/");
const char* name=GetTagName(tags[i]);
const char* name=GetTagName(aTags[i]);
strcat(path,name);
mkdir(path);
#endif
}
//ok, now see if we understand this vector
result=aDTD->VerifyContextVector(theArray,count);
result=aDTD->VerifyContextVector(aTags,count);
}
if(PR_FALSE==result){
//add debugging code here to record the fact that we just encountered
@ -389,9 +401,12 @@ PRBool nsHTMLParser::IterateTokens() {
eHTMLTokenTypes type=eHTMLTokenTypes(theToken->GetTokenType());
iteration++; //debug purposes...
switch(eHTMLTokenTypes(type)){
switch(type){
case eToken_start:
case eToken_text:
case eToken_newline:
case eToken_whitespace:
result=HandleStartToken(theToken); break;
case eToken_end:
@ -399,11 +414,6 @@ PRBool nsHTMLParser::IterateTokens() {
case eToken_entity:
result=HandleEntityToken(theToken); break;
case eToken_text:
case eToken_newline:
case eToken_whitespace:
result=HandleSimpleContentToken(theToken); break;
case eToken_skippedcontent:
//used in cases like <SCRIPT> where we skip over script content.
@ -412,9 +422,6 @@ PRBool nsHTMLParser::IterateTokens() {
case eToken_attribute:
result=HandleAttributeToken(theToken); break;
case eToken_script:
result=HandleScriptToken(theToken); break;
case eToken_style:
result=HandleStyleToken(theToken); break;
@ -428,7 +435,7 @@ PRBool nsHTMLParser::IterateTokens() {
}
VerifyContextVector(mContextStack,mContextStackPos,mDTD);
++(*mCurrentPos);
done=PRBool(e==*mCurrentPos);
done=PRBool(!(*mCurrentPos<e));
}
//One last thing...close any open containers.
@ -438,6 +445,7 @@ PRBool nsHTMLParser::IterateTokens() {
return result;
}
/**
* This is the main controlling routine in the parsing process.
* Note that it may get called multiple times for the same scanner,
@ -454,7 +462,7 @@ PRBool nsHTMLParser::Parse(nsIURL* aURL){
char* theModeStr= PR_GetEnv("PARSE_MODE");
if(theModeStr)
if(0!=PL_strncasecmp("nav",theModeStr,3))
if(0==strnicmp("other",theModeStr,5))
theMode=eParseMode_other;
return Parse(aURL,theMode);
@ -501,6 +509,8 @@ PRBool nsHTMLParser::Parse(nsIURL* aURL,eParseMode aMode){
return PR_FALSE;
}
if(mDTD)
mDTD->SetParser(this);
mTokenizer=new CTokenizer(aURL, theDelegate, mParseMode);
mTokenizer->Tokenize();
result=IterateTokens();
@ -525,6 +535,17 @@ PRBool nsHTMLParser::ResumeParse() {
}
/**
*
* @update gess4/22/98
* @param
* @return
*/
PRInt32 nsHTMLParser::GetStack(PRInt32* aStackPtr) {
aStackPtr=&mContextStack[0];
return mContextStackPos;
}
/**
*
@ -537,15 +558,18 @@ PRInt32 nsHTMLParser::CollectAttributes(nsCParserNode& aNode){
nsDeque& deque=mTokenizer->GetDeque();
nsDequeIterator end=deque.End();
while((*mCurrentPos!=end) && (eToken_attribute==subtype)) {
while((*mCurrentPos<end) && (eToken_attribute==subtype)) {
CHTMLToken* tkn=(CHTMLToken*)(++(*mCurrentPos));
subtype=eHTMLTokenTypes(tkn->GetTokenType());
if(eToken_attribute==subtype) {
aNode.AddAttribute(tkn);
}
else (*mCurrentPos)--;
if(tkn){
subtype=eHTMLTokenTypes(tkn->GetTokenType());
if(eToken_attribute==subtype) {
aNode.AddAttribute(tkn);
}
else (*mCurrentPos)--;
}
}
return aNode.GetAttributeCount();
}
@ -601,7 +625,7 @@ PRBool nsHTMLParser::HandleDefaultStartToken(CToken* aToken,eHTMLTags aChildTag,
//You must determine what container hierarchy you need to hold aToken,
//and create that on the parsestack.
result=ReduceContextStackFor(aChildTag);
if(PR_FALSE==mDTD->CanContain(parentTag,aChildTag)) {
if(PR_FALSE==mDTD->CanContain(GetTopNode(),aChildTag)) {
//we unwound too far; now we have to recreate a valid context stack.
result=CreateContextStackFor(aChildTag);
}
@ -615,35 +639,6 @@ PRBool nsHTMLParser::HandleDefaultStartToken(CToken* aToken,eHTMLTags aChildTag,
result=AddLeaf(aNode);
}
return result;
/*
if(kNotFound!=parentTag) {
if(PR_FALSE==mDTD->CanContain(parentTag,aTag)){
//if you're here, then the new topmost container can't contain aToken.
//You must determine what container hierarchy you need to hold aToken,
//and create that on the parsestack.
ReduceContextStackFor(aTag);
if(PR_FALSE==mDTD->CanContain(parentTag,aTag)) {
//we unwound too far; now we have to recreate a valid context stack.
result=CreateContextStackFor(aTag);
}
}
}
else {
CreateContextStackFor(aTag);
}
//now we think the stack is correct, so fall through
//and push our newest token onto the stack...
if(mDTD->IsContainer(aTag)){
result=OpenContainer(aNode);
}
else {
result=AddLeaf(aNode);
}
return result;
*/
}
/**
@ -672,6 +667,11 @@ PRBool nsHTMLParser::HandleStartToken(CToken* aToken) {
nsCParserNode attrNode((CHTMLToken*)aToken);
PRInt32 attrCount=CollectAttributes(attrNode);
//now check to see if this token should be omitted...
if(PR_TRUE==mDTD->CanOmit(GetTopNode(),tokenTagType)) {
return PR_TRUE;
}
switch(tokenTagType) {
case eHTMLTag_html:
@ -695,7 +695,7 @@ PRBool nsHTMLParser::HandleStartToken(CToken* aToken) {
break;
case eHTMLTag_form:
result = mSink->OpenForm(attrNode);
result = OpenForm(attrNode);
break;
case eHTMLTag_meta:
@ -718,6 +718,10 @@ PRBool nsHTMLParser::HandleStartToken(CToken* aToken) {
}
break;
case eHTMLTag_script:
result=HandleScriptToken(st); break;
case eHTMLTag_map:
// Put map into the head section
result=OpenHead(attrNode);
@ -727,6 +731,9 @@ PRBool nsHTMLParser::HandleStartToken(CToken* aToken) {
case eHTMLTag_head:
result=PR_TRUE; break; //ignore head tags...
case eHTMLTag_select:
result=PR_FALSE;
default:
result=HandleDefaultStartToken(aToken,tokenTagType,attrNode);
break;
@ -755,6 +762,11 @@ PRBool nsHTMLParser::HandleEndToken(CToken* aToken) {
CEndToken* st = (CEndToken*)(aToken);
eHTMLTags tokenTagType=st->GetHTMLTag();
//now check to see if this token should be omitted...
if(PR_TRUE==mDTD->CanOmit(GetTopNode(),tokenTagType)) {
return PR_TRUE;
}
nsCParserNode theNode((CHTMLToken*)aToken);
switch(tokenTagType) {
case eHTMLTag_html:
@ -780,15 +792,19 @@ PRBool nsHTMLParser::HandleEndToken(CToken* aToken) {
case eHTMLTag_form:
{
nsCParserNode aNode((CHTMLToken*)aToken);
result=mSink->CloseForm(aNode);
result=CloseForm(aNode);
}
break;
case eHTMLTag_script:
result=PR_TRUE; break;
default:
if(mDTD->IsContainer(tokenTagType)){
result=CloseContainer(theNode);
}
result=PR_TRUE;
//actually, you want to push the start token and ALL it's attributes...
//
break;
}
return result;
@ -826,43 +842,6 @@ PRBool nsHTMLParser::HandleCommentToken(CToken* aToken) {
return PR_TRUE;
}
/**
* This method gets called when a text, whitespace and newline
* tokens have been encountered in the parse process. After
* verifying that the topmost container can contain this data,
* we call AddLeaf to store this token in the top container.
*
* @update gess 3/25/98
* @param aToken -- next (text) token to be handled
* @return PR_TRUE if all went well; PR_FALSE if error occured
*/
PRBool nsHTMLParser::HandleSimpleContentToken(CToken* aToken) {
NS_PRECONDITION(0!=aToken,kNullToken);
CHTMLToken* theToken=(CHTMLToken*)aToken;
eHTMLTags type=theToken->GetHTMLTag();
PRBool result=PR_TRUE;
nsCParserNode aNode(theToken);
switch(type) {
case eHTMLTag_text:
result=HandleDefaultStartToken(theToken,type,aNode);
break;
case eHTMLTag_newline:
case eHTMLTag_whitespace:
if((mContextStackPos>0) &&
(PR_FALSE==mDTD->CanOmit(mContextStack[mContextStackPos-1],type))) {
result=HandleDefaultStartToken(theToken,type,aNode);
}
break;
default:
break;
} //switch
return result;
}
/**
* This method gets called when a skippedcontent token has
* been encountered in the parse process. After verifying
@ -996,6 +975,7 @@ PRBool nsHTMLParser::CloseHTML(const nsIParserNode& aNode){
* @return TRUE if ok, FALSE if error
*/
PRBool nsHTMLParser::OpenHead(const nsIParserNode& aNode){
mContextStack[mContextStackPos++]=eHTMLTag_head;
PRBool result=mSink->OpenHead(aNode);
return result;
}
@ -1010,6 +990,7 @@ PRBool nsHTMLParser::OpenHead(const nsIParserNode& aNode){
*/
PRBool nsHTMLParser::CloseHead(const nsIParserNode& aNode){
PRBool result=mSink->CloseHead(aNode);
mContextStack[--mContextStackPos]=eHTMLTag_unknown;
return result;
}
@ -1084,10 +1065,10 @@ PRBool nsHTMLParser::CloseBody(const nsIParserNode& aNode){
* @return TRUE if ok, FALSE if error
*/
PRBool nsHTMLParser::OpenForm(const nsIParserNode& aNode){
NS_PRECONDITION(mContextStackPos >= 0, kInvalidTagStackPos);
PRBool result=mSink->OpenForm(aNode);
mContextStack[mContextStackPos++]=(eHTMLTags)aNode.GetNodeType();
return result;
if(mHasOpenForm)
CloseForm(aNode);
mHasOpenForm=PR_TRUE;
return mSink->OpenForm(aNode);
}
/**
@ -1099,9 +1080,11 @@ PRBool nsHTMLParser::OpenForm(const nsIParserNode& aNode){
* @return TRUE if ok, FALSE if error
*/
PRBool nsHTMLParser::CloseForm(const nsIParserNode& aNode){
NS_PRECONDITION(mContextStackPos > 0, kInvalidTagStackPos);
PRBool result=mSink->CloseContainer(aNode);
mContextStack[--mContextStackPos]=eHTMLTag_unknown;
PRBool result=PR_TRUE;
if(mHasOpenForm) {
mHasOpenForm=PR_FALSE;
result=mSink->CloseForm(aNode);
}
return result;
}
@ -1230,12 +1213,12 @@ PRBool nsHTMLParser::CloseContainersTo(PRInt32 anIndex){
PRBool result=PR_TRUE;
nsAutoString empty;
CHTMLToken aToken(empty);
CEndToken aToken(empty);
nsCParserNode theNode(&aToken);
if((anIndex<mContextStackPos) && (anIndex>=0)) {
while(mContextStackPos>anIndex) {
aToken.SetHTMLTag(mContextStack[mContextStackPos-1]);
aToken.SetHTMLTag((eHTMLTags)mContextStack[mContextStackPos-1]);
result=CloseContainer(theNode);
}
}
@ -1287,7 +1270,7 @@ PRBool nsHTMLParser::CloseTopmostContainer(){
nsAutoString empty;
CEndToken aToken(empty);
aToken.SetHTMLTag(mContextStack[mContextStackPos-1]);
aToken.SetHTMLTag((eHTMLTags)mContextStack[mContextStackPos-1]);
nsCParserNode theNode(&aToken);
return CloseContainer(theNode);
@ -1323,14 +1306,17 @@ PRBool nsHTMLParser::CreateContextStackFor(PRInt32 aChildTag){
PRBool result=PR_FALSE;
PRInt32 pos=0;
PRInt32 cnt=0;
PRInt32 theTop=GetTopNode();
if(PR_TRUE==mDTD->ForwardPropagate(theVector,GetTopNode(),aChildTag)){
if(PR_TRUE==mDTD->ForwardPropagate(theVector,theTop,aChildTag)){
//add code here to build up context stack based on forward propagated context vector...
pos=0;
cnt=theVector.Length()-1;
result=PRBool(mContextStack[mContextStackPos-1]==theVector[cnt]);
}
else if(PR_TRUE==mDTD->BackwardPropagate(theVector,eHTMLTag_html,aChildTag)) {
else if(PR_TRUE==mDTD->BackwardPropagate(theVector,theTop,aChildTag)) {
if(theTop!=eHTMLTag_html)
mDTD->BackwardPropagate(theVector,eHTMLTag_html,theTop);
//propagation worked, so pop unwanted containers, push new ones, then exit...
pos=0;

9
htmlparser/src/nsHTMLParser.h
Просмотреть файл

@ -87,13 +87,15 @@ friend class CTokenHandler;
virtual PRBool Parse(nsIURL* aURL);
virtual PRBool Parse(nsIURL* aURL,eParseMode aMode);
virtual PRBool ResumeParse();
virtual PRInt32 GetStack(PRInt32* aStackPtr);
virtual PRBool HasOpenForm() const;
PRBool HandleStartToken(CToken* aToken);
PRBool HandleDefaultStartToken(CToken* aToken,eHTMLTags aChildTag,nsCParserNode& aNode);
PRBool HandleEndToken(CToken* aToken);
PRBool HandleEntityToken(CToken* aToken);
PRBool HandleCommentToken(CToken* aToken);
PRBool HandleSimpleContentToken(CToken* aToken);
PRBool HandleSkippedContentToken(CToken* aToken);
PRBool HandleAttributeToken(CToken* aToken);
PRBool HandleScriptToken(CToken* aToken);
@ -140,12 +142,12 @@ friend class CTokenHandler;
PRInt32 GetTopmostIndex(eHTMLTags aTag) const;
PRBool ReduceContextStackFor(PRInt32 aChildTag);
PRBool CreateContextStackFor(PRInt32 aChildTag);
PRBool HandleDefaultStartToken(CToken* aToken,eHTMLTags aTag,nsCParserNode& aNode);
nsIHTMLContentSink* mSink;
CTokenizer* mTokenizer;
eHTMLTags mContextStack[50];
// eHTMLTags mContextStack[50];
PRInt32 mContextStack[50];
PRInt32 mContextStackPos;
CTokenHandler* mTokenHandlers[100];
@ -154,6 +156,7 @@ friend class CTokenHandler;
nsIDTD* mDTD;
eParseMode mParseMode;
PRBool mHasOpenForm;
};

5
htmlparser/src/nsHTMLTokens.cpp
Просмотреть файл

@ -547,6 +547,7 @@ PRInt32 CTextToken::Consume(PRUnichar aChar, CScanner& aScanner) {
*------------------------------------------------------*/
CCommentToken::CCommentToken(const nsString& aName) : CHTMLToken(aName) {
mOrdinalValue=eToken_comment;
mTagType=eHTMLTag_comment;
}
/**-------------------------------------------------------
@ -561,11 +562,11 @@ CCommentToken::CCommentToken(const nsString& aName) : CHTMLToken(aName) {
PRInt32 CCommentToken::Consume(PRUnichar aChar, CScanner& aScanner) {
PRUnichar ch,ch2;
PRInt32 result;
static nsAutoString terminals(">");
aScanner.GetChar(ch);
PRInt32 result=aScanner.GetChar(ch);
mTextValue="<!";
if(kMinus==ch) {
aScanner.GetChar(ch2);
@ -1285,7 +1286,7 @@ eHTMLTags DetermineHTMLTagType(const nsString& aString)
* @param
* @return
*/
const char* GetTagName(eHTMLTags aTag) {
const char* GetTagName(PRInt32 aTag) {
const char* result=0;
PRInt32 cnt=sizeof(gHTMLTagTable)/sizeof(HTMLTagEntry);

11
htmlparser/src/nsHTMLTokens.h
Просмотреть файл

@ -58,9 +58,10 @@ enum eHTMLTags
eHTMLTag_blink, eHTMLTag_blockquote, eHTMLTag_body, eHTMLTag_br,
eHTMLTag_button, eHTMLTag_caption, eHTMLTag_center,
eHTMLTag_certificate, eHTMLTag_cite,
eHTMLTag_code, eHTMLTag_col, eHTMLTag_colgroup, eHTMLTag_dd,
eHTMLTag_del, eHTMLTag_dfn, eHTMLTag_div, eHTMLTag_dir,
eHTMLTag_dl, eHTMLTag_dt, eHTMLTag_em, eHTMLTag_embed,
eHTMLTag_code, eHTMLTag_col, eHTMLTag_colgroup, eHTMLTag_comment,
eHTMLTag_dd, eHTMLTag_del, eHTMLTag_dfn, eHTMLTag_div,
eHTMLTag_dir, eHTMLTag_dl, eHTMLTag_dt,
eHTMLTag_em, eHTMLTag_embed,
eHTMLTag_fieldset, eHTMLTag_font, eHTMLTag_footer,
eHTMLTag_form, eHTMLTag_frame, eHTMLTag_frameset,
eHTMLTag_h1, eHTMLTag_h2, eHTMLTag_h3, eHTMLTag_h4,
@ -77,7 +78,7 @@ enum eHTMLTags
eHTMLTag_option, eHTMLTag_paragraph, eHTMLTag_param, eHTMLTag_plaintext,
eHTMLTag_pre, eHTMLTag_quotation, eHTMLTag_strike, eHTMLTag_samp,
eHTMLTag_script, eHTMLTag_select,
eHTMLTag_server, eHTMLTag_small,
eHTMLTag_server, eHTMLTag_small,
eHTMLTag_spacer, eHTMLTag_span,
eHTMLTag_strong, eHTMLTag_style, eHTMLTag_sub, eHTMLTag_sup,
eHTMLTag_table, eHTMLTag_tbody, eHTMLTag_td,
@ -113,7 +114,7 @@ PRInt32 ConsumeAttributeText(PRUnichar aChar,nsString& aString,CScanner&
PRInt32 FindEntityIndex(const char* aBuffer,PRInt32 aBufLen=-1);
eHTMLTags DetermineHTMLTagType(const nsString& aString);
eHTMLTokenTypes DetermineTokenType(const nsString& aString);
const char* GetTagName(eHTMLTags aTag);
const char* GetTagName(PRInt32 aTag);
/** -----------------------------------------------------

14
htmlparser/src/nsIDTD.h
Просмотреть файл

@ -35,6 +35,7 @@
{0xaa, 0xda, 0x00, 0x80, 0x5f, 0x8a, 0x3e, 0x14}}
class nsIParser;
class nsIDTD : public nsISupports {
@ -49,7 +50,18 @@ class nsIDTD : public nsISupports {
* @param aChild -- tag enum of child container
* @return PR_TRUE if parent can contain child
*/
virtual PRBool CanContain(PRInt32 aParent,PRInt32 aChild) const =0;
virtual PRBool CanContain(PRInt32 aParent,PRInt32 aChild) const=0;
/**
* This method is called to determine whether or not a tag
* of one type can contain a tag of another type.
*
* @update gess 3/25/98
* @param aParent -- tag enum of parent container
* @param aChild -- tag enum of child container
* @return PR_TRUE if parent can contain child
*/
virtual void SetParser(nsIParser* aParser)=0;
/**
* This method is called to determine whether or not a tag

2
htmlparser/src/nsIParser.h
Просмотреть файл

@ -54,6 +54,8 @@ class nsIParser : public nsISupports {
virtual nsIContentSink* SetContentSink(nsIContentSink* aContentSink)=0;
virtual PRBool Parse(nsIURL* aURL)=0;
virtual PRBool ResumeParse()=0;
virtual PRInt32 GetStack(PRInt32* aStackPtr)=0;
virtual PRBool HasOpenForm() const=0;
};
extern NS_HTMLPARS nsresult NS_NewHTMLParser(nsIParser** aInstancePtrResult);

8
htmlparser/src/nsITokenizerDelegate.h
Просмотреть файл

@ -48,13 +48,13 @@ class CToken;
class ITokenizerDelegate {
public:
virtual PRBool WillTokenize()=0;
virtual PRBool DidTokenize()=0;
virtual PRBool WillTokenize()=0;
virtual PRBool DidTokenize()=0;
virtual CToken* GetToken(CScanner& aScanner,PRInt32& anErrorCode)=0;
virtual CToken* GetToken(CScanner& aScanner,PRInt32& anErrorCode)=0;
virtual PRBool WillAddToken(CToken& aToken)=0;
virtual eParseMode GetParseMode() const=0;
virtual eParseMode GetParseMode(void) const=0;
virtual nsIDTD* GetDTD(void) const=0;
};

51
htmlparser/src/nsParserTypes.h
Просмотреть файл

@ -40,35 +40,34 @@ enum eParseMode {
eParseMode_other
};
const PRInt32 kNotFound = -1;
const PRInt32 kNoError = 0;
const PRInt32 kEOF = 1000;
const PRInt32 kNotFound = -1;
const PRInt32 kNoError = 0;
const PRInt32 kEOF = 1000000L;
const PRUint32 kNewLine = '\n';
const PRUint32 kCR = '\r';
const PRUint32 kLF = '\n';
const PRUint32 kTab = '\t';
const PRUint32 kSpace = ' ';
const PRUint32 kQuote = '"';
const PRUint32 kApostrophe = '\'';
const PRUint32 kLessThan = '<';
const PRUint32 kGreaterThan = '>';
const PRUint32 kAmpersand = '&';
const PRUint32 kNullPRInt8 = '\0';
const PRUint32 kForwardSlash = '/';
const PRUint32 kEqual = '=';
const PRUint32 kMinus = '-';
const PRUint32 kPlus = '+';
const PRUint32 kExclamation = '!';
const PRUint32 kSemicolon = ';';
const PRUint32 kHashsign = '#';
const PRUint32 kCR = '\r';
const PRUint32 kLF = '\n';
const PRUint32 kTab = '\t';
const PRUint32 kSpace = ' ';
const PRUint32 kQuote = '"';
const PRUint32 kApostrophe = '\'';
const PRUint32 kLessThan = '<';
const PRUint32 kGreaterThan = '>';
const PRUint32 kAmpersand = '&';
const PRUint32 kForwardSlash = '/';
const PRUint32 kEqual = '=';
const PRUint32 kMinus = '-';
const PRUint32 kPlus = '+';
const PRUint32 kExclamation = '!';
const PRUint32 kSemicolon = ';';
const PRUint32 kHashsign = '#';
const PRUint32 kAsterisk = '*';
const PRUint32 kUnderbar = '_';
const PRUint32 kComma = ',';
const PRUint32 kLeftParen = '(';
const PRUint32 kRightParen = ')';
const PRUint32 kLeftBrace = '{';
const PRUint32 kRightBrace = '}';
const PRUint32 kUnderbar = '_';
const PRUint32 kComma = ',';
const PRUint32 kLeftParen = '(';
const PRUint32 kRightParen = ')';
const PRUint32 kLeftBrace = '{';
const PRUint32 kRightBrace = '}';
#endif

6
htmlparser/src/nsTokenHandler.cpp
Просмотреть файл

@ -325,7 +325,7 @@ CWhitespaceTokenHandler::~CWhitespaceTokenHandler(){
PRBool CWhitespaceTokenHandler::operator()(CToken* aToken,nsHTMLParser* aParser){
NS_ASSERTION(0!=aParser,kNullParserGiven);
if(aParser){
return aParser->HandleSimpleContentToken(aToken);
return aParser->HandleStartToken(aToken);
}
return PR_FALSE;
}
@ -375,7 +375,7 @@ CNewlineTokenHandler::~CNewlineTokenHandler(){
PRBool CNewlineTokenHandler::operator()(CToken* aToken,nsHTMLParser* aParser){
NS_ASSERTION(0!=aParser,kNullParserGiven);
if(aParser){
return aParser->HandleSimpleContentToken(aToken);
return aParser->HandleStartToken(aToken);
}
return PR_FALSE;
}
@ -424,7 +424,7 @@ CTextTokenHandler::~CTextTokenHandler(){
PRBool CTextTokenHandler::operator()(CToken* aToken,nsHTMLParser* aParser){
NS_ASSERTION(0!=aParser,kNullParserGiven);
if(aParser){
return aParser->HandleSimpleContentToken(aToken);
return aParser->HandleStartToken(aToken);
}
return PR_FALSE;
}

106
parser/htmlparser/src/CNavDTD.cpp
Просмотреть файл

@ -27,6 +27,7 @@
#include "nsHTMLTokens.h"
#include "nsCRT.h"
#include "nsParserTypes.h"
#include "nsIParser.h"
static NS_DEFINE_IID(kISupportsIID, NS_ISUPPORTS_IID);
static NS_DEFINE_IID(kIDTDIID, NS_IDTD_IID);
@ -98,6 +99,7 @@ NS_IMPL_RELEASE(CNavDTD)
* @return
*/
CNavDTD::CNavDTD() : nsIDTD() {
mParser=0;
}
/**
@ -110,6 +112,36 @@ CNavDTD::CNavDTD() : nsIDTD() {
CNavDTD::~CNavDTD(){
}
/**
*
*
* @update gess 3/25/98
* @param
* @return
*/
void CNavDTD::SetParser(nsIParser* aParser) {
mParser=aParser;
}
static char formElementTags[]= {
eHTMLTag_button, eHTMLTag_fieldset, eHTMLTag_input,
eHTMLTag_isindex, eHTMLTag_label, eHTMLTag_legend,
eHTMLTag_select, eHTMLTag_textarea,0};
/**
* This method is called to determine whether or not a tag
* of one type can contain a tag of another type.
*
* @update gess 4/8/98
* @param aParent -- tag enum of parent container
* @param aChild -- tag enum of child container
* @return PR_TRUE if parent can contain child
*/
PRBool CNavDTD::CanContainFormElement(PRInt32 aParent,PRInt32 aChild) const {
PRBool result=(mParser) ? mParser->HasOpenForm() : PR_FALSE;
return result;
}
/**
* This method is called to determine whether or not a tag
@ -121,6 +153,7 @@ CNavDTD::~CNavDTD(){
* @return PR_TRUE if parent can contain child
*/
PRBool CNavDTD::CanContain(PRInt32 aParent,PRInt32 aChild) const {
PRBool result=PR_FALSE;
//tagset1 has 64 members...
@ -194,6 +227,11 @@ PRBool CNavDTD::CanContain(PRInt32 aParent,PRInt32 aChild) const {
if(eHTMLTag_userdefined==aChild) // XXX Hack: For now...
result=PR_TRUE;
//handle form elements (this is very much a WIP!!!)
if(0!=strchr(formElementTags,aChild)){
return CanContainFormElement(aParent,aChild);
}
switch(aParent) {
case eHTMLTag_a:
result=PRBool(0!=strchr(gTagSet2,aChild)); break;
@ -302,7 +340,15 @@ PRBool CNavDTD::CanContain(PRInt32 aParent,PRInt32 aChild) const {
case eHTMLTag_h1: case eHTMLTag_h2:
case eHTMLTag_h3: case eHTMLTag_h4:
case eHTMLTag_h5: case eHTMLTag_h6:
result=PRBool(0!=strchr(gTagSet1,aChild)); break;
{
static char badTags[]={
eHTMLTag_h1, eHTMLTag_h2, eHTMLTag_h3,
eHTMLTag_h4, eHTMLTag_h5, eHTMLTag_h6, 0};
if(0!=strchr(badTags,aChild))
result=PR_FALSE;
else result=PRBool(0!=strchr(gTagSet1,aChild));
}
break;
case eHTMLTag_head:
{
@ -418,7 +464,7 @@ PRBool CNavDTD::CanContain(PRInt32 aParent,PRInt32 aChild) const {
break; //unadorned script text...
case eHTMLTag_select:
result=PRBool(eHTMLTag_option==aChild); break;
result=PR_TRUE; break; //for now, allow select to contain anything...
case eHTMLTag_small:
result=PRBool(0!=strchr(gTagSet2,aChild)); break;
@ -483,7 +529,7 @@ PRBool CNavDTD::CanContain(PRInt32 aParent,PRInt32 aChild) const {
case eHTMLTag_userdefined:
result=PR_TRUE; break; //XXX for now...
case eHTMLTag_xmp:
case eHTMLTag_xmp:
default:
break;
} //switch
@ -529,6 +575,8 @@ PRBool CNavDTD::CanContainIndirect(PRInt32 aParent,PRInt32 aChild) const {
}
break;
result=PR_TRUE; break;
default:
break;
}
@ -546,29 +594,43 @@ PRBool CNavDTD::CanContainIndirect(PRInt32 aParent,PRInt32 aChild) const {
PRBool CNavDTD::CanOmit(PRInt32 aParent,PRInt32 aChild) const {
PRBool result=PR_FALSE;
switch((eHTMLTags)aParent) {
case eHTMLTag_html:
case eHTMLTag_body:
case eHTMLTag_head:
case eHTMLTag_title:
case eHTMLTag_tr:
// case eHTMLTag_td:
case eHTMLTag_table:
case eHTMLTag_thead:
case eHTMLTag_tfoot:
case eHTMLTag_tbody:
case eHTMLTag_col:
case eHTMLTag_colgroup:
if((aChild==eHTMLTag_newline) ||
(aChild==eHTMLTag_whitespace))
result=PR_TRUE;
//begin with some simple (and obvious) cases...
switch((eHTMLTags)aChild) {
case eHTMLTag_userdefined:
case eHTMLTag_comment:
result=PR_TRUE;
break;
case eHTMLTag_button: case eHTMLTag_fieldset:
case eHTMLTag_input: case eHTMLTag_isindex:
case eHTMLTag_label: case eHTMLTag_legend:
case eHTMLTag_select: case eHTMLTag_textarea:
if(PR_FALSE==mParser->HasOpenForm())
result=PR_TRUE;
break;
case eHTMLTag_newline:
case eHTMLTag_whitespace:
switch((eHTMLTags)aParent) {
case eHTMLTag_html: case eHTMLTag_head:
case eHTMLTag_title: case eHTMLTag_map:
case eHTMLTag_tr: case eHTMLTag_table:
case eHTMLTag_thead: case eHTMLTag_tfoot:
case eHTMLTag_tbody: case eHTMLTag_col:
case eHTMLTag_colgroup: case eHTMLTag_unknown:
result=PR_TRUE;
default:
break;
} //switch
break;
default:
result=PR_FALSE;
if(eHTMLTag_unknown==aParent)
result=PRBool(eHTMLTag_html!=aChild);
break;
}
} //switch
return result;
}

15
parser/htmlparser/src/CNavDTD.h
Просмотреть файл

@ -36,6 +36,7 @@
{0xaa, 0xda, 0x00, 0x80, 0x5f, 0x8a, 0x3e, 0x14}}
class nsIParser;
class CNavDTD : public nsIDTD {
@ -73,6 +74,15 @@ class CNavDTD : public nsIDTD {
*/
virtual PRBool CanContain(PRInt32 aParent,PRInt32 aChild) const;
/**
*
*
* @update gess 3/25/98
* @param
* @return
*/
virtual void SetParser(nsIParser* aParser);
/**
* This method is called to determine whether or not a tag
* of one type can contain a tag of another type.
@ -150,6 +160,11 @@ class CNavDTD : public nsIDTD {
*/
virtual PRBool BackwardPropagate(nsString& aVector,PRInt32 aParentTag,PRInt32 aChildTag) const;
protected:
PRBool CanContainFormElement(PRInt32 aParent,PRInt32 aChild) const;
nsIParser* mParser;
};

4
parser/htmlparser/src/CNavDelegate.cpp
Просмотреть файл

@ -61,7 +61,7 @@ CNavDelegate::CNavDelegate(CNavDelegate& aDelegate) :
* @param
* @return
*/
eParseMode CNavDelegate::GetParseMode() const {
eParseMode CNavDelegate::GetParseMode(void) const {
return eParseMode_unknown;
}
@ -104,7 +104,7 @@ CToken* CNavDelegate::ConsumeTag(PRUnichar aChar,CScanner& aScanner,PRInt32& anE
default:
if(nsString::IsAlpha(aChar))
return ConsumeStartTag(aChar,aScanner,anErrorCode);
else if(kNotFound!=aChar) {
else if(kEOF!=aChar) {
nsAutoString temp("<");
return ConsumeText(temp,aScanner,anErrorCode);
}

11
parser/htmlparser/src/CNavDelegate.h
Просмотреть файл

@ -49,15 +49,15 @@ class CNavDelegate : public ITokenizerDelegate {
CNavDelegate();
CNavDelegate(CNavDelegate& aDelegate);
virtual CToken* GetToken(CScanner& aScanner,PRInt32& anErrorCode);
virtual CToken* GetToken(CScanner& aScanner,PRInt32& anErrorCode);
virtual PRBool WillAddToken(CToken& aToken);
virtual PRBool WillTokenize();
virtual PRBool DidTokenize();
virtual PRBool WillTokenize();
virtual PRBool DidTokenize();
virtual eParseMode GetParseMode() const;
virtual eParseMode GetParseMode(void) const;
virtual nsIDTD* GetDTD(void) const;
static void SelfTest();
static void SelfTest();
protected:
@ -82,4 +82,3 @@ class CNavDelegate : public ITokenizerDelegate {
#endif

16
parser/htmlparser/src/COtherDTD.cpp
Просмотреть файл

@ -27,6 +27,7 @@
#include "nsHTMLTokens.h"
#include "nsCRT.h"
#include "nsParserTypes.h"
#include "nsIParser.h"
static NS_DEFINE_IID(kISupportsIID, NS_ISUPPORTS_IID);
static NS_DEFINE_IID(kIDTDIID, NS_IDTD_IID);
@ -98,6 +99,7 @@ NS_IMPL_RELEASE(COtherDTD)
* @return
*/
COtherDTD::COtherDTD() : nsIDTD() {
mParser=0;
}
/**
@ -110,6 +112,18 @@ COtherDTD::COtherDTD() : nsIDTD() {
COtherDTD::~COtherDTD(){
}
/**
*
*
* @update gess 3/25/98
* @param
* @return
*/
void COtherDTD::SetParser(nsIParser* aParser) {
mParser=aParser;
}
/**
* This method is called to determine whether or not a tag
* of one type can contain a tag of another type.
@ -503,6 +517,8 @@ PRBool COtherDTD::CanContainIndirect(PRInt32 aParent,PRInt32 aChild) const {
}
break;
result=PR_TRUE; break;
default:
break;
}

15
parser/htmlparser/src/COtherDTD.h
Просмотреть файл

@ -37,6 +37,7 @@
class nsIParser;
class COtherDTD : public nsIDTD {
@ -52,7 +53,7 @@ class COtherDTD : public nsIDTD {
* @param
* @return
*/
COtherDTD();
COtherDTD();
/**
*
@ -63,6 +64,16 @@ class COtherDTD : public nsIDTD {
*/
virtual ~COtherDTD();
/**
*
*
* @update gess 3/25/98
* @param
* @return
*/
virtual void SetParser(nsIParser* aParser);
/**
* This method is called to determine whether or not a tag
* of one type can contain a tag of another type.
@ -153,6 +164,8 @@ class COtherDTD : public nsIDTD {
*/
virtual PRBool BackwardPropagate(nsString& aVector,PRInt32 aParentTag,PRInt32 aChildTag) const;
protected:
nsIParser* mParser;
};

4
parser/htmlparser/src/COtherDelegate.cpp
Просмотреть файл

@ -62,7 +62,7 @@ COtherDelegate::COtherDelegate(COtherDelegate& aDelegate) :
* @param
* @return
*/
eParseMode COtherDelegate::GetParseMode() const {
eParseMode COtherDelegate::GetParseMode(void) const {
return eParseMode_unknown;
}
@ -104,7 +104,7 @@ CToken* COtherDelegate::ConsumeTag(PRUnichar aChar,CScanner& aScanner,PRInt32& a
default:
if(nsString::IsAlpha(aChar))
return ConsumeStartTag(aChar,aScanner,anErrorCode);
else if(kNotFound!=aChar) {
else if(kEOF!=aChar) {
nsAutoString temp("<");
return ConsumeText(temp,aScanner,anErrorCode);
}

10
parser/htmlparser/src/COtherDelegate.h
Просмотреть файл

@ -49,15 +49,15 @@ class COtherDelegate : public ITokenizerDelegate {
COtherDelegate();
COtherDelegate(COtherDelegate& aDelegate);
virtual CToken* GetToken(CScanner& aScanner,PRInt32& anErrorCode);
virtual CToken* GetToken(CScanner& aScanner,PRInt32& anErrorCode);
virtual PRBool WillAddToken(CToken& aToken);
virtual PRBool WillTokenize();
virtual PRBool DidTokenize();
virtual PRBool WillTokenize();
virtual PRBool DidTokenize();
virtual eParseMode GetParseMode() const;
virtual eParseMode GetParseMode(void) const;
virtual nsIDTD* GetDTD(void) const;
static void SelfTest();
static void SelfTest();
protected:

202
parser/htmlparser/src/nsHTMLParser.cpp
Просмотреть файл

@ -100,6 +100,7 @@ nsHTMLParser::nsHTMLParser() {
nsCRT::zero(mContextStack,sizeof(mContextStack));
nsCRT::zero(mTokenHandlers,sizeof(mTokenHandlers));
mDTD=0;
mHasOpenForm=PR_FALSE;
mTokenHandlerCount=0;
InitializeDefaultTokenHandlers();
gVerificationOutputDir = PR_GetEnv("VERIFY_PARSER");
@ -176,8 +177,20 @@ nsresult nsHTMLParser::QueryInterface(const nsIID& aIID, void** aInstancePtr)
eHTMLTags nsHTMLParser::NodeAt(PRInt32 aPos) const {
NS_PRECONDITION(0 <= aPos, "bad nodeAt");
if((aPos>-1) && (aPos<mContextStackPos))
return mContextStack[aPos];
return (eHTMLTags)kNotFound;
return (eHTMLTags)mContextStack[aPos];
return eHTMLTag_unknown;
}
/**
* This method allows the caller to determine if a form
* element is currently open.
*
* @update gess 4/2/98
* @param
* @return
*/
PRBool nsHTMLParser::HasOpenForm() const {
return mHasOpenForm;
}
/**
@ -190,10 +203,11 @@ eHTMLTags nsHTMLParser::NodeAt(PRInt32 aPos) const {
*/
eHTMLTags nsHTMLParser::GetTopNode() const {
if(mContextStackPos)
return mContextStack[mContextStackPos-1];
return (eHTMLTags)kNotFound;
return (eHTMLTags)mContextStack[mContextStackPos-1];
return eHTMLTag_unknown;
}
/**
* Determine whether the given tag is open anywhere
* in our context stack.
@ -326,14 +340,13 @@ nsIContentSink* nsHTMLParser::SetContentSink(nsIContentSink* aSink) {
* @param aDTD is the DTD we plan to ask for verification
* @return TRUE if we know how to handle it, else false
*/
PRBool VerifyContextVector(eHTMLTags tags[],PRInt32 count,nsIDTD* aDTD) {
PRBool VerifyContextVector(PRInt32 aTags[],PRInt32 count,nsIDTD* aDTD) {
PRBool result=PR_TRUE;
if(0!=gVerificationOutputDir) {
if(aDTD){
PRInt32 theArray[50];
#ifdef XP_PC
char path[_MAX_PATH+1];
@ -341,17 +354,16 @@ PRBool VerifyContextVector(eHTMLTags tags[],PRInt32 count,nsIDTD* aDTD) {
#endif
for(int i=0;i<count;i++){
theArray[i]=tags[i];
#ifdef NS_WIN32
strcat(path,"/");
const char* name=GetTagName(tags[i]);
const char* name=GetTagName(aTags[i]);
strcat(path,name);
mkdir(path);
#endif
}
//ok, now see if we understand this vector
result=aDTD->VerifyContextVector(theArray,count);
result=aDTD->VerifyContextVector(aTags,count);
}
if(PR_FALSE==result){
//add debugging code here to record the fact that we just encountered
@ -389,9 +401,12 @@ PRBool nsHTMLParser::IterateTokens() {
eHTMLTokenTypes type=eHTMLTokenTypes(theToken->GetTokenType());
iteration++; //debug purposes...
switch(eHTMLTokenTypes(type)){
switch(type){
case eToken_start:
case eToken_text:
case eToken_newline:
case eToken_whitespace:
result=HandleStartToken(theToken); break;
case eToken_end:
@ -399,11 +414,6 @@ PRBool nsHTMLParser::IterateTokens() {
case eToken_entity:
result=HandleEntityToken(theToken); break;
case eToken_text:
case eToken_newline:
case eToken_whitespace:
result=HandleSimpleContentToken(theToken); break;
case eToken_skippedcontent:
//used in cases like <SCRIPT> where we skip over script content.
@ -412,9 +422,6 @@ PRBool nsHTMLParser::IterateTokens() {
case eToken_attribute:
result=HandleAttributeToken(theToken); break;
case eToken_script:
result=HandleScriptToken(theToken); break;
case eToken_style:
result=HandleStyleToken(theToken); break;
@ -428,7 +435,7 @@ PRBool nsHTMLParser::IterateTokens() {
}
VerifyContextVector(mContextStack,mContextStackPos,mDTD);
++(*mCurrentPos);
done=PRBool(e==*mCurrentPos);
done=PRBool(!(*mCurrentPos<e));
}
//One last thing...close any open containers.
@ -438,6 +445,7 @@ PRBool nsHTMLParser::IterateTokens() {
return result;
}
/**
* This is the main controlling routine in the parsing process.
* Note that it may get called multiple times for the same scanner,
@ -454,7 +462,7 @@ PRBool nsHTMLParser::Parse(nsIURL* aURL){
char* theModeStr= PR_GetEnv("PARSE_MODE");
if(theModeStr)
if(0!=PL_strncasecmp("nav",theModeStr,3))
if(0==strnicmp("other",theModeStr,5))
theMode=eParseMode_other;
return Parse(aURL,theMode);
@ -501,6 +509,8 @@ PRBool nsHTMLParser::Parse(nsIURL* aURL,eParseMode aMode){
return PR_FALSE;
}
if(mDTD)
mDTD->SetParser(this);
mTokenizer=new CTokenizer(aURL, theDelegate, mParseMode);
mTokenizer->Tokenize();
result=IterateTokens();
@ -525,6 +535,17 @@ PRBool nsHTMLParser::ResumeParse() {
}
/**
*
* @update gess4/22/98
* @param
* @return
*/
PRInt32 nsHTMLParser::GetStack(PRInt32* aStackPtr) {
aStackPtr=&mContextStack[0];
return mContextStackPos;
}
/**
*
@ -537,15 +558,18 @@ PRInt32 nsHTMLParser::CollectAttributes(nsCParserNode& aNode){
nsDeque& deque=mTokenizer->GetDeque();
nsDequeIterator end=deque.End();
while((*mCurrentPos!=end) && (eToken_attribute==subtype)) {
while((*mCurrentPos<end) && (eToken_attribute==subtype)) {
CHTMLToken* tkn=(CHTMLToken*)(++(*mCurrentPos));
subtype=eHTMLTokenTypes(tkn->GetTokenType());
if(eToken_attribute==subtype) {
aNode.AddAttribute(tkn);
}
else (*mCurrentPos)--;
if(tkn){
subtype=eHTMLTokenTypes(tkn->GetTokenType());
if(eToken_attribute==subtype) {
aNode.AddAttribute(tkn);
}
else (*mCurrentPos)--;
}
}
return aNode.GetAttributeCount();
}
@ -601,7 +625,7 @@ PRBool nsHTMLParser::HandleDefaultStartToken(CToken* aToken,eHTMLTags aChildTag,
//You must determine what container hierarchy you need to hold aToken,
//and create that on the parsestack.
result=ReduceContextStackFor(aChildTag);
if(PR_FALSE==mDTD->CanContain(parentTag,aChildTag)) {
if(PR_FALSE==mDTD->CanContain(GetTopNode(),aChildTag)) {
//we unwound too far; now we have to recreate a valid context stack.
result=CreateContextStackFor(aChildTag);
}
@ -615,35 +639,6 @@ PRBool nsHTMLParser::HandleDefaultStartToken(CToken* aToken,eHTMLTags aChildTag,
result=AddLeaf(aNode);
}
return result;
/*
if(kNotFound!=parentTag) {
if(PR_FALSE==mDTD->CanContain(parentTag,aTag)){
//if you're here, then the new topmost container can't contain aToken.
//You must determine what container hierarchy you need to hold aToken,
//and create that on the parsestack.
ReduceContextStackFor(aTag);
if(PR_FALSE==mDTD->CanContain(parentTag,aTag)) {
//we unwound too far; now we have to recreate a valid context stack.
result=CreateContextStackFor(aTag);
}
}
}
else {
CreateContextStackFor(aTag);
}
//now we think the stack is correct, so fall through
//and push our newest token onto the stack...
if(mDTD->IsContainer(aTag)){
result=OpenContainer(aNode);
}
else {
result=AddLeaf(aNode);
}
return result;
*/
}
/**
@ -672,6 +667,11 @@ PRBool nsHTMLParser::HandleStartToken(CToken* aToken) {
nsCParserNode attrNode((CHTMLToken*)aToken);
PRInt32 attrCount=CollectAttributes(attrNode);
//now check to see if this token should be omitted...
if(PR_TRUE==mDTD->CanOmit(GetTopNode(),tokenTagType)) {
return PR_TRUE;
}
switch(tokenTagType) {
case eHTMLTag_html:
@ -695,7 +695,7 @@ PRBool nsHTMLParser::HandleStartToken(CToken* aToken) {
break;
case eHTMLTag_form:
result = mSink->OpenForm(attrNode);
result = OpenForm(attrNode);
break;
case eHTMLTag_meta:
@ -718,6 +718,10 @@ PRBool nsHTMLParser::HandleStartToken(CToken* aToken) {
}
break;
case eHTMLTag_script:
result=HandleScriptToken(st); break;
case eHTMLTag_map:
// Put map into the head section
result=OpenHead(attrNode);
@ -727,6 +731,9 @@ PRBool nsHTMLParser::HandleStartToken(CToken* aToken) {
case eHTMLTag_head:
result=PR_TRUE; break; //ignore head tags...
case eHTMLTag_select:
result=PR_FALSE;
default:
result=HandleDefaultStartToken(aToken,tokenTagType,attrNode);
break;
@ -755,6 +762,11 @@ PRBool nsHTMLParser::HandleEndToken(CToken* aToken) {
CEndToken* st = (CEndToken*)(aToken);
eHTMLTags tokenTagType=st->GetHTMLTag();
//now check to see if this token should be omitted...
if(PR_TRUE==mDTD->CanOmit(GetTopNode(),tokenTagType)) {
return PR_TRUE;
}
nsCParserNode theNode((CHTMLToken*)aToken);
switch(tokenTagType) {
case eHTMLTag_html:
@ -780,15 +792,19 @@ PRBool nsHTMLParser::HandleEndToken(CToken* aToken) {
case eHTMLTag_form:
{
nsCParserNode aNode((CHTMLToken*)aToken);
result=mSink->CloseForm(aNode);
result=CloseForm(aNode);
}
break;
case eHTMLTag_script:
result=PR_TRUE; break;
default:
if(mDTD->IsContainer(tokenTagType)){
result=CloseContainer(theNode);
}
result=PR_TRUE;
//actually, you want to push the start token and ALL it's attributes...
//
break;
}
return result;
@ -826,43 +842,6 @@ PRBool nsHTMLParser::HandleCommentToken(CToken* aToken) {
return PR_TRUE;
}
/**
* This method gets called when a text, whitespace and newline
* tokens have been encountered in the parse process. After
* verifying that the topmost container can contain this data,
* we call AddLeaf to store this token in the top container.
*
* @update gess 3/25/98
* @param aToken -- next (text) token to be handled
* @return PR_TRUE if all went well; PR_FALSE if error occured
*/
PRBool nsHTMLParser::HandleSimpleContentToken(CToken* aToken) {
NS_PRECONDITION(0!=aToken,kNullToken);
CHTMLToken* theToken=(CHTMLToken*)aToken;
eHTMLTags type=theToken->GetHTMLTag();
PRBool result=PR_TRUE;
nsCParserNode aNode(theToken);
switch(type) {
case eHTMLTag_text:
result=HandleDefaultStartToken(theToken,type,aNode);
break;
case eHTMLTag_newline:
case eHTMLTag_whitespace:
if((mContextStackPos>0) &&
(PR_FALSE==mDTD->CanOmit(mContextStack[mContextStackPos-1],type))) {
result=HandleDefaultStartToken(theToken,type,aNode);
}
break;
default:
break;
} //switch
return result;
}
/**
* This method gets called when a skippedcontent token has
* been encountered in the parse process. After verifying
@ -996,6 +975,7 @@ PRBool nsHTMLParser::CloseHTML(const nsIParserNode& aNode){
* @return TRUE if ok, FALSE if error
*/
PRBool nsHTMLParser::OpenHead(const nsIParserNode& aNode){
mContextStack[mContextStackPos++]=eHTMLTag_head;
PRBool result=mSink->OpenHead(aNode);
return result;
}
@ -1010,6 +990,7 @@ PRBool nsHTMLParser::OpenHead(const nsIParserNode& aNode){
*/
PRBool nsHTMLParser::CloseHead(const nsIParserNode& aNode){
PRBool result=mSink->CloseHead(aNode);
mContextStack[--mContextStackPos]=eHTMLTag_unknown;
return result;
}
@ -1084,10 +1065,10 @@ PRBool nsHTMLParser::CloseBody(const nsIParserNode& aNode){
* @return TRUE if ok, FALSE if error
*/
PRBool nsHTMLParser::OpenForm(const nsIParserNode& aNode){
NS_PRECONDITION(mContextStackPos >= 0, kInvalidTagStackPos);
PRBool result=mSink->OpenForm(aNode);
mContextStack[mContextStackPos++]=(eHTMLTags)aNode.GetNodeType();
return result;
if(mHasOpenForm)
CloseForm(aNode);
mHasOpenForm=PR_TRUE;
return mSink->OpenForm(aNode);
}
/**
@ -1099,9 +1080,11 @@ PRBool nsHTMLParser::OpenForm(const nsIParserNode& aNode){
* @return TRUE if ok, FALSE if error
*/
PRBool nsHTMLParser::CloseForm(const nsIParserNode& aNode){
NS_PRECONDITION(mContextStackPos > 0, kInvalidTagStackPos);
PRBool result=mSink->CloseContainer(aNode);
mContextStack[--mContextStackPos]=eHTMLTag_unknown;
PRBool result=PR_TRUE;
if(mHasOpenForm) {
mHasOpenForm=PR_FALSE;
result=mSink->CloseForm(aNode);
}
return result;
}
@ -1230,12 +1213,12 @@ PRBool nsHTMLParser::CloseContainersTo(PRInt32 anIndex){
PRBool result=PR_TRUE;
nsAutoString empty;
CHTMLToken aToken(empty);
CEndToken aToken(empty);
nsCParserNode theNode(&aToken);
if((anIndex<mContextStackPos) && (anIndex>=0)) {
while(mContextStackPos>anIndex) {
aToken.SetHTMLTag(mContextStack[mContextStackPos-1]);
aToken.SetHTMLTag((eHTMLTags)mContextStack[mContextStackPos-1]);
result=CloseContainer(theNode);
}
}
@ -1287,7 +1270,7 @@ PRBool nsHTMLParser::CloseTopmostContainer(){
nsAutoString empty;
CEndToken aToken(empty);
aToken.SetHTMLTag(mContextStack[mContextStackPos-1]);
aToken.SetHTMLTag((eHTMLTags)mContextStack[mContextStackPos-1]);
nsCParserNode theNode(&aToken);
return CloseContainer(theNode);
@ -1323,14 +1306,17 @@ PRBool nsHTMLParser::CreateContextStackFor(PRInt32 aChildTag){
PRBool result=PR_FALSE;
PRInt32 pos=0;
PRInt32 cnt=0;
PRInt32 theTop=GetTopNode();
if(PR_TRUE==mDTD->ForwardPropagate(theVector,GetTopNode(),aChildTag)){
if(PR_TRUE==mDTD->ForwardPropagate(theVector,theTop,aChildTag)){
//add code here to build up context stack based on forward propagated context vector...
pos=0;
cnt=theVector.Length()-1;
result=PRBool(mContextStack[mContextStackPos-1]==theVector[cnt]);
}
else if(PR_TRUE==mDTD->BackwardPropagate(theVector,eHTMLTag_html,aChildTag)) {
else if(PR_TRUE==mDTD->BackwardPropagate(theVector,theTop,aChildTag)) {
if(theTop!=eHTMLTag_html)
mDTD->BackwardPropagate(theVector,eHTMLTag_html,theTop);
//propagation worked, so pop unwanted containers, push new ones, then exit...
pos=0;

9
parser/htmlparser/src/nsHTMLParser.h
Просмотреть файл

@ -87,13 +87,15 @@ friend class CTokenHandler;
virtual PRBool Parse(nsIURL* aURL);
virtual PRBool Parse(nsIURL* aURL,eParseMode aMode);
virtual PRBool ResumeParse();
virtual PRInt32 GetStack(PRInt32* aStackPtr);
virtual PRBool HasOpenForm() const;
PRBool HandleStartToken(CToken* aToken);
PRBool HandleDefaultStartToken(CToken* aToken,eHTMLTags aChildTag,nsCParserNode& aNode);
PRBool HandleEndToken(CToken* aToken);
PRBool HandleEntityToken(CToken* aToken);
PRBool HandleCommentToken(CToken* aToken);
PRBool HandleSimpleContentToken(CToken* aToken);
PRBool HandleSkippedContentToken(CToken* aToken);
PRBool HandleAttributeToken(CToken* aToken);
PRBool HandleScriptToken(CToken* aToken);
@ -140,12 +142,12 @@ friend class CTokenHandler;
PRInt32 GetTopmostIndex(eHTMLTags aTag) const;
PRBool ReduceContextStackFor(PRInt32 aChildTag);
PRBool CreateContextStackFor(PRInt32 aChildTag);
PRBool HandleDefaultStartToken(CToken* aToken,eHTMLTags aTag,nsCParserNode& aNode);
nsIHTMLContentSink* mSink;
CTokenizer* mTokenizer;
eHTMLTags mContextStack[50];
// eHTMLTags mContextStack[50];
PRInt32 mContextStack[50];
PRInt32 mContextStackPos;
CTokenHandler* mTokenHandlers[100];
@ -154,6 +156,7 @@ friend class CTokenHandler;
nsIDTD* mDTD;
eParseMode mParseMode;
PRBool mHasOpenForm;
};

5
parser/htmlparser/src/nsHTMLTokens.cpp
Просмотреть файл

@ -547,6 +547,7 @@ PRInt32 CTextToken::Consume(PRUnichar aChar, CScanner& aScanner) {
*------------------------------------------------------*/
CCommentToken::CCommentToken(const nsString& aName) : CHTMLToken(aName) {
mOrdinalValue=eToken_comment;
mTagType=eHTMLTag_comment;
}
/**-------------------------------------------------------
@ -561,11 +562,11 @@ CCommentToken::CCommentToken(const nsString& aName) : CHTMLToken(aName) {
PRInt32 CCommentToken::Consume(PRUnichar aChar, CScanner& aScanner) {
PRUnichar ch,ch2;
PRInt32 result;
static nsAutoString terminals(">");
aScanner.GetChar(ch);
PRInt32 result=aScanner.GetChar(ch);
mTextValue="<!";
if(kMinus==ch) {
aScanner.GetChar(ch2);
@ -1285,7 +1286,7 @@ eHTMLTags DetermineHTMLTagType(const nsString& aString)
* @param
* @return
*/
const char* GetTagName(eHTMLTags aTag) {
const char* GetTagName(PRInt32 aTag) {
const char* result=0;
PRInt32 cnt=sizeof(gHTMLTagTable)/sizeof(HTMLTagEntry);

11
parser/htmlparser/src/nsHTMLTokens.h
Просмотреть файл

@ -58,9 +58,10 @@ enum eHTMLTags
eHTMLTag_blink, eHTMLTag_blockquote, eHTMLTag_body, eHTMLTag_br,
eHTMLTag_button, eHTMLTag_caption, eHTMLTag_center,
eHTMLTag_certificate, eHTMLTag_cite,
eHTMLTag_code, eHTMLTag_col, eHTMLTag_colgroup, eHTMLTag_dd,
eHTMLTag_del, eHTMLTag_dfn, eHTMLTag_div, eHTMLTag_dir,
eHTMLTag_dl, eHTMLTag_dt, eHTMLTag_em, eHTMLTag_embed,
eHTMLTag_code, eHTMLTag_col, eHTMLTag_colgroup, eHTMLTag_comment,
eHTMLTag_dd, eHTMLTag_del, eHTMLTag_dfn, eHTMLTag_div,
eHTMLTag_dir, eHTMLTag_dl, eHTMLTag_dt,
eHTMLTag_em, eHTMLTag_embed,
eHTMLTag_fieldset, eHTMLTag_font, eHTMLTag_footer,
eHTMLTag_form, eHTMLTag_frame, eHTMLTag_frameset,
eHTMLTag_h1, eHTMLTag_h2, eHTMLTag_h3, eHTMLTag_h4,
@ -77,7 +78,7 @@ enum eHTMLTags
eHTMLTag_option, eHTMLTag_paragraph, eHTMLTag_param, eHTMLTag_plaintext,
eHTMLTag_pre, eHTMLTag_quotation, eHTMLTag_strike, eHTMLTag_samp,
eHTMLTag_script, eHTMLTag_select,
eHTMLTag_server, eHTMLTag_small,
eHTMLTag_server, eHTMLTag_small,
eHTMLTag_spacer, eHTMLTag_span,
eHTMLTag_strong, eHTMLTag_style, eHTMLTag_sub, eHTMLTag_sup,
eHTMLTag_table, eHTMLTag_tbody, eHTMLTag_td,
@ -113,7 +114,7 @@ PRInt32 ConsumeAttributeText(PRUnichar aChar,nsString& aString,CScanner&
PRInt32 FindEntityIndex(const char* aBuffer,PRInt32 aBufLen=-1);
eHTMLTags DetermineHTMLTagType(const nsString& aString);
eHTMLTokenTypes DetermineTokenType(const nsString& aString);
const char* GetTagName(eHTMLTags aTag);
const char* GetTagName(PRInt32 aTag);
/** -----------------------------------------------------

14
parser/htmlparser/src/nsIDTD.h
Просмотреть файл

@ -35,6 +35,7 @@
{0xaa, 0xda, 0x00, 0x80, 0x5f, 0x8a, 0x3e, 0x14}}
class nsIParser;
class nsIDTD : public nsISupports {
@ -49,7 +50,18 @@ class nsIDTD : public nsISupports {
* @param aChild -- tag enum of child container
* @return PR_TRUE if parent can contain child
*/
virtual PRBool CanContain(PRInt32 aParent,PRInt32 aChild) const =0;
virtual PRBool CanContain(PRInt32 aParent,PRInt32 aChild) const=0;
/**
* This method is called to determine whether or not a tag
* of one type can contain a tag of another type.
*
* @update gess 3/25/98
* @param aParent -- tag enum of parent container
* @param aChild -- tag enum of child container
* @return PR_TRUE if parent can contain child
*/
virtual void SetParser(nsIParser* aParser)=0;
/**
* This method is called to determine whether or not a tag

2
parser/htmlparser/src/nsIParser.h
Просмотреть файл

@ -54,6 +54,8 @@ class nsIParser : public nsISupports {
virtual nsIContentSink* SetContentSink(nsIContentSink* aContentSink)=0;
virtual PRBool Parse(nsIURL* aURL)=0;
virtual PRBool ResumeParse()=0;
virtual PRInt32 GetStack(PRInt32* aStackPtr)=0;
virtual PRBool HasOpenForm() const=0;
};
extern NS_HTMLPARS nsresult NS_NewHTMLParser(nsIParser** aInstancePtrResult);

8
parser/htmlparser/src/nsITokenizerDelegate.h
Просмотреть файл

@ -48,13 +48,13 @@ class CToken;
class ITokenizerDelegate {
public:
virtual PRBool WillTokenize()=0;
virtual PRBool DidTokenize()=0;
virtual PRBool WillTokenize()=0;
virtual PRBool DidTokenize()=0;
virtual CToken* GetToken(CScanner& aScanner,PRInt32& anErrorCode)=0;
virtual CToken* GetToken(CScanner& aScanner,PRInt32& anErrorCode)=0;
virtual PRBool WillAddToken(CToken& aToken)=0;
virtual eParseMode GetParseMode() const=0;
virtual eParseMode GetParseMode(void) const=0;
virtual nsIDTD* GetDTD(void) const=0;
};

51
parser/htmlparser/src/nsParserTypes.h
Просмотреть файл

@ -40,35 +40,34 @@ enum eParseMode {
eParseMode_other
};
const PRInt32 kNotFound = -1;
const PRInt32 kNoError = 0;
const PRInt32 kEOF = 1000;
const PRInt32 kNotFound = -1;
const PRInt32 kNoError = 0;
const PRInt32 kEOF = 1000000L;
const PRUint32 kNewLine = '\n';
const PRUint32 kCR = '\r';
const PRUint32 kLF = '\n';
const PRUint32 kTab = '\t';
const PRUint32 kSpace = ' ';
const PRUint32 kQuote = '"';
const PRUint32 kApostrophe = '\'';
const PRUint32 kLessThan = '<';
const PRUint32 kGreaterThan = '>';
const PRUint32 kAmpersand = '&';
const PRUint32 kNullPRInt8 = '\0';
const PRUint32 kForwardSlash = '/';
const PRUint32 kEqual = '=';
const PRUint32 kMinus = '-';
const PRUint32 kPlus = '+';
const PRUint32 kExclamation = '!';
const PRUint32 kSemicolon = ';';
const PRUint32 kHashsign = '#';
const PRUint32 kCR = '\r';
const PRUint32 kLF = '\n';
const PRUint32 kTab = '\t';
const PRUint32 kSpace = ' ';
const PRUint32 kQuote = '"';
const PRUint32 kApostrophe = '\'';
const PRUint32 kLessThan = '<';
const PRUint32 kGreaterThan = '>';
const PRUint32 kAmpersand = '&';
const PRUint32 kForwardSlash = '/';
const PRUint32 kEqual = '=';
const PRUint32 kMinus = '-';
const PRUint32 kPlus = '+';
const PRUint32 kExclamation = '!';
const PRUint32 kSemicolon = ';';
const PRUint32 kHashsign = '#';
const PRUint32 kAsterisk = '*';
const PRUint32 kUnderbar = '_';
const PRUint32 kComma = ',';
const PRUint32 kLeftParen = '(';
const PRUint32 kRightParen = ')';
const PRUint32 kLeftBrace = '{';
const PRUint32 kRightBrace = '}';
const PRUint32 kUnderbar = '_';
const PRUint32 kComma = ',';
const PRUint32 kLeftParen = '(';
const PRUint32 kRightParen = ')';
const PRUint32 kLeftBrace = '{';
const PRUint32 kRightBrace = '}';
#endif

6
parser/htmlparser/src/nsTokenHandler.cpp
Просмотреть файл

@ -325,7 +325,7 @@ CWhitespaceTokenHandler::~CWhitespaceTokenHandler(){
PRBool CWhitespaceTokenHandler::operator()(CToken* aToken,nsHTMLParser* aParser){
NS_ASSERTION(0!=aParser,kNullParserGiven);
if(aParser){
return aParser->HandleSimpleContentToken(aToken);
return aParser->HandleStartToken(aToken);
}
return PR_FALSE;
}
@ -375,7 +375,7 @@ CNewlineTokenHandler::~CNewlineTokenHandler(){
PRBool CNewlineTokenHandler::operator()(CToken* aToken,nsHTMLParser* aParser){
NS_ASSERTION(0!=aParser,kNullParserGiven);
if(aParser){
return aParser->HandleSimpleContentToken(aToken);
return aParser->HandleStartToken(aToken);
}
return PR_FALSE;
}
@ -424,7 +424,7 @@ CTextTokenHandler::~CTextTokenHandler(){
PRBool CTextTokenHandler::operator()(CToken* aToken,nsHTMLParser* aParser){
NS_ASSERTION(0!=aParser,kNullParserGiven);
if(aParser){
return aParser->HandleSimpleContentToken(aToken);
return aParser->HandleStartToken(aToken);
}
return PR_FALSE;
}