dynamic DTD support

git-svn-id: svn://10.0.0.236/trunk@5278 18797224-902f-48f8-a5cc-f745e15eee43

mozilla/content/html/document/src/nsHTMLContentSink.cpp

@@ -904,7 +904,7 @@ void HTMLContentSink::ReflowNewContent()
       shell->ContentAppended(mBody);
       NS_RELEASE(shell);
     }
-  }
+  } 
 #endif
 }
 

mozilla/content/html/document/src/nsHTMLDocument.cpp

@@ -36,6 +36,12 @@
 #include "nsIWebWidget.h"
 #include "CNavDTD.h"
 
+
+//#define rickgdebug 1
+#ifdef rickgdebug
+#include "nsHTMLContentSinkStream.h"
+#endif
+
 static NS_DEFINE_IID(kIDocumentIID, NS_IDOCUMENT_IID);
 static NS_DEFINE_IID(kIDOMElementIID, NS_IDOMELEMENT_IID);
 static NS_DEFINE_IID(kIDOMTextIID, NS_IDOMTEXT_IID);
@@ -104,6 +110,13 @@ nsHTMLDocument::StartDocumentLoad(nsIURL *aURL,
   nsresult rv = NS_NewParser(&parser);
   if (NS_OK == rv) {
     nsIHTMLContentSink* sink;
+
+#ifdef rickgdebug
+    rv = NS_New_HTML_ContentSinkStream(&sink);
+#else
+    rv = NS_NewHTMLContentSink(&sink, this, aURL, aWebWidget);
+#endif
+
     rv = NS_NewHTMLContentSink(&sink, this, aURL, aWebWidget);
     if (NS_OK == rv) {
       nsIHTMLCSSStyleSheet* styleAttrSheet;
@@ -129,10 +142,9 @@ nsHTMLDocument::StartDocumentLoad(nsIURL *aURL,
         nsIDTD* theDTD=0;
         NS_NewNavHTMLDTD(&theDTD);
         parser->RegisterDTD(theDTD);
-        parser->RegisterDTD(theDTD);
 
         parser->SetContentSink(sink);
-        parser->BeginParse(aURL);
+        parser->Parse(aURL);
       }
       NS_RELEASE(sink);
     }

mozilla/htmlparser/src/CNavDTD.cpp

@@ -282,18 +282,33 @@ void CNavDTD::SetDTDDebug(nsIDTDDebug * aDTDDebug)
 }
 
 /**
- * This method gets called by the parser to determine if this DTD can handle 
- * the requested process for the requested content type.
- *
+ * This method is called to determine if the given DTD can parse
+ * a document in a given source-type. 
+ * NOTE: Parsing always assumes that the end result will involve
+ *       storing the result in the main content model.
  * @update	gess6/24/98
- * @param 
- * @return  TRUE if the DTD can handle the process for this type; FALSE otherwise.
+ * @param   
+ * @return  TRUE if this DTD can satisfy the request; FALSE otherwise.
  */
-PRBool CNavDTD::IsCapableOf(eProcessType aProcessType, nsString& aContentType,PRInt32 aVersion){
+PRBool CNavDTD::CanParse(nsString& aContentType, PRInt32 aVersion){
   PRBool result=aContentType.Equals(kHTMLTextContentType);
   return result;
 }
 
+/**
+ * 
+ * @update	gess7/7/98
+ * @param 
+ * @return
+ */
+eAutoDetectResult CNavDTD::AutoDetectContentType(nsString& aBuffer,nsString& aType){
+  eAutoDetectResult result=eUnknownDetect;
+  if(PR_TRUE==aType.Equals(kHTMLTextContentType)) 
+    result=eValidDetect;
+  return result;
+}
+
+
 
 /**
  * 

mozilla/htmlparser/src/CNavDTD.h

@@ -70,13 +70,24 @@ class CNavDTD : public nsIDTD {
      */
     virtual ~CNavDTD();
 
+    /**
+     * This method is called to determine if the given DTD can parse
+     * a document in a given source-type. 
+     * NOTE: Parsing always assumes that the end result will involve
+     *       storing the result in the main content model.
+     * @update	gess6/24/98
+     * @param   
+     * @return  TRUE if this DTD can satisfy the request; FALSE otherwise.
+     */
+    virtual PRBool CanParse(nsString& aContentType, PRInt32 aVersion);
+
     /**
      * 
-     * @update	gess7/1/98
+     * @update	gess7/7/98
      * @param 
      * @return
      */
-    virtual PRBool IsCapableOf(eProcessType aProcessType, nsString& aString, PRInt32 aVersion);
+    virtual eAutoDetectResult AutoDetectContentType(nsString& aBuffer,nsString& aType);
 
     /**
      * 

mozilla/htmlparser/src/COtherDTD.cpp

@@ -149,17 +149,29 @@ COtherDTD::~COtherDTD(){
 }
 
 /**
- * This method gets called by the parser to determine if this DTD can handle 
- * the requested process for the requested content type.
- *
+ * This method is called to determine if the given DTD can parse
+ * a document in a given source-type. 
+ * NOTE: Parsing always assumes that the end result will involve
+ *       storing the result in the main content model.
  * @update	gess6/24/98
- * @param 
- * @return  TRUE if the DTD can handle the process for this type; FALSE otherwise.
+ * @param   
+ * @return  TRUE if this DTD can satisfy the request; FALSE otherwise.
  */
-PRBool COtherDTD::IsCapableOf(eProcessType aProcessType, nsString& aContentType,PRInt32 aVersion){
-  return CNavDTD::IsCapableOf(aProcessType,aContentType,aVersion);
+PRBool COtherDTD::CanParse(nsString& aContentType, PRInt32 aVersion){
+  return CNavDTD::CanParse(aContentType,aVersion);
 }
 
+/**
+ * 
+ * @update	gess7/7/98
+ * @param 
+ * @return
+ */
+eAutoDetectResult COtherDTD::AutoDetectContentType(nsString& aBuffer,nsString& aType){
+  return CNavDTD::AutoDetectContentType(aBuffer,aType);
+}
+
+
 /**
  * 
  * @update	gess5/18/98

mozilla/htmlparser/src/COtherDTD.h

@@ -62,7 +62,24 @@ class COtherDTD : public CNavDTD {
      */
     virtual ~COtherDTD();
 
-    virtual PRBool IsCapableOf(eProcessType aProcessType, nsString& aString,PRInt32 aVersion);
+    /**
+     * This method is called to determine if the given DTD can parse
+     * a document in a given source-type. 
+     * NOTE: Parsing always assumes that the end result will involve
+     *       storing the result in the main content model.
+     * @update	gess6/24/98
+     * @param   
+     * @return  TRUE if this DTD can satisfy the request; FALSE otherwise.
+     */
+    virtual PRBool CanParse(nsString& aContentType, PRInt32 aVersion);
+
+    /**
+     * 
+     * @update	gess7/7/98
+     * @param 
+     * @return
+     */
+    virtual eAutoDetectResult AutoDetectContentType(nsString& aBuffer,nsString& aType);
     
     /**
      * 

mozilla/htmlparser/src/CRtfDTD.cpp

@@ -0,0 +1,622 @@
+/* -*- Mode: C++; tab-width: 2; indent-tabs-mode: nil; c-basic-offset: 2 -*- */
+/*
+ * The contents of this file are subject to the Netscape Public License
+ * Version 1.0 (the "NPL"); you may not use this file except in
+ * compliance with the NPL.  You may obtain a copy of the NPL at
+ * http://www.mozilla.org/NPL/
+ *
+ * Software distributed under the NPL is distributed on an "AS IS" basis,
+ * WITHOUT WARRANTY OF ANY KIND, either express or implied. See the NPL
+ * for the specific language governing rights and limitations under the
+ * NPL.
+ *
+ * The Initial Developer of this code under the NPL is Netscape
+ * Communications Corporation.  Portions created by Netscape are
+ * Copyright (C) 1998 Netscape Communications Corporation.  All Rights
+ * Reserved.
+ */
+
+/**
+ * MODULE NOTES:
+ * @update  gess 4/8/98
+ * 
+ *         
+ */
+
+/**
+ * TRANSIENT STYLE-HANDLING NOTES:
+ * @update  gess 6/15/98
+ * 
+ * ...add comments here about transient style stack.
+ *         
+   */
+
+#include "nsIDTDDebug.h"
+#include "CRtfDTD.h"
+#include "nsCRT.h"
+#include "nsParser.h"
+#include "nsScanner.h"
+#include "nsParserTypes.h"
+#include "nsTokenHandler.h"
+
+#include "prenv.h"  //this is here for debug reasons...
+#include "prtypes.h"  //this is here for debug reasons...
+#include "prio.h"
+#include "plstr.h"
+
+#ifdef XP_PC
+#include <direct.h> //this is here for debug reasons...
+#endif
+#include "prmem.h"
+
+
+static NS_DEFINE_IID(kISupportsIID, NS_ISUPPORTS_IID);                 
+static NS_DEFINE_IID(kIDTDIID,      NS_IDTD_IID);
+static NS_DEFINE_IID(kClassIID,     NS_RTF_DTD_IID); 
+
+
+static const char* kNullURL = "Error: Null URL given";
+static const char* kNullFilename= "Error: Null filename given";
+static const char* kNullTokenizer = "Error: Unable to construct tokenizer";
+static const char* kNullToken = "Error: Null token given";
+static const char* kInvalidTagStackPos = "Error: invalid tag stack position";
+static const char* kRTFTextContentType = "application/rtf";
+static const char* kRTFDocHeader= "{\\rtf0";
+static nsString     gAlphaChars("abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ");
+static nsAutoString gDigits("-0123456789");
+
+static nsAutoString gEmpty;
+
+struct RTFEntry {
+  char      mName[10];
+  eRTFTags  mTagID;
+};
+
+
+static RTFEntry gRTFTable[] = {
+
+  {"$",eRTFCtrl_unknown},
+  {"'",eRTFCtrl_quote},
+  {"*",eRTFCtrl_star},   
+  {"0x0a",eRTFCtrl_linefeed},
+  {"0x0d",eRTFCtrl_return},
+  {"\\",eRTFCtrl_begincontrol},
+  {"b",eRTFCtrl_bold}, 
+  {"bin",eRTFCtrl_bin},
+  {"blue",eRTFCtrl_blue},
+  {"cols",eRTFCtrl_cols},
+  {"comment",eRTFCtrl_comment},
+  {"f",eRTFCtrl_font},
+  {"fonttbl",eRTFCtrl_fonttable},
+  {"green",eRTFCtrl_green},
+  {"i",eRTFCtrl_italic},
+  {"margb",eRTFCtrl_bottommargin},
+  {"margl",eRTFCtrl_leftmargin},
+  {"margr",eRTFCtrl_rightmargin},
+  {"margt",eRTFCtrl_topmargin},
+  {"par",eRTFCtrl_par},
+  {"pard",eRTFCtrl_pard},
+  {"plain",eRTFCtrl_plain},
+  {"qc",eRTFCtrl_justified},
+  {"qj",eRTFCtrl_fulljustified},
+  {"ql",eRTFCtrl_leftjustified},
+  {"qr",eRTFCtrl_rightjustified},
+  {"rdblquote",eRTFCtrl_rdoublequote},
+  {"red",eRTFCtrl_red},
+  {"rtf",eRTFCtrl_rtf},
+  {"tab",eRTFCtrl_tab},
+  {"title",eRTFCtrl_title},
+  {"u",eRTFCtrl_underline},
+  {"{",eRTFCtrl_startgroup},
+  {"}",eRTFCtrl_endgroup},
+  {"~",eRTFCtrl_last} //make sure this stays the last token...
+};
+
+/**
+ * 
+ * @update	gess4/25/98
+ * @param 
+ * @return
+ */
+const char* GetTagName(eRTFTags aTag) {
+  const    char* result=0;
+  PRInt32  cnt=sizeof(gRTFTable)/sizeof(RTFEntry);
+  PRInt32  low=0; 
+  PRInt32  high=cnt-1;
+  PRInt32  middle=kNotFound;
+  
+  while(low<=high) {
+    middle=(PRInt32)(low+high)/2;
+    if(aTag==gRTFTable[middle].mTagID)   
+      return gRTFTable[middle].mName;
+    if(aTag<gRTFTable[middle].mTagID)
+      high=middle-1; 
+    else low=middle+1; 
+  }
+  return "";
+}
+
+/**
+ *  This method gets called as part of our COM-like interfaces.
+ *  Its purpose is to create an interface to parser object
+ *  of some type.
+ *  
+ *  @update   gess 4/8/98
+ *  @param    nsIID  id of object to discover
+ *  @param    aInstancePtr ptr to newly discovered interface
+ *  @return   NS_xxx result code
+ */
+nsresult CRtfDTD::QueryInterface(const nsIID& aIID, void** aInstancePtr)  
+{                                                                        
+  if (NULL == aInstancePtr) {                                            
+    return NS_ERROR_NULL_POINTER;                                        
+  }                                                                      
+
+  if(aIID.Equals(kISupportsIID))    {  //do IUnknown...
+    *aInstancePtr = (nsIDTD*)(this);                                        
+  }
+  else if(aIID.Equals(kIDTDIID)) {  //do IParser base class...
+    *aInstancePtr = (nsIDTD*)(this);                                        
+  }
+  else if(aIID.Equals(kClassIID)) {  //do this class...
+    *aInstancePtr = (CRtfDTD*)(this);                                        
+  }                 
+  else {
+    *aInstancePtr=0;
+    return NS_NOINTERFACE;
+  }
+  ((nsISupports*) *aInstancePtr)->AddRef();
+  return NS_OK;                                                        
+}
+
+/**
+ *  This method is defined in nsIParser. It is used to 
+ *  cause the COM-like construction of an nsParser.
+ *  
+ *  @update  gess 4/8/98
+ *  @param   nsIParser** ptr to newly instantiated parser
+ *  @return  NS_xxx error result
+ */
+NS_HTMLPARS nsresult NS_NewRTF_DTD(nsIDTD** aInstancePtrResult)
+{
+  CRtfDTD* it = new CRtfDTD();
+
+  if (it == 0) {
+    return NS_ERROR_OUT_OF_MEMORY;
+  }
+
+  return it->QueryInterface(kClassIID, (void **) aInstancePtrResult);
+}
+
+
+NS_IMPL_ADDREF(CRtfDTD)
+NS_IMPL_RELEASE(CRtfDTD)
+
+
+/**
+ *  Default constructor
+ *  
+ *  @update  gess 4/9/98
+ *  @param   
+ *  @return  
+ */
+CRtfDTD::CRtfDTD() : nsIDTD() {
+  NS_INIT_REFCNT();
+  mParser=0;
+  mFilename=0;
+}
+
+/**
+ *  Default destructor
+ *  
+ *  @update  gess 4/9/98
+ *  @param   
+ *  @return  
+ */
+CRtfDTD::~CRtfDTD(){
+}
+
+/**
+ * This method is called to determine if the given DTD can parse
+ * a document in a given source-type. 
+ * NOTE: Parsing always assumes that the end result will involve
+ *       storing the result in the main content model.
+ * @update	gess6/24/98
+ * @param   
+ * @return  TRUE if this DTD can satisfy the request; FALSE otherwise.
+ */
+PRBool CRtfDTD::CanParse(nsString& aContentType, PRInt32 aVersion){
+  PRBool result=aContentType.Equals(kRTFTextContentType);
+  return result;
+}
+
+
+/**
+ * 
+ * @update	gess7/7/98
+ * @param 
+ * @return
+ */
+eAutoDetectResult CRtfDTD::AutoDetectContentType(nsString& aBuffer,nsString& aType){
+  eAutoDetectResult result=eUnknownDetect;
+  if(kNotFound!=aBuffer.Find(kRTFDocHeader))
+    if(PR_TRUE==aType.Equals(kRTFTextContentType))
+      result=eValidDetect;
+  return result;
+}
+
+/**
+ * 
+ * @update	gess5/18/98
+ * @param 
+ * @return
+ */
+PRInt32 CRtfDTD::WillBuildModel(const char* aFilename){
+  PRInt32 result=0;
+  return result;
+}
+
+/**
+ * 
+ * @update	gess5/18/98
+ * @param 
+ * @return
+ */
+PRInt32 CRtfDTD::DidBuildModel(PRInt32 anErrorCode){
+  PRInt32 result=0;
+
+  return result;
+}
+
+/**
+ * 
+ *  
+ *  @update  gess 3/25/98
+ *  @param   
+ *  @return 
+ */
+void CRtfDTD::SetParser(nsIParser* aParser) {
+  mParser=(nsParser*)aParser;
+}
+
+/**
+ *  This method gets called in order to set the content
+ *  sink for this parser to dump nodes to.
+ *  
+ *  @update  gess 3/25/98
+ *  @param   nsIContentSink interface for node receiver
+ *  @return  
+ */
+nsIContentSink* CRtfDTD::SetContentSink(nsIContentSink* aSink) {
+  return 0;
+}
+
+
+/*******************************************************************
+  These methods use to be hidden in the tokenizer-delegate. 
+  That file merged with the DTD, since the separation wasn't really
+  buying us anything.
+ *******************************************************************/
+
+/**
+ *  This gets called when we've just read a '\' char. 
+ *  It means we need to read an RTF control word.
+ *  
+ *  @update gess 3/25/98
+ *  @param  aToken ptr-ref to new token we create
+ *  @return error code -- preferably kNoError
+ */
+PRInt32 CRtfDTD::ConsumeControlWord(CToken*& aToken){
+  PRInt32 result=kNoError;
+  CRTFControlWord* cw=new CRTFControlWord("");
+  if(cw){
+    CScanner* theScanner=mParser->GetScanner();
+    cw->Consume(*theScanner);
+    aToken=cw;
+  }
+  return result;
+}
+
+/**
+ *  This gets called when we've just read a '{' or '}' group char. 
+ *  It means we have just started or ended an RTF group.
+ *  
+ *  @update gess 3/25/98
+ *  @param  aToken ptr-ref to new token we create
+ *  @return error code -- preferably kNoError
+ */
+static char* keys[] = {"}","{"};
+
+PRInt32 CRtfDTD::ConsumeGroupTag(CToken*& aToken,PRBool aStartGroup){
+  PRInt32 result=kNoError;
+  CRTFGroup* cw=new CRTFGroup(keys[PR_TRUE==aStartGroup],aStartGroup);
+  if(cw){
+    CScanner* theScanner=mParser->GetScanner();
+    cw->Consume(*theScanner);
+    aToken=cw;
+  }
+  return result;
+}
+
+/**
+ *  This gets called when we've just read plain text char.
+ *  It means we need to read RTF content.
+ *  
+ *  @update gess 3/25/98
+ *  @param  aToken ptr-ref to new token we create
+ *  @return error code -- preferably kNoError
+ */
+PRInt32 CRtfDTD::ConsumeContent(PRUnichar aChar,CToken*& aToken){
+  PRInt32 result=kNoError;
+  PRUnichar buffer[2] = {0,0};
+  buffer[0]=aChar;
+  CRTFContent * cw=new CRTFContent(buffer);
+  if(cw){
+    CScanner* theScanner=mParser->GetScanner();
+    cw->Consume(*theScanner);
+    aToken=cw;
+  }
+  return result;
+}
+
+/**
+ *  This method repeatedly called by the tokenizer. 
+ *  Each time, we determine the kind of token were about to 
+ *  read, and then we call the appropriate method to handle
+ *  that token type.
+ *  
+ *  @update gess 3/25/98
+ *  @param  aChar: last char read
+ *  @param  aScanner: see nsScanner.h
+ *  @param  anErrorCode: arg that will hold error condition
+ *  @return new token or null 
+ */
+PRInt32 CRtfDTD::ConsumeToken(CToken*& aToken){
+  
+  CScanner* theScanner=mParser->GetScanner();
+
+  PRUnichar aChar;
+  PRInt32   result=theScanner->GetChar(aChar);
+
+  switch(result) {
+    case kEOF:
+      break;
+
+    case kInterrupted:
+      theScanner->RewindToMark();
+      break; 
+
+    case kNoError:
+    default:
+      switch(aChar) {
+        case kLeftBrace:
+          result=ConsumeGroupTag(aToken,PR_TRUE); break;
+
+        case kRightBrace:
+          result=ConsumeGroupTag(aToken,PR_FALSE); break;
+
+        case kBackSlash:
+          result=ConsumeControlWord(aToken); break;
+
+        case kCR:
+        case kSpace:
+        case kTab:
+        case kLF:
+          break;
+
+        default:
+          result=ConsumeContent(aChar,aToken); break;
+      } //switch
+      break; 
+  } //switch
+  if(kNoError==result)
+    result=theScanner->Eof();
+  return result;
+}
+
+/**
+ * 
+ * @update	gess5/18/98
+ * @param 
+ * @return
+ */
+void CRtfDTD::WillResumeParse(void){
+  return;
+}
+
+/**
+ * 
+ * @update	gess5/18/98
+ * @param 
+ * @return
+ */
+void CRtfDTD::WillInterruptParse(void){
+  return;
+}
+
+/**
+ * 
+ * @update	jevering6/23/98
+ * @param 
+ * @return
+ */
+void CRtfDTD::SetDTDDebug(nsIDTDDebug * aDTDDebug) {
+}
+
+/**
+ *  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 -- int tag of parent container
+ *  @param   aChild -- int tag of child container
+ *  @return  PR_TRUE if parent can contain child
+ */
+PRBool CRtfDTD::CanContain(PRInt32 aParent,PRInt32 aChild){
+  PRBool result=PR_FALSE;
+  return result;
+}
+
+/**
+ *  
+ *  @update  gess 3/25/98
+ *  @param   aToken -- token object to be put into content model
+ *  @return  0 if all is well; non-zero is an error
+ */
+PRInt32 CRtfDTD::HandleGroup(CToken* aToken){
+  PRInt32 result=kNoError;
+  return result;
+}
+
+/**
+ *  
+ *  @update  gess 3/25/98
+ *  @param   aToken -- token object to be put into content model
+ *  @return  0 if all is well; non-zero is an error
+ */
+PRInt32 CRtfDTD::HandleControlWord(CToken* aToken){
+  PRInt32 result=kNoError;
+  return result;
+}
+
+/**
+ *  
+ *  @update  gess 3/25/98
+ *  @param   aToken -- token object to be put into content model
+ *  @return  0 if all is well; non-zero is an error
+ */
+PRInt32 CRtfDTD::HandleContent(CToken* aToken){
+  PRInt32 result=kNoError;
+  return result;
+}
+
+/**
+ *  
+ *  @update  gess 3/25/98
+ *  @param   aToken -- token object to be put into content model
+ *  @return  0 if all is well; non-zero is an error
+ */
+PRInt32 CRtfDTD::HandleToken(CToken* aToken) {
+  PRInt32 result=0;
+
+  if(aToken) {
+    eRTFTokenTypes theType=eRTFTokenTypes(aToken->GetTokenType());
+    
+    switch(theType) {
+      case eRTFToken_group:
+        result=HandleGroup(aToken); break;
+
+      case eRTFToken_controlword:
+        result=HandleControlWord(aToken); break;
+
+      case eRTFToken_content:
+        result=HandleContent(aToken); break;
+
+      default:
+        break;
+      //everything else is just text or attributes of controls...
+    }
+
+  }//if
+
+  return result;
+}
+
+
+
+/***************************************************************
+  Heres's the RTFControlWord subclass...
+ ***************************************************************/
+
+CRTFControlWord::CRTFControlWord(char* aKey) : CToken(aKey), mArgument("") {
+}
+
+
+PRInt32 CRTFControlWord::GetTokenType() {
+  return eRTFToken_controlword;
+}
+
+
+PRInt32 CRTFControlWord::Consume(CScanner& aScanner){
+  PRInt32 result=aScanner.ReadWhile(mTextValue,gAlphaChars,PR_FALSE);
+  if(kNoError==result) {
+    //ok, now look for an option parameter...
+    PRUnichar ch;
+    result=aScanner.Peek(ch);
+
+    switch(ch) {
+      case '0': case '1': case '2': case '3': case '4': 
+      case '5': case '6': case '7': case '8': case '9': 
+      case kMinus:
+        result=aScanner.ReadWhile(mArgument,gDigits,PR_FALSE);
+        break;
+
+      case kSpace:
+      default:
+        break;
+    }
+  }
+  if(kNoError==result)
+    result=aScanner.SkipWhitespace();
+  return result;
+}
+
+
+/***************************************************************
+  Heres's the RTFGroup subclass...
+ ***************************************************************/
+
+CRTFGroup::CRTFGroup(char* aKey,PRBool aStartGroup) : CToken(aKey) {
+  mStart=aStartGroup;
+}
+
+
+PRInt32 CRTFGroup::GetTokenType() {
+  return eRTFToken_group;
+}
+
+void CRTFGroup::SetGroupStart(PRBool aFlag){
+  mStart=aFlag;
+}
+
+PRBool CRTFGroup::IsGroupStart(){
+  return mStart;
+}
+
+PRInt32 CRTFGroup::Consume(CScanner& aScanner){
+  PRInt32 result=kNoError;
+  if(PR_FALSE==mStart)
+    result=aScanner.SkipWhitespace();
+  return result;
+}
+
+
+/***************************************************************
+  Heres's the RTFContent subclass...
+ ***************************************************************/
+
+CRTFContent::CRTFContent(PRUnichar* aKey) : CToken(aKey) {
+}
+
+
+PRInt32 CRTFContent::GetTokenType() {
+  return eRTFToken_content;
+}
+
+
+
+/**
+ * We're supposed to read text until we encounter one
+ * of the RTF control characters: \.{,}.
+ * @update	gess7/9/98
+ * @param 
+ * @return
+ */
+static nsString textTerminators("\\{}");
+PRInt32 CRTFContent::Consume(CScanner& aScanner){
+  PRInt32 result=aScanner.ReadUntil(mTextValue,textTerminators,PR_FALSE);
+  return result;
+}
+
+
+

mozilla/htmlparser/src/CRtfDTD.h

@@ -0,0 +1,339 @@
+
+/* -*- Mode: C++; tab-width: 2; indent-tabs-mode: nil; c-basic-offset: 2 -*- */
+/*
+ * The contents of this file are subject to the Netscape Public License
+ * Version 1.0 (the "NPL"); you may not use this file except in
+ * compliance with the NPL.  You may obtain a copy of the NPL at
+ * http://www.mozilla.org/NPL/
+ *
+ * Software distributed under the NPL is distributed on an "AS IS" basis,
+ * WITHOUT WARRANTY OF ANY KIND, either express or implied. See the NPL
+ * for the specific language governing rights and limitations under the
+ * NPL.
+ *
+ * The Initial Developer of this code under the NPL is Netscape
+ * Communications Corporation.  Portions created by Netscape are
+ * Copyright (C) 1998 Netscape Communications Corporation.  All Rights
+ * Reserved.
+ */
+
+/**
+ * MODULE NOTES:
+ * @update  gess 4/8/98
+ * 
+ *         
+ */
+
+#ifndef NS_NAVHTMLDTD__
+#define NS_NAVHTMLDTD__
+
+#include "nsIDTD.h"
+#include "nsISupports.h"
+#include "nsHTMLTokens.h"
+#include "nshtmlpars.h"
+#include "nsVoidArray.h"
+#include "nsDeque.h"
+
+#define NS_RTF_DTD_IID      \
+  {0xa39c6bfc, 0x15f0,  0x11d2,  \
+  {0x80, 0x41, 0x00,    0x10, 0x4b, 0x98, 0x3f, 0xd4}}
+
+class nsIParserNode;
+class nsParser;
+
+
+enum eRTFTokenTypes {
+  eRTFToken_unknown=0,
+  eRTFToken_group,        // '{'
+  eRTFToken_controlword,  // '\'
+  eRTFToken_content,      // contains all document content
+    //everything else is just text or attributes of controls...
+  eRTFToken_last //make sure this stays the last token...
+};
+
+
+enum eRTFTags {
+  eRTFCtrl_unknown=0,
+  eRTFCtrl_quote,
+  eRTFCtrl_star,   
+  eRTFCtrl_linefeed,
+  eRTFCtrl_return,
+  eRTFCtrl_begincontrol,
+  eRTFCtrl_bold, 
+  eRTFCtrl_bin,
+  eRTFCtrl_blue,
+  eRTFCtrl_cols,
+  eRTFCtrl_comment,
+  eRTFCtrl_italic,
+  eRTFCtrl_font,
+  eRTFCtrl_fonttable,
+  eRTFCtrl_green,
+  eRTFCtrl_bottommargin,
+  eRTFCtrl_leftmargin,
+  eRTFCtrl_rightmargin,
+  eRTFCtrl_topmargin,
+  eRTFCtrl_par,
+  eRTFCtrl_pard,
+  eRTFCtrl_plain,
+  eRTFCtrl_justified,
+  eRTFCtrl_fulljustified,
+  eRTFCtrl_leftjustified,
+  eRTFCtrl_rightjustified,
+  eRTFCtrl_rdoublequote,
+  eRTFCtrl_red,
+  eRTFCtrl_rtf,
+  eRTFCtrl_tab,
+  eRTFCtrl_title,
+  eRTFCtrl_underline,
+  eRTFCtrl_startgroup,
+  eRTFCtrl_endgroup,
+  eRTFCtrl_last //make sure this stays the last token...
+};
+
+
+
+/**
+ * 
+ * @update	gess7/8/98
+ * @param 
+ * @return
+ */
+class CRTFControlWord : public CToken {
+public:
+                  CRTFControlWord(char* aKey);
+  virtual PRInt32 GetTokenType();
+  virtual PRInt32 Consume(CScanner& aScanner);
+protected:
+  nsString  mArgument;
+};
+
+
+/**
+ * 
+ * @update	gess7/8/98
+ * @param 
+ * @return
+ */
+class CRTFGroup: public CToken {
+public:
+                  CRTFGroup(char* aKey,PRBool aStartGroup);
+  virtual PRInt32 GetTokenType();
+  virtual void    SetGroupStart(PRBool aFlag);
+  virtual PRBool  IsGroupStart();
+  virtual PRInt32 Consume(CScanner& aScanner);
+protected:
+          PRBool  mStart;
+};
+
+
+/**
+ * 
+ * @update	gess7/8/98
+ * @param 
+ * @return
+ */
+class CRTFContent: public CToken {
+public:
+                  CRTFContent(PRUnichar* aValue);
+  virtual PRInt32 GetTokenType();
+  virtual PRInt32 Consume(CScanner& aScanner);
+};
+
+
+class CRtfDTD : public nsIDTD {
+            
+  public:
+
+    NS_DECL_ISUPPORTS
+
+
+    /**
+     *  
+     *  
+     *  @update  gess 4/9/98
+     *  @param   
+     *  @return  
+     */
+    CRtfDTD();
+
+    /**
+     *  
+     *  
+     *  @update  gess 4/9/98
+     *  @param   
+     *  @return  
+     */
+    virtual ~CRtfDTD();
+
+    /**
+     * This method is called to determine if the given DTD can parse
+     * a document in a given source-type. 
+     * NOTE: Parsing always assumes that the end result will involve
+     *       storing the result in the main content model.
+     * @update	gess6/24/98
+     * @param   
+     * @return  TRUE if this DTD can satisfy the request; FALSE otherwise.
+     */
+    virtual PRBool CanParse(nsString& aContentType, PRInt32 aVersion);
+
+    /**
+     * 
+     * @update	gess7/7/98
+     * @param 
+     * @return
+     */
+    virtual eAutoDetectResult AutoDetectContentType(nsString& aBuffer,nsString& aType);
+
+    /**
+     * 
+     * @update	gess5/18/98
+     * @param 
+     * @return
+     */
+    virtual PRInt32 WillBuildModel(const char* aFilename=0);
+
+    /**
+     * 
+     * @update	gess5/18/98
+     * @param 
+     * @return
+     */
+    virtual PRInt32 DidBuildModel(PRInt32 anErrorCode);
+
+    /**
+     *  
+     *  @update  gess 3/25/98
+     *  @param   aToken -- token object to be put into content model
+     *  @return  0 if all is well; non-zero is an error
+     */
+    virtual PRInt32 HandleGroup(CToken* aToken);
+
+    /**
+     *  
+     *  @update  gess 3/25/98
+     *  @param   aToken -- token object to be put into content model
+     *  @return  0 if all is well; non-zero is an error
+     */
+    virtual PRInt32 HandleControlWord(CToken* aToken);
+
+    /**
+     *  
+     *  @update  gess 3/25/98
+     *  @param   aToken -- token object to be put into content model
+     *  @return  0 if all is well; non-zero is an error
+     */
+    virtual PRInt32 HandleContent(CToken* aToken);
+
+    /**
+     *  
+     *  @update  gess 3/25/98
+     *  @param   aToken -- token object to be put into content model
+     *  @return  0 if all is well; non-zero is an error
+     */
+    virtual PRInt32 HandleToken(CToken* aToken);
+
+    /**
+     * 
+     *  
+     *  @update  gess 3/25/98
+     *  @param   
+     *  @return 
+     */
+    virtual void SetParser(nsIParser* aParser);
+
+    /**
+     *  Cause the tokenizer to consume the next token, and 
+     *  return an error result.
+     *  
+     *  @update  gess 3/25/98
+     *  @param   anError -- ref to error code
+     *  @return  new token or null
+     */
+    virtual PRInt32 ConsumeControlWord(CToken*& aToken);
+
+    /**
+     *  Cause the tokenizer to consume the next token, and 
+     *  return an error result.
+     *  
+     *  @update  gess 3/25/98
+     *  @param   anError -- ref to error code
+     *  @return  new token or null
+     */
+    virtual PRInt32 ConsumeGroupTag(CToken*& aToken,PRBool aStartTag);
+
+    /**
+     *  Cause the tokenizer to consume the next token, and 
+     *  return an error result.
+     *  
+     *  @update  gess 3/25/98
+     *  @param   anError -- ref to error code
+     *  @return  new token or null
+     */
+    virtual PRInt32 ConsumeContent(PRUnichar aChar,CToken*& aToken);
+
+    /**
+     *  Cause the tokenizer to consume the next token, and 
+     *  return an error result.
+     *  
+     *  @update  gess 3/25/98
+     *  @param   anError -- ref to error code
+     *  @return  new token or null
+     */
+    virtual PRInt32 ConsumeToken(CToken*& aToken);
+
+
+    /**
+     * 
+     * @update	gess5/18/98
+     * @param 
+     * @return
+     */
+    virtual void WillResumeParse(void);
+
+    /**
+     * 
+     * @update	gess5/18/98
+     * @param 
+     * @return
+     */
+    virtual void WillInterruptParse(void);
+
+   /**
+     * Select given content sink into parser for parser output
+     * @update	gess5/11/98
+     * @param   aSink is the new sink to be used by parser
+     * @return  old sink, or NULL
+     */
+    virtual nsIContentSink* SetContentSink(nsIContentSink* aSink);
+
+    /**
+     * 
+     * @update	jevering6/23/98
+     * @param 
+     * @return
+     */
+	  virtual void SetDTDDebug(nsIDTDDebug * aDTDDebug);
+
+    /**
+     *  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 -- int tag of parent container
+     *  @param   aChild -- int tag of child container
+     *  @return  PR_TRUE if parent can contain child
+     */
+    virtual PRBool CanContain(PRInt32 aParent,PRInt32 aChild);
+	
+protected:
+    
+    nsParser*           mParser;
+    char*               mFilename;
+};
+
+extern NS_HTMLPARS nsresult NS_NewRTF_DTD(nsIDTD** aInstancePtrResult);
+
+#endif 
+
+
+

mozilla/htmlparser/src/MANIFEST

@@ -5,6 +5,7 @@
 nshtmlpars.h
 nsIContentSink.h
 nsIHTMLContentSink.h
+nsHTMLContentSinkStream.h
 nsHTMLTokens.h
 nsIParserNode.h
 nsIParser.h
@@ -12,3 +13,4 @@ nsToken.h
 nsIDTDDebug.h
 nsIDTD.h
 CNavDTD.h
+CRTFDTD.h

mozilla/htmlparser/src/Makefile

@@ -30,16 +30,21 @@ CPPSRCS =     \
   nsTokenHandler.cpp \
   CNavDTD.cpp \
   COtherDTD.cpp \
+  CRtfDTD.cpp \
+  nsValidDTD.cpp \
+  nsWellFormedDTD.cpp \
   nsParser.cpp  \
   nsHTMLTokens.cpp  \
   nsHTMLTags.cpp \
   prstrm.cpp \
+  nsHTMLContentSinkStream.cpp \
   $(NULL)
 
 EXPORTS =     \
   nshtmlpars.h    \
   nsIContentSink.h  \
   nsIHTMLContentSink.h  \
+  nsHTMLContentSinkStream.h \
   nsHTMLTokens.h    \
   nsIParserNode.h   \
   nsIParser.h   \
@@ -48,6 +53,9 @@ EXPORTS =     \
   nsToken.h   \
   CNavDTD.h \
   COtherDTD.h \
+  nsWellFormedDTD.h \
+  nsValidDTD.h \
+  CRTFDTD.h \
   $(NULL)
 
 EXTRA_EXPORTS =					\

mozilla/htmlparser/src/makefile.win

@@ -26,6 +26,9 @@ REQUIRES=xpcom raptor netlib
 CPPSRCS=					\
 	CNavDTD.obj				\
 	COtherDTD.obj				\
+	CRTFDTD.obj				\
+	nsWellFormedDTD.obj	 \
+	nsValidDTD.obj	 \
 	nsDTDDebug.obj				\
 	nsHTMLContentSink.obj			\
 	nsHTMLTags.obj				\
@@ -35,12 +38,16 @@ CPPSRCS=					\
 	nsScanner.obj				\
 	nsToken.obj				\
 	nsTokenHandler.obj			\
+	nsHTMLContentSinkStream.obj			\
 	prstrm.obj				\
 	$(NULL)
 
 CPP_OBJS=					\
 	.\$(OBJDIR)\CNavDTD.obj			\
+	.\$(OBJDIR)\CRTFDTD.obj			\
 	.\$(OBJDIR)\COtherDTD.obj		\
+	.\$(OBJDIR)\nsWellFormedDTD.obj		\
+	.\$(OBJDIR)\nsValidDTD.obj		\
 	.\$(OBJDIR)\nsDTDDebug.obj		\
 	.\$(OBJDIR)\nsHTMLContentSink.obj	\
 	.\$(OBJDIR)\nsHTMLTags.obj		\
@@ -50,6 +57,7 @@ CPP_OBJS=					\
 	.\$(OBJDIR)\nsScanner.obj		\
 	.\$(OBJDIR)\nsToken.obj			\
 	.\$(OBJDIR)\nsTokenHandler.obj		\
+	.\$(OBJDIR)\nsHTMLContentSinkStream.obj		\
 	.\$(OBJDIR)\prstrm.obj			\
 	$(NULL)
 
@@ -57,6 +65,7 @@ EXPORTS=					\
 	nshtmlpars.h				\
 	nsIContentSink.h			\
 	nsIHTMLContentSink.h			\
+	nsHTMLContentSinkStream.h			\
 	nsHTMLTokens.h				\
 	nsIParserNode.h				\
 	nsIParser.h				\
@@ -64,6 +73,9 @@ EXPORTS=					\
 	nsIDTDDebug.h				\
 	nsIParserFilter.h			\
 	CNavDTD.h				\
+	nsWellFormedDTD.h	\
+	nsValidDTD.h	\
+	CRTFDTD.h				\
 	COtherDTD.h				\
 	nsIDTD.h				\
 	$(NULL)

mozilla/htmlparser/src/nsHTMLContentSinkStream.cpp

@@ -0,0 +1,646 @@
+/* -*- Mode: C++; tab-width: 2; indent-tabs-mode: nil; c-basic-offset: 2 -*- */
+/*
+ * The contents of this file are subject to the Netscape Public License
+ * Version 1.0 (the "NPL"); you may not use this file except in
+ * compliance with the NPL.  You may obtain a copy of the NPL at
+ * http://www.mozilla.org/NPL/
+ *
+ * Software distributed under the NPL is distributed on an "AS IS" basis,
+ * WITHOUT WARRANTY OF ANY KIND, either express or implied. See the NPL
+ * for the specific language governing rights and limitations under the
+ * NPL.
+ *
+ * The Initial Developer of this code under the NPL is Netscape
+ * Communications Corporation.  Portions created by Netscape are
+ * Copyright (C) 1998 Netscape Communications Corporation.  All Rights
+ * Reserved.
+ */
+
+/**
+ * MODULE NOTES:
+ * @update  gess 4/1/98
+ * 
+ * This file declares the concrete HTMLContentSink class.
+ * This class is used during the parsing process as the
+ * primary interface between the parser and the content
+ * model.
+ */
+
+
+#include "nsHTMLContentSinkStream.h"
+#include "nsHTMLTokens.h"
+#include <iostream.h>
+#include "nsstring.h"
+#include "nsParserTypes.h"
+
+static NS_DEFINE_IID(kISupportsIID,   NS_ISUPPORTS_IID);                 
+static NS_DEFINE_IID(kIContentSinkIID,NS_IHTMLCONTENTSINK_IID);
+static NS_DEFINE_IID(kClassIID,       NS_HTMLCONTENTSINK_STREAM_IID); 
+
+
+static char*    gHeaderComment = "<!-- This page was created by the NGLayout output system. -->";
+static char*    gDocTypeHeader = "<!DOCTYPE HTML PUBLIC \"-//W3C//DTD HTML 3.2//EN\">";
+const  int      gTabSize=2;
+static char     gBuffer[500];
+
+/**
+ *  This method gets called as part of our COM-like interfaces.
+ *  Its purpose is to create an interface to parser object
+ *  of some type.
+ *  
+ *  @update   gess 4/8/98
+ *  @param    nsIID  id of object to discover
+ *  @param    aInstancePtr ptr to newly discovered interface
+ *  @return   NS_xxx result code
+ */
+nsresult CHTMLContentSinkStream::QueryInterface(const nsIID& aIID, void** aInstancePtr)  
+{                                                                        
+  if (NULL == aInstancePtr) {                                            
+    return NS_ERROR_NULL_POINTER;                                        
+  }                                                                      
+
+  if(aIID.Equals(kISupportsIID))    {  //do IUnknown...
+    *aInstancePtr = (nsIContentSink*)(this);                                        
+  }
+  else if(aIID.Equals(kIContentSinkIID)) {  //do IParser base class...
+    *aInstancePtr = (nsIContentSink*)(this);                                        
+  }
+  else if(aIID.Equals(kClassIID)) {  //do this class...
+    *aInstancePtr = (CHTMLContentSinkStream*)(this);                                        
+  }                 
+  else {
+    *aInstancePtr=0;
+    return NS_NOINTERFACE;
+  }
+  ((nsISupports*) *aInstancePtr)->AddRef();
+  return NS_OK;                                                        
+}
+
+
+NS_IMPL_ADDREF(CHTMLContentSinkStream)
+NS_IMPL_RELEASE(CHTMLContentSinkStream)
+
+
+/**
+ *  This method is defined in nsIParser. It is used to 
+ *  cause the COM-like construction of an nsParser.
+ *  
+ *  @update  gess 4/8/98
+ *  @param   nsIParser** ptr to newly instantiated parser
+ *  @return  NS_xxx error result
+ */
+NS_HTMLPARS nsresult NS_New_HTML_ContentSinkStream(CHTMLContentSinkStream** aInstancePtrResult) {
+  CHTMLContentSinkStream* it = new CHTMLContentSinkStream();
+
+  if (it == 0) {
+    return NS_ERROR_OUT_OF_MEMORY;
+  }
+
+  return it->QueryInterface(kClassIID, (void **) aInstancePtrResult);
+}
+
+
+/**
+ * Construct a content sink stream.
+ * @update	gess7/7/98
+ * @param 
+ * @return
+ */
+CHTMLContentSinkStream::CHTMLContentSinkStream()  {
+  mOutput=&cout;
+}
+
+/**
+ * Construct a content sink stream.
+ * @update	gess7/7/98
+ * @param 
+ * @return
+ */
+CHTMLContentSinkStream::CHTMLContentSinkStream(ostream& aStream)  {
+  mOutput=&aStream;
+}
+
+
+/**
+ * 
+ * @update	gess7/7/98
+ * @param 
+ * @return
+ */
+CHTMLContentSinkStream::~CHTMLContentSinkStream() {
+  mOutput=0;  //we don't own the stream we're given; just forget it.
+}
+
+
+/**
+ * 
+ * @update	gess7/7/98
+ * @param 
+ * @return
+ */
+void CHTMLContentSinkStream::SetOutputStream(ostream& aStream) {
+  mOutput=&aStream;
+}
+
+
+/**
+ * 
+ * @update	gess7/7/98
+ * @param 
+ * @return
+ */
+void WriteAttributes(const nsIParserNode& aNode,ostream& aStream) {
+  int theCount=aNode.GetAttributeCount();
+  if(theCount) {
+    int i=0;
+    for(i=0;i<theCount;i++){
+      const nsString& key=aNode.GetKeyAt(i);
+      key.ToCString(gBuffer,sizeof(gBuffer)-1);
+      aStream << " " << gBuffer << char(kEqual);
+      const nsString& value=aNode.GetValueAt(i);
+      value.ToCString(gBuffer,sizeof(gBuffer)-1);
+      aStream << gBuffer;
+    }
+  }
+}
+
+
+/**
+ * 
+ * @update	gess7/5/98
+ * @param 
+ * @return
+ */
+void OpenTagWithAttributes(const char* theTag,const nsIParserNode& aNode,int tab,ostream& aStream,bool aNewline) {
+  int i=0;
+  for(i=0;i<tab*gTabSize;i++) 
+    aStream << " ";
+  aStream << (char)kLessThan << theTag;
+  WriteAttributes(aNode,aStream);
+  aStream << (char)kGreaterThan;
+  if(aNewline)
+    aStream << endl;
+}
+
+
+/**
+ * 
+ * @update	gess7/5/98
+ * @param 
+ * @return
+ */
+void OpenTag(const char* theTag,int tab,ostream& aStream,bool aNewline) {
+  int i=0;
+  for(i=0;i<tab*gTabSize;i++) 
+    aStream << " ";
+  aStream << (char)kLessThan << theTag << (char)kGreaterThan;
+  if(aNewline)
+    aStream << endl;
+}
+
+
+/**
+ * 
+ * @update	gess7/5/98
+ * @param 
+ * @return
+ */
+void CloseTag(const char* theTag,int tab,ostream& aStream) {
+  int i=0;
+  for(i=0;i<tab*gTabSize;i++) 
+    aStream << " ";
+  aStream << (char)kLessThan << (char)kForwardSlash << theTag << (char)kGreaterThan << endl;
+}
+
+
+/**
+ * 
+ * @update	gess7/5/98
+ * @param 
+ * @return
+ */
+void WritePair(eHTMLTags aTag,const nsString& theContent,int tab,ostream& aStream) {
+  const char* titleStr = GetTagName(aTag);
+  OpenTag(titleStr,tab,aStream,false);
+  theContent.ToCString(gBuffer,sizeof(gBuffer)-1);
+  aStream << gBuffer;
+  CloseTag(titleStr,0,aStream);
+}
+
+
+
+/**
+ * 
+ * @update	gess7/5/98
+ * @param 
+ * @return
+ */
+void WriteSingleton(eHTMLTags aTag,const nsString& theContent,int tab,ostream& aStream) {
+  const char* titleStr = GetTagName(aTag);
+  OpenTag(titleStr,tab,aStream,false);
+  if(theContent.Length()) {
+    theContent.ToCString(gBuffer,sizeof(gBuffer)-1);
+    aStream << gBuffer;
+  }
+  aStream << endl;
+}
+
+
+/**
+  * This method gets called by the parser when it encounters
+  * a title tag and wants to set the document title in the sink.
+  *
+  * @update 4/1/98 gess
+  * @param  nsString reference to new title value
+  * @return PR_TRUE if successful. 
+  */     
+PRBool CHTMLContentSinkStream::SetTitle(const nsString& aValue){
+  PRBool result=PR_FALSE;
+  if(mOutput) {
+    WritePair(eHTMLTag_title,aValue,1+mTabLevel,*mOutput);
+  }
+  return result;
+}
+
+
+/**
+  * This method is used to open the outer HTML container.
+  *
+  * @update 4/1/98 gess
+  * @param  nsIParserNode reference to parser node interface
+  * @return PR_TRUE if successful. 
+  */     
+PRInt32 CHTMLContentSinkStream::OpenHTML(const nsIParserNode& aNode){
+  PRInt32 result=0;
+  mTabLevel++;
+  if(mOutput) {
+    const char* theTag= GetTagName(eHTMLTag_html);
+    OpenTagWithAttributes(theTag,aNode,mTabLevel,*mOutput,true);
+  }
+  return result;
+}
+
+
+/**
+  * This method is used to close the outer HTML container.
+  *
+  * @update 4/1/98 gess
+  * @param  nsIParserNode reference to parser node interface
+  * @return PR_TRUE if successful. 
+  */     
+PRInt32 CHTMLContentSinkStream::CloseHTML(const nsIParserNode& aNode){
+  PRInt32 result=0;
+  if(mOutput) {
+    const char* theTag= GetTagName(eHTMLTag_html);
+    CloseTag(theTag,0,*mOutput);
+  }
+  mTabLevel--;
+  return result;
+}
+
+
+/**
+  * This method is used to open the only HEAD container.
+  *
+  * @update 4/1/98 gess
+  * @param  nsIParserNode reference to parser node interface
+  * @return PR_TRUE if successful. 
+  */     
+PRInt32 CHTMLContentSinkStream::OpenHead(const nsIParserNode& aNode){
+  PRInt32 result=0;
+  mTabLevel++;
+  if(mOutput) {
+    const char* theTag= GetTagName(eHTMLTag_head);
+    OpenTagWithAttributes(theTag,aNode,mTabLevel,*mOutput,true);
+  }
+  return result;
+}
+
+
+/**
+  * This method is used to close the only HEAD container.
+  *
+  * @update 4/1/98 gess
+  * @param  nsIParserNode reference to parser node interface
+  * @return PR_TRUE if successful. 
+  */     
+PRInt32 CHTMLContentSinkStream::CloseHead(const nsIParserNode& aNode){
+  PRInt32 result=0;
+  if(mOutput) {
+    const char* theTag= GetTagName(eHTMLTag_head);
+    CloseTag(theTag,mTabLevel,*mOutput);
+  }
+  mTabLevel--;
+  return result;
+}
+
+
+/**
+  * This method is used to open the main BODY container.
+  *
+  * @update 4/1/98 gess
+  * @param  nsIParserNode reference to parser node interface
+  * @return PR_TRUE if successful. 
+  */     
+PRInt32 CHTMLContentSinkStream::OpenBody(const nsIParserNode& aNode){
+  PRInt32 result=0;
+  mTabLevel++;
+  if(mOutput) {
+    const char* theTag= GetTagName(eHTMLTag_body);
+    OpenTagWithAttributes(theTag,aNode,mTabLevel,*mOutput,true);
+  }
+  return result;
+}
+
+
+/**
+  * This method is used to close the main BODY container.
+  *
+  * @update 4/1/98 gess
+  * @param  nsIParserNode reference to parser node interface
+  * @return PR_TRUE if successful. 
+  */     
+PRInt32 CHTMLContentSinkStream::CloseBody(const nsIParserNode& aNode){
+  PRInt32 result=0;
+  if(mOutput) {
+    const char* theTag= GetTagName(eHTMLTag_body);
+    CloseTag(theTag,mTabLevel,*mOutput);
+  }
+  mTabLevel--;
+  return result;
+}
+
+
+/**
+  * This method is used to open a new FORM container.
+  *
+  * @update 4/1/98 gess
+  * @param  nsIParserNode reference to parser node interface
+  * @return PR_TRUE if successful. 
+  */     
+PRInt32 CHTMLContentSinkStream::OpenForm(const nsIParserNode& aNode){
+  PRInt32 result=0;
+  mTabLevel++;
+  if(mOutput) {
+    const char* theTag= GetTagName(eHTMLTag_form);
+    OpenTagWithAttributes(theTag,aNode,mTabLevel,*mOutput,true);
+  }
+  return result;
+}
+
+
+/**
+  * This method is used to close the outer FORM container.
+  *
+  * @update 4/1/98 gess
+  * @param  nsIParserNode reference to parser node interface
+  * @return PR_TRUE if successful. 
+  */     
+PRInt32 CHTMLContentSinkStream::CloseForm(const nsIParserNode& aNode){
+  PRInt32 result=0;
+  if(mOutput) {
+    const char* theTag= GetTagName(eHTMLTag_form);
+    CloseTag(theTag,mTabLevel,*mOutput);
+  }
+  mTabLevel--;
+  return result;
+}
+
+    
+/**
+  * This method is used to open the FRAMESET container.
+  *
+  * @update 4/1/98 gess
+  * @param  nsIParserNode reference to parser node interface
+  * @return PR_TRUE if successful. 
+  */     
+PRInt32 CHTMLContentSinkStream::OpenFrameset(const nsIParserNode& aNode){
+  PRInt32 result=0;
+  mTabLevel++;
+  if(mOutput) {
+    const char* theTag= GetTagName(eHTMLTag_frameset);
+    OpenTagWithAttributes(theTag,aNode,mTabLevel,*mOutput,true);
+  }
+  return result;
+}
+
+
+/**
+  * This method is used to close the FRAMESET container.
+  *
+  * @update 4/1/98 gess
+  * @param  nsIParserNode reference to parser node interface
+  * @return PR_TRUE if successful. 
+  */     
+PRInt32 CHTMLContentSinkStream::CloseFrameset(const nsIParserNode& aNode){
+  PRInt32 result=0;
+  if(mOutput) {
+    const char* theTag= GetTagName(eHTMLTag_frameset);
+    CloseTag(theTag,mTabLevel,*mOutput);
+  }
+  mTabLevel--;
+  return result;
+}
+
+    
+/**
+  * This method is used to a general container. 
+  * This includes: OL,UL,DIR,SPAN,TABLE,H[1..6],etc.
+  *
+  * @update 4/1/98 gess
+  * @param  nsIParserNode reference to parser node interface
+  * @return PR_TRUE if successful. 
+  */     
+PRInt32 CHTMLContentSinkStream::OpenContainer(const nsIParserNode& aNode){
+  PRInt32 result=0;
+  mTabLevel++;
+  if(mOutput) {
+
+    eHTMLTags nodeType=(eHTMLTags)aNode.GetNodeType();
+
+    switch(nodeType) {
+
+      case eHTMLTag_html:
+        result=OpenHTML(aNode); break;
+
+      case eHTMLTag_body:
+        result=OpenBody(aNode); break;
+
+      case eHTMLTag_title:
+        {
+          const nsString& title=aNode.GetName();
+          result=SetTitle(title); 
+        }
+        break;
+
+      case eHTMLTag_head:
+        result=OpenHead(aNode); break;
+
+      case eHTMLTag_style:
+        break;
+
+      case eHTMLTag_form:
+        result=OpenForm(aNode); break;
+
+      default:
+        //this is the hard one. You have to decode aNode manually.
+        {
+          const nsString& name=aNode.GetName();
+          const char* tagName= GetTagName(nodeType);
+          OpenTagWithAttributes(tagName,aNode,mTabLevel,*mOutput,true);
+        }
+        break;
+    }
+
+
+  }
+  return result;
+}
+
+
+/**
+  * This method is used to close a generic container.
+  *
+  * @update 4/1/98 gess
+  * @param  nsIParserNode reference to parser node interface
+  * @return PR_TRUE if successful. 
+  */     
+PRInt32 CHTMLContentSinkStream::CloseContainer(const nsIParserNode& aNode){
+  PRInt32 result=0;
+  if(mOutput) {
+
+    eHTMLTags nodeType=(eHTMLTags)aNode.GetNodeType();
+  
+    switch(nodeType) {
+
+      case eHTMLTag_html:
+        result=CloseHTML(aNode); break;
+
+      case eHTMLTag_style:
+        break;
+
+      case eHTMLTag_head:
+        result=CloseHead(aNode); break; 
+
+      case eHTMLTag_body:
+        result=CloseBody(aNode); break;
+
+      case eHTMLTag_form:
+        result=CloseForm(aNode); break;
+
+      case eHTMLTag_title:
+//        result=SetTitle(aNode); break;
+
+      default:
+        {
+          const nsString& name=aNode.GetName();
+          const char* tagName= GetTagName(nodeType);
+          //(*mOutput) << endl;
+          CloseTag(tagName,mTabLevel,*mOutput);
+        }
+        break;
+    }
+  }
+  mTabLevel--;
+  return result;
+}
+
+
+/**
+  * This method is used to add a leaf to the currently 
+  * open container.
+  *
+  * @update 4/1/98 gess
+  * @param  nsIParserNode reference to parser node interface
+  * @return PR_TRUE if successful. 
+  */     
+PRInt32 CHTMLContentSinkStream::AddLeaf(const nsIParserNode& aNode){
+  PRInt32 result=0;
+  if(mOutput) {
+
+    eHTMLTags nodeType=(eHTMLTags)aNode.GetNodeType();
+
+    switch(nodeType) {
+      case eHTMLTag_hr:
+      case eHTMLTag_br:
+        {
+          nsAutoString empty;
+          WriteSingleton(nodeType,empty,1+mTabLevel,*mOutput);
+        }
+        break;
+
+      case eHTMLTag_style:
+        {
+          const nsString& skipped=aNode.GetSkippedContent();
+          WritePair(nodeType,skipped,1+mTabLevel,*mOutput);
+        }
+        break;
+
+      case eHTMLTag_newline:
+      case eHTMLTag_whitespace:
+        break;
+
+      default:
+        {
+          int i=0;
+          for(i=0;i<(1+mTabLevel)*gTabSize;i++) 
+            (*mOutput) << " ";
+    
+          const nsString& text=aNode.GetText();
+          text.ToCString(gBuffer,sizeof(gBuffer)-1);
+          (*mOutput) << gBuffer << endl;
+        }
+    }//switch
+  }
+  return result;
+}
+
+
+/**
+  * This method gets called when the parser begins the process
+  * of building the content model via the content sink.
+  *
+  * @update 5/7/98 gess
+  */     
+void CHTMLContentSinkStream::WillBuildModel(void){
+  mTabLevel=-1;
+  if(mOutput) {
+    (*mOutput) << gHeaderComment << endl;
+    (*mOutput) << gDocTypeHeader << endl;
+  }
+}
+
+
+/**
+  * This method gets called when the parser concludes the process
+  * of building the content model via the content sink.
+  *
+  * @param  aQualityLevel describes how well formed the doc was.
+  *         0=GOOD; 1=FAIR; 2=POOR;
+  * @update 5/7/98 gess
+  */     
+void CHTMLContentSinkStream::DidBuildModel(PRInt32 aQualityLevel) {
+}
+
+
+/**
+  * This method gets called when the parser gets i/o blocked,
+  * and wants to notify the sink that it may be a while before
+  * more data is available.
+  *
+  * @update 5/7/98 gess
+  */     
+void CHTMLContentSinkStream::WillInterrupt(void) {
+}
+
+
+/**
+  * This method gets called when the parser i/o gets unblocked,
+  * and we're about to start dumping content again to the sink.
+  *
+  * @update 5/7/98 gess
+  */     
+void CHTMLContentSinkStream::WillResume(void) {
+}
+
+

mozilla/htmlparser/src/nsHTMLContentSinkStream.h

@@ -0,0 +1,259 @@
+/* -*- Mode: C++; tab-width: 2; indent-tabs-mode: nil; c-basic-offset: 2 -*- */
+/*
+ * The contents of this file are subject to the Netscape Public License
+ * Version 1.0 (the "NPL"); you may not use this file except in
+ * compliance with the NPL.  You may obtain a copy of the NPL at
+ * http://www.mozilla.org/NPL/
+ *
+ * Software distributed under the NPL is distributed on an "AS IS" basis,
+ * WITHOUT WARRANTY OF ANY KIND, either express or implied. See the NPL
+ * for the specific language governing rights and limitations under the
+ * NPL.
+ *
+ * The Initial Developer of this code under the NPL is Netscape
+ * Communications Corporation.  Portions created by Netscape are
+ * Copyright (C) 1998 Netscape Communications Corporation.  All Rights
+ * Reserved.
+ */
+
+/**
+ * MODULE NOTES:
+ * @update  gess 4/1/98
+ * 
+ * This file declares the concrete HTMLContentSink class.
+ * This class is used during the parsing process as the
+ * primary interface between the parser and the content
+ * model.
+ */
+
+#ifndef  NS_HTMLCONTENTSINK_STREAM
+#define  NS_ITMLCONTENTSINK_STREAM
+
+#include "nsIParserNode.h"
+#include "nsIHTMLContentSink.h"
+#include "nshtmlpars.h"
+
+#define NS_HTMLCONTENTSINK_STREAM_IID  \
+  {0xa39c6bff, 0x15f0, 0x11d2, \
+  {0x80, 0x41, 0x0, 0x10, 0x4b, 0x98, 0x3f, 0xd4}}
+
+
+
+class ostream;
+
+class CHTMLContentSinkStream : public nsIHTMLContentSink {
+  public:
+
+  NS_DECL_ISUPPORTS
+
+  /**
+   * 
+   * @update	gess7/7/98
+   * @param 
+   * @return
+   */
+  CHTMLContentSinkStream(); 
+
+  /**
+   * 
+   * @update	gess7/7/98
+   * @param 
+   * @return
+   */
+  CHTMLContentSinkStream(ostream& aStream); 
+
+  /**
+   * 
+   * @update	gess7/7/98
+   * @param 
+   * @return
+   */
+  virtual ~CHTMLContentSinkStream();
+
+
+  /**
+   * 
+   * @update	gess7/7/98
+   * @param 
+   * @return
+   */
+  void SetOutputStream(ostream& aStream);
+
+   /**
+    * This method gets called by the parser when it encounters
+    * a title tag and wants to set the document title in the sink.
+    *
+    * @update 4/1/98 gess
+    * @param  nsString reference to new title value
+    * @return PR_TRUE if successful. 
+    */     
+  virtual PRBool SetTitle(const nsString& aValue);
+
+   /**
+    * This method is used to open the outer HTML container.
+    *
+    * @update 4/1/98 gess
+    * @param  nsIParserNode reference to parser node interface
+    * @return PR_TRUE if successful. 
+    */     
+    virtual PRInt32 OpenHTML(const nsIParserNode& aNode);
+
+   /**
+    * This method is used to close the outer HTML container.
+    *
+    * @update 4/1/98 gess
+    * @param  nsIParserNode reference to parser node interface
+    * @return PR_TRUE if successful. 
+    */     
+    virtual PRInt32 CloseHTML(const nsIParserNode& aNode);
+
+   /**
+    * This method is used to open the only HEAD container.
+    *
+    * @update 4/1/98 gess
+    * @param  nsIParserNode reference to parser node interface
+    * @return PR_TRUE if successful. 
+    */     
+    virtual PRInt32 OpenHead(const nsIParserNode& aNode);
+
+   /**
+    * This method is used to close the only HEAD container.
+    *
+    * @update 4/1/98 gess
+    * @param  nsIParserNode reference to parser node interface
+    * @return PR_TRUE if successful. 
+    */     
+    virtual PRInt32 CloseHead(const nsIParserNode& aNode);
+  
+   /**
+    * This method is used to open the main BODY container.
+    *
+    * @update 4/1/98 gess
+    * @param  nsIParserNode reference to parser node interface
+    * @return PR_TRUE if successful. 
+    */     
+    virtual PRInt32 OpenBody(const nsIParserNode& aNode);
+
+   /**
+    * This method is used to close the main BODY container.
+    *
+    * @update 4/1/98 gess
+    * @param  nsIParserNode reference to parser node interface
+    * @return PR_TRUE if successful. 
+    */     
+    virtual PRInt32 CloseBody(const nsIParserNode& aNode);
+
+   /**
+    * This method is used to open a new FORM container.
+    *
+    * @update 4/1/98 gess
+    * @param  nsIParserNode reference to parser node interface
+    * @return PR_TRUE if successful. 
+    */     
+    virtual PRInt32 OpenForm(const nsIParserNode& aNode);
+
+   /**
+    * This method is used to close the outer FORM container.
+    *
+    * @update 4/1/98 gess
+    * @param  nsIParserNode reference to parser node interface
+    * @return PR_TRUE if successful. 
+    */     
+    virtual PRInt32 CloseForm(const nsIParserNode& aNode);
+        
+   /**
+    * This method is used to open the FRAMESET container.
+    *
+    * @update 4/1/98 gess
+    * @param  nsIParserNode reference to parser node interface
+    * @return PR_TRUE if successful. 
+    */     
+    virtual PRInt32 OpenFrameset(const nsIParserNode& aNode);
+
+   /**
+    * This method is used to close the FRAMESET container.
+    *
+    * @update 4/1/98 gess
+    * @param  nsIParserNode reference to parser node interface
+    * @return PR_TRUE if successful. 
+    */     
+    virtual PRInt32 CloseFrameset(const nsIParserNode& aNode);
+        
+   /**
+    * This method is used to a general container. 
+    * This includes: OL,UL,DIR,SPAN,TABLE,H[1..6],etc.
+    *
+    * @update 4/1/98 gess
+    * @param  nsIParserNode reference to parser node interface
+    * @return PR_TRUE if successful. 
+    */     
+    virtual PRInt32 OpenContainer(const nsIParserNode& aNode);
+    
+   /**
+    * This method is used to close a generic container.
+    *
+    * @update 4/1/98 gess
+    * @param  nsIParserNode reference to parser node interface
+    * @return PR_TRUE if successful. 
+    */     
+    virtual PRInt32 CloseContainer(const nsIParserNode& aNode);
+
+   /**
+    * This method is used to add a leaf to the currently 
+    * open container.
+    *
+    * @update 4/1/98 gess
+    * @param  nsIParserNode reference to parser node interface
+    * @return PR_TRUE if successful. 
+    */     
+    virtual PRInt32 AddLeaf(const nsIParserNode& aNode);
+
+   /**
+    * This method gets called when the parser begins the process
+    * of building the content model via the content sink.
+    *
+    * @update 5/7/98 gess
+    */     
+    virtual void WillBuildModel(void);
+
+   /**
+    * This method gets called when the parser concludes the process
+    * of building the content model via the content sink.
+    *
+    * @param  aQualityLevel describes how well formed the doc was.
+    *         0=GOOD; 1=FAIR; 2=POOR;
+    * @update 5/7/98 gess
+    */     
+    virtual void DidBuildModel(PRInt32 aQualityLevel);
+
+   /**
+    * This method gets called when the parser gets i/o blocked,
+    * and wants to notify the sink that it may be a while before
+    * more data is available.
+    *
+    * @update 5/7/98 gess
+    */     
+    virtual void WillInterrupt(void);
+
+   /**
+    * This method gets called when the parser i/o gets unblocked,
+    * and we're about to start dumping content again to the sink.
+    *
+    * @update 5/7/98 gess
+    */     
+    virtual void WillResume(void);
+
+protected:
+    ostream*  mOutput;
+    int       mTabLevel;
+
+};
+
+extern NS_HTMLPARS nsresult NS_New_HTML_ContentSinkStream(CHTMLContentSinkStream** aInstancePtrResult);
+
+
+#endif
+
+
+
+

mozilla/htmlparser/src/nsHTMLTokens.cpp

@@ -247,7 +247,7 @@ PRInt32 CStartToken::Consume(PRUnichar aChar, CScanner& aScanner) {
    //and see if the next char is ">". If so, we have a complete
    //tag without attributes.
   if(kNoError==result) {  
-    result=aScanner.SkipWhite();
+    result=aScanner.SkipWhitespace();
     if(kNoError==result) {
       result=aScanner.GetChar(aChar);
       if(kNoError==result) {
@@ -678,7 +678,7 @@ PRInt32 ConsumeAttributeValueText(PRUnichar aChar,nsString& aString,CScanner& aS
  */
 PRInt32 CAttributeToken::Consume(PRUnichar aChar, CScanner& aScanner) {
 
-  aScanner.SkipWhite();             //skip leading whitespace                                      
+  aScanner.SkipWhitespace();             //skip leading whitespace                                      
   PRInt32 result=aScanner.Peek(aChar);
   if(kNoError==result) {
     if(kQuote==aChar) {               //if you're here, handle quoted key...
@@ -703,12 +703,12 @@ PRInt32 CAttributeToken::Consume(PRUnichar aChar, CScanner& aScanner) {
     }
 
       //now it's time to Consume the (optional) value...
-    if(!(result=aScanner.SkipWhite())) { 
+    if(!(result=aScanner.SkipWhitespace())) { 
       if(!(result=aScanner.Peek(aChar))) {
         if(kEqual==aChar){
           result=aScanner.GetChar(aChar);  //skip the equal sign...
           if(kNoError==result) {
-            result=aScanner.SkipWhite();     //now skip any intervening whitespace
+            result=aScanner.SkipWhitespace();     //now skip any intervening whitespace
             if(kNoError==result) {
               result=aScanner.GetChar(aChar);  //and grab the next char.    
               if(kNoError==result) {
@@ -722,7 +722,7 @@ PRInt32 CAttributeToken::Consume(PRUnichar aChar, CScanner& aScanner) {
                 } 
               }//if
               if(kNoError==result)
-                result=aScanner.SkipWhite();     
+                result=aScanner.SkipWhitespace();     
             }//if
           }//if
         }//if

mozilla/htmlparser/src/nsIContentSink.h

@@ -60,15 +60,6 @@ class nsIContentSink : public nsISupports {
     */     
   virtual PRInt32 CloseContainer(const nsIParserNode& aNode) = 0;
 
-   /**
-    * This method gets called by the parser when a the
-    * topmost container in the content sink needs to be closed.
-    *
-    * @update 4/1/98 gess
-    * @return PR_TRUE if successful. 
-    */     
-  virtual PRInt32 CloseTopmostContainer() = 0;
-
    /**
     * This gets called by the parser when you want to add
     * a leaf node to the current container in the content

mozilla/htmlparser/src/nsIDTD.h

@@ -41,7 +41,6 @@ class nsIDTDDebug;
 class nsIURL;
 class nsString;
 
-enum eProcessType {eParsing, eConverting};
 enum eAutoDetectResult {eUnknownDetect, eValidDetect, eInvalidDetect};
 
 class nsIDTD : public nsISupports {
@@ -76,12 +75,23 @@ class nsIDTD : public nsISupports {
     virtual nsIContentSink* SetContentSink(nsIContentSink* aSink)=0;
 
     /**
-     * 
+     * This method is called to determine if the given DTD can parse
+     * a document in a given source-type. 
+     * NOTE: Parsing always assumes that the end result will involve
+     *       storing the result in the main content model.
      * @update	gess6/24/98
+     * @param   
+     * @return  TRUE if this DTD can satisfy the request; FALSE otherwise.
+     */
+    virtual PRBool CanParse(nsString& aContentType, PRInt32 aVersion)=0;
+
+    /**
+     * 
+     * @update	gess7/7/98
      * @param 
      * @return
      */
-    virtual PRBool IsCapableOf(eProcessType aProcessType, nsString& aContentType, PRInt32 aVersion)=0;
+    virtual eAutoDetectResult AutoDetectContentType(nsString& aBuffer,nsString& aType)=0;
 
     /**
      * 
@@ -136,13 +146,16 @@ class nsIDTD : public nsISupports {
     virtual void WillInterruptParse(void)=0;
 
     /**
-     * 
-     * @update	jevering 6/18/98
-     * @param  aParent  parent tag
-     * @param  aChild   child tag
-     * @return PR_TRUE if valid container
+     *  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 -- int tag of parent container
+     *  @param   aChild -- int tag of child container
+     *  @return  PR_TRUE if parent can contain child
      */
-    virtual PRBool CanContain(PRInt32 aParent, PRInt32 aChild) = 0;
+    virtual PRBool CanContain(PRInt32 aParent,PRInt32 aChild)=0;
+
 
     /**
      * 

mozilla/htmlparser/src/nsIHTMLContentSink.h

@@ -62,7 +62,7 @@
  */
 
 #ifndef  NS_IHTMLCONTENTSINK
-#define  NS_ITMLCONTENTSINK
+#define  NS_IHTMLCONTENTSINK
 
 #include "nsIParserNode.h"
 #include "nsIContentSink.h"
@@ -194,16 +194,6 @@ class nsIHTMLContentSink : public nsIContentSink {
     */     
     virtual PRInt32 CloseContainer(const nsIParserNode& aNode)=0;
 
-   /**
-    * This method is used to close the topmost container, regardless
-    * of the type.
-    *
-    * @update 4/1/98 gess
-    * @param  nsIParserNode reference to parser node interface
-    * @return PR_TRUE if successful. 
-    */     
-    virtual PRInt32 CloseTopmostContainer()=0;
-
    /**
     * This method is used to add a leaf to the currently 
     * open container.

mozilla/htmlparser/src/nsIParser.h

@@ -35,6 +35,7 @@ class nsString;
 class CToken;
 class nsIURL;
 class nsIDTDDebug;
+class fstream;
 
 /**
  *  This class defines the iparser interface. This XPCOM
@@ -49,7 +50,7 @@ class nsIParser : public nsISupports {
 
     virtual nsIContentSink* SetContentSink(nsIContentSink* aContentSink)=0;
 
-    virtual eAutoDetectResult AutoDetectContentType(nsString& aBuffer)=0;
+    virtual eAutoDetectResult AutoDetectContentType(nsString& aBuffer,nsString& aType)=0;
 
     /**
      *  Cause the tokenizer to consume the next token, and 
@@ -61,26 +62,15 @@ class nsIParser : public nsISupports {
      */
     virtual PRInt32 ConsumeToken(CToken*& aToken)=0;
 
-    virtual PRInt32 BeginParse(nsIURL* aURL,
-                               nsIStreamObserver* aListener = nsnull,
-                               nsIDTDDebug * aDTDDebug = 0) = 0;
-
     /******************************************************************************************
      *  Parse methods always begin with an input source, and perform conversions 
      *  until you wind up with HTML in your actual content model.
      ******************************************************************************************/
     virtual PRInt32 Parse(nsIURL* aURL,nsIStreamObserver* aListener = nsnull,nsIDTDDebug * aDTDDebug = 0) = 0;
     virtual PRInt32 Parse(const char* aFilename)=0;
+    virtual PRInt32 Parse(fstream& aStream)=0;
     virtual PRInt32 Parse(nsString& anHTMLString,PRBool appendTokens)=0;
 
-    /******************************************************************************************
-     *  Convert methods start input source (of known or unknown form), and perform conversions 
-     *  until you wind up with a <i>stream</i> in your target form.
-     *  The internal content model is never effected.
-     ******************************************************************************************/
-    virtual PRInt32 Convert(nsIURL* aURL,char* aSourceForm,char* aTargetForm,nsIStreamListener* aListener) = 0;
-    virtual PRInt32 Convert(const char* aFilename,char* aSourceForm,char* aTargetForm)=0;
-    virtual PRInt32 Convert(nsString& anHTMLString,char* aSourceForm,char* aTargetForm,PRBool appendTokens)=0;
 
     virtual PRInt32 ResumeParse(void)=0;
 

mozilla/htmlparser/src/nsParser.cpp

@@ -31,6 +31,13 @@
 #include "nsIDTDDebug.h"
 #include "nshtmlpars.h"
 
+#define rickgdebug 1
+#ifdef  rickgdebug
+#include "CRtfDTD.h"
+#include "nsWellFormedDTD.h"
+#endif
+
+
 static NS_DEFINE_IID(kISupportsIID, NS_ISUPPORTS_IID);                 
 static NS_DEFINE_IID(kClassIID, NS_PARSER_IID); 
 static NS_DEFINE_IID(kIParserIID, NS_IPARSER_IID);
@@ -136,7 +143,8 @@ CSharedParserObjects gSharedParserObjects;
  */
 nsParser::nsParser() : 
   mTokenDeque(gTokenDeallocator),
-  mContentType()  
+  mSourceType(),
+  mTargetType()
 {
   NS_INIT_REFCNT();
   mDTDDebug = 0;
@@ -270,6 +278,15 @@ nsIContentSink* nsParser::SetContentSink(nsIContentSink* aSink) {
  *  @return  nothing.
  */
 void nsParser::RegisterDTD(nsIDTD* aDTD){
+
+#ifdef rickgdebug
+  nsIDTD* rv=0;
+  NS_NewRTF_DTD(&rv);
+  gSharedParserObjects.RegisterDTD(rv);
+  NS_NewWellFormed_DTD(&rv);
+  gSharedParserObjects.RegisterDTD(rv);
+#endif
+
   gSharedParserObjects.RegisterDTD(aDTD);
 }
 
@@ -309,28 +326,31 @@ eParseMode DetermineParseMode() {
  *  @param   
  *  @return  
  */
-PRBool FindSuitableDTD(eProcessType aProcessType,eParseMode aMode,nsString& aContentType, nsIDTD*& aDefaultDTD) {
- PRBool result=PR_FALSE;
+PRBool FindSuitableDTD( eParseMode aMode,
+                        nsString& aSourceType, 
+                        nsString& aTargetType, 
+                        nsIDTD*& aDefaultDTD) {
 
-  if(aDefaultDTD) {
-    if(aDefaultDTD->IsCapableOf(aProcessType,aContentType,0)) {
-      result=PR_TRUE;
-    }
-  }
-  else {
-    nsDequeIterator b=gSharedParserObjects.mDTDDeque.Begin(); 
-    nsDequeIterator e=gSharedParserObjects.mDTDDeque.End(); 
+    //Let's start by tring the defaultDTD, if one exists...
+  if(aDefaultDTD && (aDefaultDTD->CanParse(aSourceType,0)))
+    return PR_TRUE;
 
-    while(b<e){
-      nsIDTD* theDTD=(nsIDTD*)b.GetCurrent();
-      if(theDTD && theDTD->IsCapableOf(aProcessType,aContentType,0)) {
+  PRBool  result=PR_FALSE;
+  nsDequeIterator b=gSharedParserObjects.mDTDDeque.Begin(); 
+  nsDequeIterator e=gSharedParserObjects.mDTDDeque.End(); 
+
+  while(b<e){
+    nsIDTD* theDTD=(nsIDTD*)b.GetCurrent();
+    if(theDTD) {
+      result=theDTD->CanParse(aSourceType,0); 
+      if(result){
         aDefaultDTD=theDTD;
-        result=PR_TRUE;
         break;
       }
-      ++b;
     }
-  }
+    b++;
+  } 
+
   return result;
 }
 
@@ -341,22 +361,30 @@ PRBool FindSuitableDTD(eProcessType aProcessType,eParseMode aMode,nsString& aCon
  *
  * @update	gess6/22/98
  * @param   aBuffer -- nsString containing sample data to be analyzed.
+ * @param   aType -- may hold typename given from netlib; will hold result given by DTD's.
  * @return  auto-detect result: eValid, eInvalid, eUnknown
  */
-eAutoDetectResult nsParser::AutoDetectContentType(nsString& aBuffer) {
+eAutoDetectResult nsParser::AutoDetectContentType(nsString& aBuffer,nsString& aType) {
 
     //The process:
     //  You should go out and ask each DTD if they
     //  recognize the content in the scanner.
     //  Somebody should say yes, or we can't continue.
 
-    //This method may change mContentType and mDTD.
+    //This method may change mSourceType and mDTD.
     //It absolutely changes mAutoDetectStatus
 
+  nsDequeIterator b=gSharedParserObjects.mDTDDeque.Begin(); 
+  nsDequeIterator e=gSharedParserObjects.mDTDDeque.End(); 
+
   mAutoDetectStatus=eUnknownDetect;
-  if(PR_TRUE==mContentType.Equals(kHTMLTextContentType)) {
-    mAutoDetectStatus=eValidDetect;
-  }
+  while((b<e) && (eUnknownDetect==mAutoDetectStatus)){
+    nsIDTD* theDTD=(nsIDTD*)b.GetCurrent();
+    if(theDTD) {
+      mAutoDetectStatus=theDTD->AutoDetectContentType(aBuffer,aType);
+    }
+    b++;
+  } 
 
   return mAutoDetectStatus;  
 }
@@ -373,13 +401,13 @@ eAutoDetectResult nsParser::AutoDetectContentType(nsString& aBuffer) {
  * @param 
  * @return
  */
-PRInt32 nsParser::WillBuildModel(eProcessType aProcessType, const char* aFilename){
+PRInt32 nsParser::WillBuildModel(const char* aFilename){
 
   mMajorIteration=-1;
   mMinorIteration=-1;
 
   mParseMode=DetermineParseMode();  
-  if(PR_TRUE==FindSuitableDTD(aProcessType,mParseMode,mContentType,mDTD)) {
+  if(PR_TRUE==FindSuitableDTD(mParseMode,mSourceType,mTargetType,mDTD)) {
     mDTD->SetParser(this);
     mDTD->SetContentSink(mSink);
     mDTD->WillBuildModel(aFilename);
@@ -439,11 +467,12 @@ PRBool nsParser::Parse(const char* aFilename){
   if(aFilename) {
 
     //ok, time to create our tokenizer and begin the process
+    mTargetType=kHTMLTextContentType;
     mScanner=new CScanner(aFilename,mParseMode);
     if(mScanner) {
       mScanner->Eof();
-      if(eValidDetect==AutoDetectContentType(mScanner->GetBuffer())) {
-        WillBuildModel(eParsing,aFilename);
+      if(eValidDetect==AutoDetectContentType(mScanner->GetBuffer(),mSourceType)) {
+        WillBuildModel(aFilename);
         status=ResumeParse();
         DidBuildModel(status);
       }
@@ -452,6 +481,18 @@ PRBool nsParser::Parse(const char* aFilename){
   return status;
 }
 
+/**
+ * Cause parser to parse input from given stream 
+ * @update	gess5/11/98
+ * @param   aStream is the i/o source
+ * @return  TRUE if all went well -- FALSE otherwise
+ */
+PRInt32 nsParser::Parse(fstream& aStream){
+  PRInt32 result=0;
+  return result;
+}
+
+
 /**
  *  This is the main controlling routine in the parsing process. 
  *  Note that it may get called multiple times for the same scanner, 
@@ -467,54 +508,6 @@ PRBool nsParser::Parse(const char* aFilename){
  *  @return  PR_TRUE if parse succeeded, PR_FALSE otherwise.
  */
 PRInt32 nsParser::Parse(nsIURL* aURL,nsIStreamObserver* aListener, nsIDTDDebug * aDTDDebug) {
-  PRInt32 status;
-
-  status = BeginParse(aURL, aListener, aDTDDebug);
-  if (NS_OK == status) {
-    status=mURL->Open(this);
-  }
-  return status;
-}
-
-
-/**
- * Call this method if all you want to do is parse 1 string full of HTML text.
- *
- * @update	gess5/11/98
- * @param   anHTMLString contains a string-full of real HTML
- * @param   appendTokens tells us whether we should insert tokens inline, or append them.
- * @return  TRUE if all went well -- FALSE otherwise
- */
-PRInt32 nsParser::Parse(nsString& aSourceBuffer,PRBool appendTokens){
-  PRInt32 result=kNoError;
-
-  mScanner=new CScanner(); 
-  mScanner->Append(aSourceBuffer);
-  
-  if(eValidDetect==AutoDetectContentType(aSourceBuffer)) {
-    WillBuildModel(eParsing,"from-string");
-    result=ResumeParse();
-    DidBuildModel(result);
-  }
-  
-  return result;
-}
-
-/**
- *  This is the main controlling routine in the parsing process. 
- *  Note that it may get called multiple times for the same scanner, 
- *  since this is a pushed based system, and all the tokens may 
- *  not have been consumed by the scanner during a given invocation 
- *  of this method. 
- *
- *  NOTE: We don't call willbuildmodel here, because it will happen
- *        as a result of calling OnStartBinding later on.
- *
- *  @update  gess 3/25/98
- *  @param   aFilename -- const char* containing file to be parsed.
- *  @return  PR_TRUE if parse succeeded, PR_FALSE otherwise.
- */
-PRInt32 nsParser::BeginParse(nsIURL* aURL,nsIStreamObserver* aListener, nsIDTDDebug * aDTDDebug) {
   NS_PRECONDITION(0!=aURL,kNullURL);
 
   PRInt32 status=kBadURL;
@@ -537,36 +530,27 @@ PRInt32 nsParser::BeginParse(nsIURL* aURL,nsIStreamObserver* aListener, nsIDTDDe
   return status;
 }
 
-/**
- * 
- * @update	gess6/23/98
- * @param 
- * @return
- */
-PRInt32 nsParser::Convert(nsIURL* aURL,char* aSourceForm,char* aTargetForm,nsIStreamListener* aListener) {
-  PRInt32 result=0;
-  return result;
-}
 
 /**
- * 
- * @update	gess6/23/98
- * @param 
- * @return
+ * Call this method if all you want to do is parse 1 string full of HTML text.
+ *
+ * @update	gess5/11/98
+ * @param   anHTMLString contains a string-full of real HTML
+ * @param   appendTokens tells us whether we should insert tokens inline, or append them.
+ * @return  TRUE if all went well -- FALSE otherwise
  */
-PRInt32 nsParser::Convert(const char* aFilename,char* aSourceForm,char* aTargetForm) {
-  PRInt32 result=0;
-  return result;
-}
+PRInt32 nsParser::Parse(nsString& aSourceBuffer,PRBool appendTokens){
+  PRInt32 result=kNoError;
 
-/** 
- * 
- * @update	gess6/23/98
- * @param 
- * @return
- */
-PRInt32 nsParser::Convert(nsString& anHTMLString,char* aSourceForm,char* aTargetForm,PRBool appendTokens){
-  PRInt32 result=0;
+  mTargetType=kHTMLTextContentType;
+  mScanner=new CScanner(); 
+  mScanner->Append(aSourceBuffer);  
+  if(eValidDetect==AutoDetectContentType(aSourceBuffer,mSourceType)) {
+    WillBuildModel("");
+    result=ResumeParse();
+    DidBuildModel(result);
+  }
+  
   return result;
 }
 
@@ -722,14 +706,14 @@ nsParser::OnProgress(PRInt32 aProgress, PRInt32 aProgressMax,
  *  @param   
  *  @return  
  */
-nsresult nsParser::OnStartBinding(const char *aContentType){
+nsresult nsParser::OnStartBinding(const char *aSourceType){
   if (nsnull != mObserver) {
-    mObserver->OnStartBinding(aContentType);
+    mObserver->OnStartBinding(aSourceType);
   }
   mAutoDetectStatus=eUnknownDetect;
   mDTD=0;
 
-  mContentType=aContentType;
+  mSourceType=aSourceType;
   return kNoError;
 }
 
@@ -773,8 +757,8 @@ nsresult nsParser::OnDataAvailable(nsIInputStream *pIStream, PRInt32 length){
         mScanner->Append(mTransferBuffer,len);
 
         if(eUnknownDetect==mAutoDetectStatus) {
-          if(eValidDetect==AutoDetectContentType(mScanner->GetBuffer())) {
-            nsresult result=WillBuildModel(eParsing,mURL->GetSpec());
+          if(eValidDetect==AutoDetectContentType(mScanner->GetBuffer(),mSourceType)) {
+            nsresult result=WillBuildModel(mURL->GetSpec());
           } //if
         }
       } //if

mozilla/htmlparser/src/nsParser.h

@@ -74,7 +74,7 @@ class nsIDTD;
 class nsIDTDDebug;
 class CScanner;
 class nsIParserFilter;
-
+class fstream;
 
 class nsParser : public nsIParser, public nsIStreamListener {
             
@@ -118,10 +118,6 @@ friend class CTokenHandler;
      */
     virtual CScanner* GetScanner(void);
 
-    virtual PRInt32 BeginParse(nsIURL* aURL,
-                               nsIStreamObserver* aListener, 
-                               nsIDTDDebug * aDTDDebug = 0);
-
     /**
      * Cause parser to parse input from given URL in given mode
      * @update	gess5/11/98
@@ -136,11 +132,18 @@ friend class CTokenHandler;
      * Cause parser to parse input from given file in given mode
      * @update	gess5/11/98
      * @param   aFilename is a path for file document
-     * @param   aMode is the desired parser mode (Nav, other, etc.)
      * @return  TRUE if all went well -- FALSE otherwise
      */
     virtual PRInt32 Parse(const char* aFilename);
 
+    /**
+     * Cause parser to parse input from given stream 
+     * @update	gess5/11/98
+     * @param   aStream is the i/o source
+     * @return  TRUE if all went well -- FALSE otherwise
+     */
+    virtual PRInt32 Parse(fstream& aStream);
+
     /**
      * @update	gess5/11/98
      * @param   anHTMLString contains a string-full of real HTML
@@ -149,15 +152,6 @@ friend class CTokenHandler;
      */
     virtual PRInt32 Parse(nsString& anHTMLString,PRBool appendTokens);
 
-    /******************************************************************************************
-     *  Convert methods start input source (of known or unknown form), and perform conversions 
-     *  until you wind up with a <i>stream</i> in your target form.
-     *  The internal content model is never effected.
-     ******************************************************************************************/
-    virtual PRInt32 Convert(nsIURL* aURL,char* aSourceForm, char* aTargetForm,nsIStreamListener* aListener);
-    virtual PRInt32 Convert(const char* aFilename,char* aSourceForm,char* aTargetForm);
-    virtual PRInt32 Convert(nsString& anHTMLString,char* aSourceForm,char* aTargetForm,PRBool appendTokens);
-
     /**
      * This method gets called (automatically) during incremental parsing
      * @update	gess5/11/98
@@ -214,7 +208,7 @@ protected:
      * @param 
      * @return
      */
-    PRInt32 WillBuildModel(eProcessType theProcessType=eParsing,const char* aFilename=0);
+    PRInt32 WillBuildModel(const char* aFilename=0);
 
     /**
      * 
@@ -309,7 +303,7 @@ private:
      * @param 
      * @return  TRUE if we figured it out.
      */
-    eAutoDetectResult AutoDetectContentType(nsString& aBuffer);
+    eAutoDetectResult AutoDetectContentType(nsString& aBuffer,nsString& aType);
 
 
 protected:
@@ -335,7 +329,8 @@ protected:
     CScanner*           mScanner;
     nsIURL*             mURL;
 	  nsIDTDDebug*		    mDTDDebug;
-    nsString            mContentType;
+    nsString            mSourceType;
+    nsString            mTargetType;
     eAutoDetectResult   mAutoDetectStatus;
 };
 

mozilla/htmlparser/src/nsParserTypes.h

@@ -42,40 +42,41 @@ enum  eParseMode {
 };
 
 
-const PRInt32   kEOF          = 1000000L;
-const PRInt32   kUnknownError = -1000;
-const PRInt32   kCantPropagate = -6;
-const PRInt32   kContextMismatch = -5;
-const PRInt32   kBadFilename  = -4;
-const PRInt32   kBadURL       = -3;
-const PRInt32   kInterrupted  = -2;
-const PRInt32   kNotFound     = -1;
-const PRInt32   kNoError      = 0;
+const PRInt32   kEOF              = 1000000L;
+const PRInt32   kUnknownError     = -1000;
+const PRInt32   kCantPropagate    = -6;
+const PRInt32   kContextMismatch  = -5;
+const PRInt32   kBadFilename      = -4;
+const PRInt32   kBadURL           = -3;
+const PRInt32   kInterrupted      = -2;
+const PRInt32   kNotFound         = -1;
+const PRInt32   kNoError          = 0;
 
-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  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  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  kForwardSlash     = '/';
+const PRUint32  kBackSlash        = '\\';
+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       = '}';
 
 
 #endif

mozilla/htmlparser/src/nsScanner.cpp

@@ -47,6 +47,7 @@ CScanner::CScanner(eParseMode aMode) : mBuffer("") {
   mNetStream=0;
   mFileStream=0;
   mIncremental=PR_TRUE;
+  mOwnsStream=PR_TRUE;
 }
 
 /**
@@ -64,6 +65,7 @@ CScanner::CScanner(const char* aFilename,eParseMode aMode) : mBuffer("") {
   mParseMode=aMode;
   mNetStream=0;
   mIncremental=PR_FALSE;
+  mOwnsStream=PR_TRUE;
 #if defined(XP_UNIX) && (defined(IRIX) || defined(MKLINUX))
   /* XXX: IRIX does not support ios::binary */
   mFileStream=new fstream(aFilename,ios::in);
@@ -72,6 +74,24 @@ CScanner::CScanner(const char* aFilename,eParseMode aMode) : mBuffer("") {
 #endif
 }
 
+/**
+ *  Use this constructor if you want i/o to be file based.
+ *
+ *  @update  gess 5/12/98
+ *  @param   aMode represents the parser mode (nav, other)
+ *  @return  
+ */
+CScanner::CScanner(fstream& aStream,eParseMode aMode) : mBuffer("") {
+  mOffset=0;
+  mMarkPos=-1;
+  mTotalRead=0;
+  mParseMode=aMode;
+  mNetStream=0;
+  mIncremental=PR_FALSE;
+  mOwnsStream=PR_FALSE;
+  mFileStream=&aStream;
+}
+
 /**
  *  Use this constructor if you want i/o to be based on a
  *  non-incremental netstream.
@@ -91,6 +111,7 @@ CScanner::CScanner(nsIURL* aURL,eParseMode aMode) : mBuffer("") {
   mIncremental=PR_FALSE;
   mNetStream=aURL->Open(&error);
   gURLRef=aURL->GetSpec();
+  mOwnsStream=PR_FALSE;
 }
 
 
@@ -104,7 +125,8 @@ CScanner::CScanner(nsIURL* aURL,eParseMode aMode) : mBuffer("") {
 CScanner::~CScanner() {
   if(mFileStream) {
     mFileStream->close();
-    delete mFileStream;
+    if(mOwnsStream)
+      delete mFileStream;
   }
   else if(mNetStream) {
     mNetStream->Close();
@@ -354,7 +376,7 @@ PRInt32 CScanner::PutBack(PRUnichar aChar) {
  *  @param   
  *  @return  error status
  */
-PRInt32 CScanner::SkipWhite(void) {
+PRInt32 CScanner::SkipWhitespace(void) {
   static nsAutoString chars(" \n\r\t");
   return SkipOver(chars);
 }

mozilla/htmlparser/src/nsScanner.h

@@ -25,7 +25,7 @@
  * how to consume characters out of an (internal) stream.
  * This class also offers a series of utility methods
  * that most tokenizers want, such as readUntil(), 
- * readWhile() and SkipWhite().
+ * readWhile() and SkipWhitespace().
  */
 
 
@@ -73,6 +73,16 @@ class CScanner {
        */
       CScanner(const char* aFilename,eParseMode aMode=eParseMode_navigator);
 
+      /**
+       *  Use this constructor if you want i/o to be stream based.
+       *
+       *  @update  gess 5/12/98
+       *  @param   aMode represents the parser mode (nav, other)
+       *  @return  
+       */
+      CScanner(fstream& aStream,eParseMode aMode=eParseMode_navigator);
+
+
       ~CScanner();
 
       /**
@@ -136,7 +146,7 @@ class CScanner {
        *  @update  gess 3/25/98
        *  @return  error status
        */
-      PRInt32 SkipWhite(void);
+      PRInt32 SkipWhitespace(void);
 
       /**
        *  Determine if the scanner has reached EOF.
@@ -265,6 +275,7 @@ class CScanner {
       PRInt32         mTotalRead;
       eParseMode      mParseMode;
       PRBool          mIncremental;
+      PRBool          mOwnsStream;
 };
 
 #endif

mozilla/htmlparser/src/nsValidDTD.cpp

@@ -0,0 +1,334 @@
+/* -*- Mode: C++; tab-width: 2; indent-tabs-mode: nil; c-basic-offset: 2 -*- */
+/*
+ * The contents of this file are subject to the Netscape Public License
+ * Version 1.0 (the "NPL"); you may not use this file except in
+ * compliance with the NPL.  You may obtain a copy of the NPL at
+ * http://www.mozilla.org/NPL/
+ *
+ * Software distributed under the NPL is distributed on an "AS IS" basis,
+ * WITHOUT WARRANTY OF ANY KIND, either express or implied. See the NPL
+ * for the specific language governing rights and limitations under the
+ * NPL.
+ *
+ * The Initial Developer of this code under the NPL is Netscape
+ * Communications Corporation.  Portions created by Netscape are
+ * Copyright (C) 1998 Netscape Communications Corporation.  All Rights
+ * Reserved.
+ */
+
+/**
+ * MODULE NOTES:
+ * @update  gess 4/8/98
+ * 
+ *         
+ */
+
+/**
+ * TRANSIENT STYLE-HANDLING NOTES:
+ * @update  gess 6/15/98
+ * 
+ * ...add comments here about transient style stack.
+ *         
+   */
+
+#include "nsIDTDDebug.h"
+#include "nsValidDTD.h"
+#include "nsCRT.h"
+#include "nsParser.h"
+#include "nsScanner.h"
+#include "nsParserTypes.h"
+#include "nsTokenHandler.h"
+
+#include "prenv.h"  //this is here for debug reasons...
+#include "prtypes.h"  //this is here for debug reasons...
+#include "prio.h"
+#include "plstr.h"
+
+#ifdef XP_PC
+#include <direct.h> //this is here for debug reasons...
+#endif
+#include "prmem.h"
+
+
+static NS_DEFINE_IID(kISupportsIID, NS_ISUPPORTS_IID);                 
+static NS_DEFINE_IID(kIDTDIID,      NS_IDTD_IID);
+static NS_DEFINE_IID(kClassIID,     NS_VALID_DTD_IID); 
+
+
+static const char* kNullURL = "Error: Null URL given";
+static const char* kNullFilename= "Error: Null filename given";
+static const char* kNullTokenizer = "Error: Unable to construct tokenizer";
+static const char* kNullToken = "Error: Null token given";
+static const char* kInvalidTagStackPos = "Error: invalid tag stack position";
+static const char* kRTFTextContentType = "text/xml";
+
+static nsAutoString gEmpty;
+
+
+/**
+ *  This method gets called as part of our COM-like interfaces.
+ *  Its purpose is to create an interface to parser object
+ *  of some type.
+ *  
+ *  @update   gess 4/8/98
+ *  @param    nsIID  id of object to discover
+ *  @param    aInstancePtr ptr to newly discovered interface
+ *  @return   NS_xxx result code
+ */
+nsresult CValidDTD::QueryInterface(const nsIID& aIID, void** aInstancePtr)  
+{                                                                        
+  if (NULL == aInstancePtr) {                                            
+    return NS_ERROR_NULL_POINTER;                                        
+  }                                                                      
+
+  if(aIID.Equals(kISupportsIID))    {  //do IUnknown...
+    *aInstancePtr = (nsIDTD*)(this);                                        
+  }
+  else if(aIID.Equals(kIDTDIID)) {  //do IParser base class...
+    *aInstancePtr = (nsIDTD*)(this);                                        
+  }
+  else if(aIID.Equals(kClassIID)) {  //do this class...
+    *aInstancePtr = (CValidDTD*)(this);                                        
+  }                 
+  else {
+    *aInstancePtr=0;
+    return NS_NOINTERFACE;
+  }
+  ((nsISupports*) *aInstancePtr)->AddRef();
+  return NS_OK;                                                        
+}
+
+/**
+ *  This method is defined in nsIParser. It is used to 
+ *  cause the COM-like construction of an nsParser.
+ *  
+ *  @update  gess 4/8/98
+ *  @param   nsIParser** ptr to newly instantiated parser
+ *  @return  NS_xxx error result
+ */
+NS_HTMLPARS nsresult NS_NewValid_DTD(nsIDTD** aInstancePtrResult)
+{
+  CValidDTD* it = new CValidDTD();
+
+  if (it == 0) {
+    return NS_ERROR_OUT_OF_MEMORY;
+  }
+
+  return it->QueryInterface(kClassIID, (void **) aInstancePtrResult);
+}
+
+
+NS_IMPL_ADDREF(CValidDTD)
+NS_IMPL_RELEASE(CValidDTD)
+
+
+/**
+ *  Default constructor
+ *  
+ *  @update  gess 4/9/98
+ *  @param   
+ *  @return  
+ */
+CValidDTD::CValidDTD() : nsIDTD() {
+  NS_INIT_REFCNT();
+  mParser=0;
+  mFilename=0;
+}
+
+/**
+ *  Default destructor
+ *  
+ *  @update  gess 4/9/98
+ *  @param   
+ *  @return  
+ */
+CValidDTD::~CValidDTD(){
+}
+
+/**
+ * This method is called to determine if the given DTD can parse
+ * a document in a given source-type. 
+ * NOTE: Parsing always assumes that the end result will involve
+ *       storing the result in the main content model.
+ * @update	gess6/24/98
+ * @param   
+ * @return  TRUE if this DTD can satisfy the request; FALSE otherwise.
+ */
+PRBool CValidDTD::CanParse(nsString& aContentType, PRInt32 aVersion){
+  PRBool result=aContentType.Equals(kRTFTextContentType);
+  return result;
+}
+
+
+/**
+ * This method is called to determine if the given DTD can perform
+ * a document conversion from a given source-type to a given target-type.
+ * NOTE: conversion always assumes an input stream and an outputstream.
+ * @update	gess6/24/98
+ * @param   
+ * @return  TRUE if this DTD can satisfy the request; FALSE otherwise.
+ */
+PRBool CValidDTD::CanConvert(nsString& aSourceType, nsString& aTargetType, PRInt32 aVersion){
+  PRBool result=PR_TRUE;
+  return result;
+}
+
+/**
+ * 
+ * @update	gess7/7/98
+ * @param 
+ * @return
+ */
+eAutoDetectResult CValidDTD::AutoDetectContentType(nsString& aBuffer,nsString& aType){
+  eAutoDetectResult result=eUnknownDetect;
+  return result;
+}
+
+/**
+ * 
+ * @update	gess5/18/98
+ * @param 
+ * @return
+ */
+PRInt32 CValidDTD::WillBuildModel(const char* aFilename){
+  PRInt32 result=0;
+  return result;
+}
+
+/**
+ * 
+ * @update	gess5/18/98
+ * @param 
+ * @return
+ */
+PRInt32 CValidDTD::DidBuildModel(PRInt32 anErrorCode){
+  PRInt32 result=0;
+
+  return result;
+}
+
+/**
+ * 
+ *  
+ *  @update  gess 3/25/98
+ *  @param   
+ *  @return 
+ */
+void CValidDTD::SetParser(nsIParser* aParser) {
+  mParser=(nsParser*)aParser;
+}
+
+/**
+ *  This method gets called in order to set the content
+ *  sink for this parser to dump nodes to.
+ *  
+ *  @update  gess 3/25/98
+ *  @param   nsIContentSink interface for node receiver
+ *  @return  
+ */
+nsIContentSink* CValidDTD::SetContentSink(nsIContentSink* aSink) {
+  return 0;
+}
+
+
+/*******************************************************************
+  These methods use to be hidden in the tokenizer-delegate. 
+  That file merged with the DTD, since the separation wasn't really
+  buying us anything.
+ *******************************************************************/
+
+/**
+ *  This method repeatedly called by the tokenizer. 
+ *  Each time, we determine the kind of token were about to 
+ *  read, and then we call the appropriate method to handle
+ *  that token type.
+ *  
+ *  @update gess 3/25/98
+ *  @param  aChar: last char read
+ *  @param  aScanner: see nsScanner.h
+ *  @param  anErrorCode: arg that will hold error condition
+ *  @return new token or null 
+ */
+PRInt32 CValidDTD::ConsumeToken(CToken*& aToken){
+  
+  CScanner* theScanner=mParser->GetScanner();
+
+  PRUnichar aChar;
+  PRInt32   result=theScanner->GetChar(aChar);
+
+  switch(result) {
+    case kEOF:
+      break;
+
+    case kInterrupted:
+      theScanner->RewindToMark();
+      break; 
+
+    case kNoError:
+    default:
+      switch(aChar) {
+        case kLessThan:
+        default:
+          break;
+      } //switch
+      break; 
+  } //switch
+  if(kNoError==result)
+    result=theScanner->Eof();
+  return result;
+}
+
+/**
+ * 
+ * @update	gess5/18/98
+ * @param 
+ * @return
+ */
+void CValidDTD::WillResumeParse(void){
+  return;
+}
+
+/**
+ * 
+ * @update	gess5/18/98
+ * @param 
+ * @return
+ */
+void CValidDTD::WillInterruptParse(void){
+  return;
+}
+
+/**
+ * 
+ * @update	jevering6/23/98
+ * @param 
+ * @return
+ */
+void CValidDTD::SetDTDDebug(nsIDTDDebug * aDTDDebug) {
+}
+
+/**
+ *  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 -- int tag of parent container
+ *  @param   aChild -- int tag of child container
+ *  @return  PR_TRUE if parent can contain child
+ */
+PRBool CValidDTD::CanContain(PRInt32 aParent,PRInt32 aChild){
+  PRBool result=PR_FALSE;
+  return result;
+}
+
+/**
+ *  
+ *  @update  gess 3/25/98
+ *  @param   aToken -- token object to be put into content model
+ *  @return  0 if all is well; non-zero is an error
+ */
+PRInt32 CValidDTD::HandleToken(CToken* aToken) {
+  PRInt32 result=0;
+  return result;
+}
+

mozilla/htmlparser/src/nsValidDTD.h

@@ -0,0 +1,205 @@
+
+/* -*- Mode: C++; tab-width: 2; indent-tabs-mode: nil; c-basic-offset: 2 -*- */
+/*
+ * The contents of this file are subject to the Netscape Public License
+ * Version 1.0 (the "NPL"); you may not use this file except in
+ * compliance with the NPL.  You may obtain a copy of the NPL at
+ * http://www.mozilla.org/NPL/
+ *
+ * Software distributed under the NPL is distributed on an "AS IS" basis,
+ * WITHOUT WARRANTY OF ANY KIND, either express or implied. See the NPL
+ * for the specific language governing rights and limitations under the
+ * NPL.
+ *
+ * The Initial Developer of this code under the NPL is Netscape
+ * Communications Corporation.  Portions created by Netscape are
+ * Copyright (C) 1998 Netscape Communications Corporation.  All Rights
+ * Reserved.
+ */
+
+/**
+ * MODULE NOTES:
+ * @update  gess 4/8/98
+ * 
+ *         
+ */
+
+#ifndef __NS_VALID_DTD
+#define __NS_VALID_DTD
+
+#include "nsIDTD.h"
+#include "nsISupports.h"
+#include "nsHTMLTokens.h"
+#include "nshtmlpars.h"
+#include "nsVoidArray.h"
+#include "nsDeque.h"
+
+#define NS_VALID_DTD_IID      \
+  {0xa39c6bfe, 0x15f0,    0x11d2, \
+  {0x80, 0x41, 0x0, 0x10, 0x4b, 0x98, 0x3f, 0xd4}}
+
+
+class nsIHTMLContentSink;
+class nsIDTDDebug;
+class nsIParserNode;
+class CITokenHandler;
+class nsParser;
+
+
+
+class CValidDTD : public nsIDTD {
+            
+  public:
+
+    NS_DECL_ISUPPORTS
+
+
+    /**
+     *  
+     *  
+     *  @update  gess 4/9/98
+     *  @param   
+     *  @return  
+     */
+    CValidDTD();
+
+    /**
+     *  
+     *  
+     *  @update  gess 4/9/98
+     *  @param   
+     *  @return  
+     */
+    virtual ~CValidDTD();
+
+    /**
+     * This method is called to determine if the given DTD can parse
+     * a document in a given source-type. 
+     * NOTE: Parsing always assumes that the end result will involve
+     *       storing the result in the main content model.
+     * @update	gess6/24/98
+     * @param   
+     * @return  TRUE if this DTD can satisfy the request; FALSE otherwise.
+     */
+    virtual PRBool CanParse(nsString& aContentType, PRInt32 aVersion);
+
+
+    /**
+     * This method is called to determine if the given DTD can perform
+     * a document conversion from a given source-type to a given target-type.
+     * NOTE: conversion always assumes an input stream and an outputstream.
+     * @update	gess6/24/98
+     * @param   
+     * @return  TRUE if this DTD can satisfy the request; FALSE otherwise.
+     */
+    virtual PRBool CanConvert(nsString& aSourceType, nsString& aTargetType, PRInt32 aVersion);
+
+    /**
+     * 
+     * @update	gess7/7/98
+     * @param 
+     * @return
+     */
+    virtual eAutoDetectResult AutoDetectContentType(nsString& aBuffer,nsString& aType);
+
+    /**
+     * 
+     * @update	gess5/18/98
+     * @param 
+     * @return
+     */
+    virtual PRInt32 WillBuildModel(const char* aFilename=0);
+
+    /**
+     * 
+     * @update	gess5/18/98
+     * @param 
+     * @return
+     */
+    virtual PRInt32 DidBuildModel(PRInt32 anErrorCode);
+
+    /**
+     *  
+     *  @update  gess 3/25/98
+     *  @param   aToken -- token object to be put into content model
+     *  @return  0 if all is well; non-zero is an error
+     */
+    virtual PRInt32 HandleToken(CToken* aToken);
+
+    /**
+     * 
+     *  
+     *  @update  gess 3/25/98
+     *  @param   
+     *  @return 
+     */
+    virtual void SetParser(nsIParser* aParser);
+
+    /**
+     *  Cause the tokenizer to consume the next token, and 
+     *  return an error result.
+     *  
+     *  @update  gess 3/25/98
+     *  @param   anError -- ref to error code
+     *  @return  new token or null
+     */
+    virtual PRInt32 ConsumeToken(CToken*& aToken);
+
+
+    /**
+     * 
+     * @update	gess5/18/98
+     * @param 
+     * @return
+     */
+    virtual void WillResumeParse(void);
+
+    /**
+     * 
+     * @update	gess5/18/98
+     * @param 
+     * @return
+     */
+    virtual void WillInterruptParse(void);
+
+   /**
+     * Select given content sink into parser for parser output
+     * @update	gess5/11/98
+     * @param   aSink is the new sink to be used by parser
+     * @return  old sink, or NULL
+     */
+    virtual nsIContentSink* SetContentSink(nsIContentSink* aSink);
+
+    /**
+     * 
+     * @update	jevering6/23/98
+     * @param 
+     * @return
+     */
+	  virtual void SetDTDDebug(nsIDTDDebug * aDTDDebug);
+
+    /**
+     *  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 -- int tag of parent container
+     *  @param   aChild -- int tag of child container
+     *  @return  PR_TRUE if parent can contain child
+     */
+    virtual PRBool CanContain(PRInt32 aParent,PRInt32 aChild);
+	
+protected:
+
+    
+    nsParser*           mParser;
+    nsIHTMLContentSink* mSink;
+    char*               mFilename;
+};
+
+extern NS_HTMLPARS nsresult NS_NewValid_DTD(nsIDTD** aInstancePtrResult);
+
+#endif 
+
+
+

mozilla/htmlparser/src/nsWellFormedDTD.cpp

@@ -0,0 +1,321 @@
+/* -*- Mode: C++; tab-width: 2; indent-tabs-mode: nil; c-basic-offset: 2 -*- */
+/*
+ * The contents of this file are subject to the Netscape Public License
+ * Version 1.0 (the "NPL"); you may not use this file except in
+ * compliance with the NPL.  You may obtain a copy of the NPL at
+ * http://www.mozilla.org/NPL/
+ *
+ * Software distributed under the NPL is distributed on an "AS IS" basis,
+ * WITHOUT WARRANTY OF ANY KIND, either express or implied. See the NPL
+ * for the specific language governing rights and limitations under the
+ * NPL.
+ *
+ * The Initial Developer of this code under the NPL is Netscape
+ * Communications Corporation.  Portions created by Netscape are
+ * Copyright (C) 1998 Netscape Communications Corporation.  All Rights
+ * Reserved.
+ */
+
+/**
+ * MODULE NOTES:
+ * @update  gess 4/8/98
+ * 
+ *         
+ */
+
+/**
+ * TRANSIENT STYLE-HANDLING NOTES:
+ * @update  gess 6/15/98
+ * 
+ * ...add comments here about transient style stack.
+ *         
+   */
+
+#include "nsIDTDDebug.h"
+#include "nsWellFormedDTD.h"
+#include "nsCRT.h"
+#include "nsParser.h"
+#include "nsScanner.h"
+#include "nsParserTypes.h"
+#include "nsTokenHandler.h"
+
+#include "prenv.h"  //this is here for debug reasons...
+#include "prtypes.h"  //this is here for debug reasons...
+#include "prio.h"
+#include "plstr.h"
+
+#ifdef XP_PC
+#include <direct.h> //this is here for debug reasons...
+#endif
+#include "prmem.h"
+
+
+static NS_DEFINE_IID(kISupportsIID, NS_ISUPPORTS_IID);                 
+static NS_DEFINE_IID(kIDTDIID,      NS_IDTD_IID);
+static NS_DEFINE_IID(kClassIID,     NS_WELLFORMED_DTD_IID); 
+
+
+static const char* kNullURL = "Error: Null URL given";
+static const char* kNullFilename= "Error: Null filename given";
+static const char* kNullTokenizer = "Error: Unable to construct tokenizer";
+static const char* kNullToken = "Error: Null token given";
+static const char* kInvalidTagStackPos = "Error: invalid tag stack position";
+static const char* kRTFTextContentType = "text/xml";
+
+static nsAutoString gEmpty;
+
+
+/**
+ *  This method gets called as part of our COM-like interfaces.
+ *  Its purpose is to create an interface to parser object
+ *  of some type.
+ *  
+ *  @update   gess 4/8/98
+ *  @param    nsIID  id of object to discover
+ *  @param    aInstancePtr ptr to newly discovered interface
+ *  @return   NS_xxx result code
+ */
+nsresult CWellFormedDTD::QueryInterface(const nsIID& aIID, void** aInstancePtr)  
+{                                                                        
+  if (NULL == aInstancePtr) {                                            
+    return NS_ERROR_NULL_POINTER;                                        
+  }                                                                      
+
+  if(aIID.Equals(kISupportsIID))    {  //do IUnknown...
+    *aInstancePtr = (nsIDTD*)(this);                                        
+  }
+  else if(aIID.Equals(kIDTDIID)) {  //do IParser base class...
+    *aInstancePtr = (nsIDTD*)(this);                                        
+  }
+  else if(aIID.Equals(kClassIID)) {  //do this class...
+    *aInstancePtr = (CWellFormedDTD*)(this);                                        
+  }                 
+  else {
+    *aInstancePtr=0;
+    return NS_NOINTERFACE;
+  }
+  ((nsISupports*) *aInstancePtr)->AddRef();
+  return NS_OK;                                                        
+}
+
+/**
+ *  This method is defined in nsIParser. It is used to 
+ *  cause the COM-like construction of an nsParser.
+ *  
+ *  @update  gess 4/8/98
+ *  @param   nsIParser** ptr to newly instantiated parser
+ *  @return  NS_xxx error result
+ */
+NS_HTMLPARS nsresult NS_NewWellFormed_DTD(nsIDTD** aInstancePtrResult)
+{
+  CWellFormedDTD* it = new CWellFormedDTD();
+
+  if (it == 0) {
+    return NS_ERROR_OUT_OF_MEMORY;
+  }
+
+  return it->QueryInterface(kClassIID, (void **) aInstancePtrResult);
+}
+
+
+NS_IMPL_ADDREF(CWellFormedDTD)
+NS_IMPL_RELEASE(CWellFormedDTD)
+
+
+/**
+ *  Default constructor
+ *  
+ *  @update  gess 4/9/98
+ *  @param   
+ *  @return  
+ */
+CWellFormedDTD::CWellFormedDTD() : nsIDTD() {
+  NS_INIT_REFCNT();
+  mParser=0;
+  mFilename=0;
+}
+
+/**
+ *  Default destructor
+ *  
+ *  @update  gess 4/9/98
+ *  @param   
+ *  @return  
+ */
+CWellFormedDTD::~CWellFormedDTD(){
+}
+
+/**
+ * This method is called to determine if the given DTD can parse
+ * a document in a given source-type. 
+ * NOTE: Parsing always assumes that the end result will involve
+ *       storing the result in the main content model.
+ * @update	gess6/24/98
+ * @param   
+ * @return  TRUE if this DTD can satisfy the request; FALSE otherwise.
+ */
+PRBool CWellFormedDTD::CanParse(nsString& aContentType, PRInt32 aVersion){
+  PRBool result=aContentType.Equals(kRTFTextContentType);
+  return result;
+}
+
+
+/**
+ * 
+ * @update	gess7/7/98
+ * @param 
+ * @return
+ */
+eAutoDetectResult CWellFormedDTD::AutoDetectContentType(nsString& aBuffer,nsString& aType){
+  eAutoDetectResult result=eUnknownDetect;
+  return result;
+}
+
+/**
+ * 
+ * @update	gess5/18/98
+ * @param 
+ * @return
+ */
+PRInt32 CWellFormedDTD::WillBuildModel(const char* aFilename){
+  PRInt32 result=0;
+  return result;
+}
+
+/**
+ * 
+ * @update	gess5/18/98
+ * @param 
+ * @return
+ */
+PRInt32 CWellFormedDTD::DidBuildModel(PRInt32 anErrorCode){
+  PRInt32 result=0;
+
+  return result;
+}
+
+/**
+ * 
+ *  
+ *  @update  gess 3/25/98
+ *  @param   
+ *  @return 
+ */
+void CWellFormedDTD::SetParser(nsIParser* aParser) {
+  mParser=(nsParser*)aParser;
+}
+
+/**
+ *  This method gets called in order to set the content
+ *  sink for this parser to dump nodes to.
+ *  
+ *  @update  gess 3/25/98
+ *  @param   nsIContentSink interface for node receiver
+ *  @return  
+ */
+nsIContentSink* CWellFormedDTD::SetContentSink(nsIContentSink* aSink) {
+  return 0;
+}
+
+
+/*******************************************************************
+  These methods use to be hidden in the tokenizer-delegate. 
+  That file merged with the DTD, since the separation wasn't really
+  buying us anything.
+ *******************************************************************/
+
+/**
+ *  This method repeatedly called by the tokenizer. 
+ *  Each time, we determine the kind of token were about to 
+ *  read, and then we call the appropriate method to handle
+ *  that token type.
+ *  
+ *  @update gess 3/25/98
+ *  @param  aChar: last char read
+ *  @param  aScanner: see nsScanner.h
+ *  @param  anErrorCode: arg that will hold error condition
+ *  @return new token or null 
+ */
+PRInt32 CWellFormedDTD::ConsumeToken(CToken*& aToken){
+  
+  CScanner* theScanner=mParser->GetScanner();
+
+  PRUnichar aChar;
+  PRInt32   result=theScanner->GetChar(aChar);
+
+  switch(result) {
+    case kEOF:
+      break;
+
+    case kInterrupted:
+      theScanner->RewindToMark();
+      break; 
+
+    case kNoError:
+    default:
+      switch(aChar) {
+        case kLessThan:
+        default:
+          break;
+      } //switch
+      break; 
+  } //switch
+  if(kNoError==result)
+    result=theScanner->Eof();
+  return result;
+}
+
+/**
+ * 
+ * @update	gess5/18/98
+ * @param 
+ * @return
+ */
+void CWellFormedDTD::WillResumeParse(void){
+  return;
+}
+
+/**
+ * 
+ * @update	gess5/18/98
+ * @param 
+ * @return
+ */
+void CWellFormedDTD::WillInterruptParse(void){
+  return;
+}
+
+/**
+ * 
+ * @update	jevering6/23/98
+ * @param 
+ * @return
+ */
+void CWellFormedDTD::SetDTDDebug(nsIDTDDebug * aDTDDebug) {
+}
+
+/**
+ *  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 -- int tag of parent container
+ *  @param   aChild -- int tag of child container
+ *  @return  PR_TRUE if parent can contain child
+ */
+PRBool CWellFormedDTD::CanContain(PRInt32 aParent,PRInt32 aChild){
+  PRBool result=PR_FALSE;
+  return result;
+}
+
+/**
+ *  
+ *  @update  gess 3/25/98
+ *  @param   aToken -- token object to be put into content model
+ *  @return  0 if all is well; non-zero is an error
+ */
+PRInt32 CWellFormedDTD::HandleToken(CToken* aToken) {
+  PRInt32 result=0;
+  return result;
+}
+

mozilla/htmlparser/src/nsWellFormedDTD.h

@@ -0,0 +1,194 @@
+
+/* -*- Mode: C++; tab-width: 2; indent-tabs-mode: nil; c-basic-offset: 2 -*- */
+/*
+ * The contents of this file are subject to the Netscape Public License
+ * Version 1.0 (the "NPL"); you may not use this file except in
+ * compliance with the NPL.  You may obtain a copy of the NPL at
+ * http://www.mozilla.org/NPL/
+ *
+ * Software distributed under the NPL is distributed on an "AS IS" basis,
+ * WITHOUT WARRANTY OF ANY KIND, either express or implied. See the NPL
+ * for the specific language governing rights and limitations under the
+ * NPL.
+ *
+ * The Initial Developer of this code under the NPL is Netscape
+ * Communications Corporation.  Portions created by Netscape are
+ * Copyright (C) 1998 Netscape Communications Corporation.  All Rights
+ * Reserved.
+ */
+
+/**
+ * MODULE NOTES:
+ * @update  gess 4/8/98
+ * 
+ *         
+ */
+
+#ifndef __NS_WELLFORMED_DTD
+#define __NS_WELLFORMED_DTD
+
+#include "nsIDTD.h"
+#include "nsISupports.h"
+#include "nsHTMLTokens.h"
+#include "nshtmlpars.h"
+#include "nsVoidArray.h"
+#include "nsDeque.h"
+
+#define NS_WELLFORMED_DTD_IID      \
+  {0xa39c6bfd, 0x15f0,  0x11d2, \
+  {0x80, 0x41, 0x0, 0x10, 0x4b, 0x98, 0x3f, 0xd4}}
+
+
+class nsIHTMLContentSink;
+class nsIDTDDebug;
+class nsIParserNode;
+class CITokenHandler;
+class nsParser;
+
+
+
+class CWellFormedDTD : public nsIDTD {
+            
+  public:
+
+    NS_DECL_ISUPPORTS
+
+
+    /**
+     *  
+     *  
+     *  @update  gess 4/9/98
+     *  @param   
+     *  @return  
+     */
+    CWellFormedDTD();
+
+    /**
+     *  
+     *  
+     *  @update  gess 4/9/98
+     *  @param   
+     *  @return  
+     */
+    virtual ~CWellFormedDTD();
+
+    /**
+     * This method is called to determine if the given DTD can parse
+     * a document in a given source-type. 
+     * NOTE: Parsing always assumes that the end result will involve
+     *       storing the result in the main content model.
+     * @update	gess6/24/98
+     * @param   
+     * @return  TRUE if this DTD can satisfy the request; FALSE otherwise.
+     */
+    virtual PRBool CanParse(nsString& aContentType, PRInt32 aVersion);
+
+    /**
+     * 
+     * @update	gess7/7/98
+     * @param 
+     * @return
+     */
+    virtual eAutoDetectResult AutoDetectContentType(nsString& aBuffer,nsString& aType);
+
+    /**
+     * 
+     * @update	gess5/18/98
+     * @param 
+     * @return
+     */
+    virtual PRInt32 WillBuildModel(const char* aFilename=0);
+
+    /**
+     * 
+     * @update	gess5/18/98
+     * @param 
+     * @return
+     */
+    virtual PRInt32 DidBuildModel(PRInt32 anErrorCode);
+
+    /**
+     *  
+     *  @update  gess 3/25/98
+     *  @param   aToken -- token object to be put into content model
+     *  @return  0 if all is well; non-zero is an error
+     */
+    virtual PRInt32 HandleToken(CToken* aToken);
+
+    /**
+     * 
+     *  
+     *  @update  gess 3/25/98
+     *  @param   
+     *  @return 
+     */
+    virtual void SetParser(nsIParser* aParser);
+
+    /**
+     *  Cause the tokenizer to consume the next token, and 
+     *  return an error result.
+     *  
+     *  @update  gess 3/25/98
+     *  @param   anError -- ref to error code
+     *  @return  new token or null
+     */
+    virtual PRInt32 ConsumeToken(CToken*& aToken);
+
+
+    /**
+     * 
+     * @update	gess5/18/98
+     * @param 
+     * @return
+     */
+    virtual void WillResumeParse(void);
+
+    /**
+     * 
+     * @update	gess5/18/98
+     * @param 
+     * @return
+     */
+    virtual void WillInterruptParse(void);
+
+   /**
+     * Select given content sink into parser for parser output
+     * @update	gess5/11/98
+     * @param   aSink is the new sink to be used by parser
+     * @return  old sink, or NULL
+     */
+    virtual nsIContentSink* SetContentSink(nsIContentSink* aSink);
+
+    /**
+     * 
+     * @update	jevering6/23/98
+     * @param 
+     * @return
+     */
+	  virtual void SetDTDDebug(nsIDTDDebug * aDTDDebug);
+
+    /**
+     *  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 -- int tag of parent container
+     *  @param   aChild -- int tag of child container
+     *  @return  PR_TRUE if parent can contain child
+     */
+    virtual PRBool CanContain(PRInt32 aParent,PRInt32 aChild);
+	
+protected:
+
+    
+    nsParser*           mParser;
+    nsIHTMLContentSink* mSink;
+    char*               mFilename;
+};
+
+extern NS_HTMLPARS nsresult NS_NewWellFormed_DTD(nsIDTD** aInstancePtrResult);
+
+#endif 
+
+
+

mozilla/layout/html/document/src/nsHTMLContentSink.cpp

@@ -904,7 +904,7 @@ void HTMLContentSink::ReflowNewContent()
       shell->ContentAppended(mBody);
       NS_RELEASE(shell);
     }
-  }
+  } 
 #endif
 }
 

mozilla/layout/html/document/src/nsHTMLDocument.cpp

@@ -36,6 +36,12 @@
 #include "nsIWebWidget.h"
 #include "CNavDTD.h"
 
+
+//#define rickgdebug 1
+#ifdef rickgdebug
+#include "nsHTMLContentSinkStream.h"
+#endif
+
 static NS_DEFINE_IID(kIDocumentIID, NS_IDOCUMENT_IID);
 static NS_DEFINE_IID(kIDOMElementIID, NS_IDOMELEMENT_IID);
 static NS_DEFINE_IID(kIDOMTextIID, NS_IDOMTEXT_IID);
@@ -104,6 +110,13 @@ nsHTMLDocument::StartDocumentLoad(nsIURL *aURL,
   nsresult rv = NS_NewParser(&parser);
   if (NS_OK == rv) {
     nsIHTMLContentSink* sink;
+
+#ifdef rickgdebug
+    rv = NS_New_HTML_ContentSinkStream(&sink);
+#else
+    rv = NS_NewHTMLContentSink(&sink, this, aURL, aWebWidget);
+#endif
+
     rv = NS_NewHTMLContentSink(&sink, this, aURL, aWebWidget);
     if (NS_OK == rv) {
       nsIHTMLCSSStyleSheet* styleAttrSheet;
@@ -129,10 +142,9 @@ nsHTMLDocument::StartDocumentLoad(nsIURL *aURL,
         nsIDTD* theDTD=0;
         NS_NewNavHTMLDTD(&theDTD);
         parser->RegisterDTD(theDTD);
-        parser->RegisterDTD(theDTD);
 
         parser->SetContentSink(sink);
-        parser->BeginParse(aURL);
+        parser->Parse(aURL);
       }
       NS_RELEASE(sink);
     }

mozilla/parser/htmlparser/src/CNavDTD.cpp

@@ -282,18 +282,33 @@ void CNavDTD::SetDTDDebug(nsIDTDDebug * aDTDDebug)
 }
 
 /**
- * This method gets called by the parser to determine if this DTD can handle 
- * the requested process for the requested content type.
- *
+ * This method is called to determine if the given DTD can parse
+ * a document in a given source-type. 
+ * NOTE: Parsing always assumes that the end result will involve
+ *       storing the result in the main content model.
  * @update	gess6/24/98
- * @param 
- * @return  TRUE if the DTD can handle the process for this type; FALSE otherwise.
+ * @param   
+ * @return  TRUE if this DTD can satisfy the request; FALSE otherwise.
  */
-PRBool CNavDTD::IsCapableOf(eProcessType aProcessType, nsString& aContentType,PRInt32 aVersion){
+PRBool CNavDTD::CanParse(nsString& aContentType, PRInt32 aVersion){
   PRBool result=aContentType.Equals(kHTMLTextContentType);
   return result;
 }
 
+/**
+ * 
+ * @update	gess7/7/98
+ * @param 
+ * @return
+ */
+eAutoDetectResult CNavDTD::AutoDetectContentType(nsString& aBuffer,nsString& aType){
+  eAutoDetectResult result=eUnknownDetect;
+  if(PR_TRUE==aType.Equals(kHTMLTextContentType)) 
+    result=eValidDetect;
+  return result;
+}
+
+
 
 /**
  * 

mozilla/parser/htmlparser/src/CNavDTD.h

@@ -70,13 +70,24 @@ class CNavDTD : public nsIDTD {
      */
     virtual ~CNavDTD();
 
+    /**
+     * This method is called to determine if the given DTD can parse
+     * a document in a given source-type. 
+     * NOTE: Parsing always assumes that the end result will involve
+     *       storing the result in the main content model.
+     * @update	gess6/24/98
+     * @param   
+     * @return  TRUE if this DTD can satisfy the request; FALSE otherwise.
+     */
+    virtual PRBool CanParse(nsString& aContentType, PRInt32 aVersion);
+
     /**
      * 
-     * @update	gess7/1/98
+     * @update	gess7/7/98
      * @param 
      * @return
      */
-    virtual PRBool IsCapableOf(eProcessType aProcessType, nsString& aString, PRInt32 aVersion);
+    virtual eAutoDetectResult AutoDetectContentType(nsString& aBuffer,nsString& aType);
 
     /**
      * 

mozilla/parser/htmlparser/src/COtherDTD.cpp

@@ -149,17 +149,29 @@ COtherDTD::~COtherDTD(){
 }
 
 /**
- * This method gets called by the parser to determine if this DTD can handle 
- * the requested process for the requested content type.
- *
+ * This method is called to determine if the given DTD can parse
+ * a document in a given source-type. 
+ * NOTE: Parsing always assumes that the end result will involve
+ *       storing the result in the main content model.
  * @update	gess6/24/98
- * @param 
- * @return  TRUE if the DTD can handle the process for this type; FALSE otherwise.
+ * @param   
+ * @return  TRUE if this DTD can satisfy the request; FALSE otherwise.
  */
-PRBool COtherDTD::IsCapableOf(eProcessType aProcessType, nsString& aContentType,PRInt32 aVersion){
-  return CNavDTD::IsCapableOf(aProcessType,aContentType,aVersion);
+PRBool COtherDTD::CanParse(nsString& aContentType, PRInt32 aVersion){
+  return CNavDTD::CanParse(aContentType,aVersion);
 }
 
+/**
+ * 
+ * @update	gess7/7/98
+ * @param 
+ * @return
+ */
+eAutoDetectResult COtherDTD::AutoDetectContentType(nsString& aBuffer,nsString& aType){
+  return CNavDTD::AutoDetectContentType(aBuffer,aType);
+}
+
+
 /**
  * 
  * @update	gess5/18/98

mozilla/parser/htmlparser/src/COtherDTD.h

@@ -62,7 +62,24 @@ class COtherDTD : public CNavDTD {
      */
     virtual ~COtherDTD();
 
-    virtual PRBool IsCapableOf(eProcessType aProcessType, nsString& aString,PRInt32 aVersion);
+    /**
+     * This method is called to determine if the given DTD can parse
+     * a document in a given source-type. 
+     * NOTE: Parsing always assumes that the end result will involve
+     *       storing the result in the main content model.
+     * @update	gess6/24/98
+     * @param   
+     * @return  TRUE if this DTD can satisfy the request; FALSE otherwise.
+     */
+    virtual PRBool CanParse(nsString& aContentType, PRInt32 aVersion);
+
+    /**
+     * 
+     * @update	gess7/7/98
+     * @param 
+     * @return
+     */
+    virtual eAutoDetectResult AutoDetectContentType(nsString& aBuffer,nsString& aType);
     
     /**
      * 

mozilla/parser/htmlparser/src/CRtfDTD.cpp

@@ -0,0 +1,622 @@
+/* -*- Mode: C++; tab-width: 2; indent-tabs-mode: nil; c-basic-offset: 2 -*- */
+/*
+ * The contents of this file are subject to the Netscape Public License
+ * Version 1.0 (the "NPL"); you may not use this file except in
+ * compliance with the NPL.  You may obtain a copy of the NPL at
+ * http://www.mozilla.org/NPL/
+ *
+ * Software distributed under the NPL is distributed on an "AS IS" basis,
+ * WITHOUT WARRANTY OF ANY KIND, either express or implied. See the NPL
+ * for the specific language governing rights and limitations under the
+ * NPL.
+ *
+ * The Initial Developer of this code under the NPL is Netscape
+ * Communications Corporation.  Portions created by Netscape are
+ * Copyright (C) 1998 Netscape Communications Corporation.  All Rights
+ * Reserved.
+ */
+
+/**
+ * MODULE NOTES:
+ * @update  gess 4/8/98
+ * 
+ *         
+ */
+
+/**
+ * TRANSIENT STYLE-HANDLING NOTES:
+ * @update  gess 6/15/98
+ * 
+ * ...add comments here about transient style stack.
+ *         
+   */
+
+#include "nsIDTDDebug.h"
+#include "CRtfDTD.h"
+#include "nsCRT.h"
+#include "nsParser.h"
+#include "nsScanner.h"
+#include "nsParserTypes.h"
+#include "nsTokenHandler.h"
+
+#include "prenv.h"  //this is here for debug reasons...
+#include "prtypes.h"  //this is here for debug reasons...
+#include "prio.h"
+#include "plstr.h"
+
+#ifdef XP_PC
+#include <direct.h> //this is here for debug reasons...
+#endif
+#include "prmem.h"
+
+
+static NS_DEFINE_IID(kISupportsIID, NS_ISUPPORTS_IID);                 
+static NS_DEFINE_IID(kIDTDIID,      NS_IDTD_IID);
+static NS_DEFINE_IID(kClassIID,     NS_RTF_DTD_IID); 
+
+
+static const char* kNullURL = "Error: Null URL given";
+static const char* kNullFilename= "Error: Null filename given";
+static const char* kNullTokenizer = "Error: Unable to construct tokenizer";
+static const char* kNullToken = "Error: Null token given";
+static const char* kInvalidTagStackPos = "Error: invalid tag stack position";
+static const char* kRTFTextContentType = "application/rtf";
+static const char* kRTFDocHeader= "{\\rtf0";
+static nsString     gAlphaChars("abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ");
+static nsAutoString gDigits("-0123456789");
+
+static nsAutoString gEmpty;
+
+struct RTFEntry {
+  char      mName[10];
+  eRTFTags  mTagID;
+};
+
+
+static RTFEntry gRTFTable[] = {
+
+  {"$",eRTFCtrl_unknown},
+  {"'",eRTFCtrl_quote},
+  {"*",eRTFCtrl_star},   
+  {"0x0a",eRTFCtrl_linefeed},
+  {"0x0d",eRTFCtrl_return},
+  {"\\",eRTFCtrl_begincontrol},
+  {"b",eRTFCtrl_bold}, 
+  {"bin",eRTFCtrl_bin},
+  {"blue",eRTFCtrl_blue},
+  {"cols",eRTFCtrl_cols},
+  {"comment",eRTFCtrl_comment},
+  {"f",eRTFCtrl_font},
+  {"fonttbl",eRTFCtrl_fonttable},
+  {"green",eRTFCtrl_green},
+  {"i",eRTFCtrl_italic},
+  {"margb",eRTFCtrl_bottommargin},
+  {"margl",eRTFCtrl_leftmargin},
+  {"margr",eRTFCtrl_rightmargin},
+  {"margt",eRTFCtrl_topmargin},
+  {"par",eRTFCtrl_par},
+  {"pard",eRTFCtrl_pard},
+  {"plain",eRTFCtrl_plain},
+  {"qc",eRTFCtrl_justified},
+  {"qj",eRTFCtrl_fulljustified},
+  {"ql",eRTFCtrl_leftjustified},
+  {"qr",eRTFCtrl_rightjustified},
+  {"rdblquote",eRTFCtrl_rdoublequote},
+  {"red",eRTFCtrl_red},
+  {"rtf",eRTFCtrl_rtf},
+  {"tab",eRTFCtrl_tab},
+  {"title",eRTFCtrl_title},
+  {"u",eRTFCtrl_underline},
+  {"{",eRTFCtrl_startgroup},
+  {"}",eRTFCtrl_endgroup},
+  {"~",eRTFCtrl_last} //make sure this stays the last token...
+};
+
+/**
+ * 
+ * @update	gess4/25/98
+ * @param 
+ * @return
+ */
+const char* GetTagName(eRTFTags aTag) {
+  const    char* result=0;
+  PRInt32  cnt=sizeof(gRTFTable)/sizeof(RTFEntry);
+  PRInt32  low=0; 
+  PRInt32  high=cnt-1;
+  PRInt32  middle=kNotFound;
+  
+  while(low<=high) {
+    middle=(PRInt32)(low+high)/2;
+    if(aTag==gRTFTable[middle].mTagID)   
+      return gRTFTable[middle].mName;
+    if(aTag<gRTFTable[middle].mTagID)
+      high=middle-1; 
+    else low=middle+1; 
+  }
+  return "";
+}
+
+/**
+ *  This method gets called as part of our COM-like interfaces.
+ *  Its purpose is to create an interface to parser object
+ *  of some type.
+ *  
+ *  @update   gess 4/8/98
+ *  @param    nsIID  id of object to discover
+ *  @param    aInstancePtr ptr to newly discovered interface
+ *  @return   NS_xxx result code
+ */
+nsresult CRtfDTD::QueryInterface(const nsIID& aIID, void** aInstancePtr)  
+{                                                                        
+  if (NULL == aInstancePtr) {                                            
+    return NS_ERROR_NULL_POINTER;                                        
+  }                                                                      
+
+  if(aIID.Equals(kISupportsIID))    {  //do IUnknown...
+    *aInstancePtr = (nsIDTD*)(this);                                        
+  }
+  else if(aIID.Equals(kIDTDIID)) {  //do IParser base class...
+    *aInstancePtr = (nsIDTD*)(this);                                        
+  }
+  else if(aIID.Equals(kClassIID)) {  //do this class...
+    *aInstancePtr = (CRtfDTD*)(this);                                        
+  }                 
+  else {
+    *aInstancePtr=0;
+    return NS_NOINTERFACE;
+  }
+  ((nsISupports*) *aInstancePtr)->AddRef();
+  return NS_OK;                                                        
+}
+
+/**
+ *  This method is defined in nsIParser. It is used to 
+ *  cause the COM-like construction of an nsParser.
+ *  
+ *  @update  gess 4/8/98
+ *  @param   nsIParser** ptr to newly instantiated parser
+ *  @return  NS_xxx error result
+ */
+NS_HTMLPARS nsresult NS_NewRTF_DTD(nsIDTD** aInstancePtrResult)
+{
+  CRtfDTD* it = new CRtfDTD();
+
+  if (it == 0) {
+    return NS_ERROR_OUT_OF_MEMORY;
+  }
+
+  return it->QueryInterface(kClassIID, (void **) aInstancePtrResult);
+}
+
+
+NS_IMPL_ADDREF(CRtfDTD)
+NS_IMPL_RELEASE(CRtfDTD)
+
+
+/**
+ *  Default constructor
+ *  
+ *  @update  gess 4/9/98
+ *  @param   
+ *  @return  
+ */
+CRtfDTD::CRtfDTD() : nsIDTD() {
+  NS_INIT_REFCNT();
+  mParser=0;
+  mFilename=0;
+}
+
+/**
+ *  Default destructor
+ *  
+ *  @update  gess 4/9/98
+ *  @param   
+ *  @return  
+ */
+CRtfDTD::~CRtfDTD(){
+}
+
+/**
+ * This method is called to determine if the given DTD can parse
+ * a document in a given source-type. 
+ * NOTE: Parsing always assumes that the end result will involve
+ *       storing the result in the main content model.
+ * @update	gess6/24/98
+ * @param   
+ * @return  TRUE if this DTD can satisfy the request; FALSE otherwise.
+ */
+PRBool CRtfDTD::CanParse(nsString& aContentType, PRInt32 aVersion){
+  PRBool result=aContentType.Equals(kRTFTextContentType);
+  return result;
+}
+
+
+/**
+ * 
+ * @update	gess7/7/98
+ * @param 
+ * @return
+ */
+eAutoDetectResult CRtfDTD::AutoDetectContentType(nsString& aBuffer,nsString& aType){
+  eAutoDetectResult result=eUnknownDetect;
+  if(kNotFound!=aBuffer.Find(kRTFDocHeader))
+    if(PR_TRUE==aType.Equals(kRTFTextContentType))
+      result=eValidDetect;
+  return result;
+}
+
+/**
+ * 
+ * @update	gess5/18/98
+ * @param 
+ * @return
+ */
+PRInt32 CRtfDTD::WillBuildModel(const char* aFilename){
+  PRInt32 result=0;
+  return result;
+}
+
+/**
+ * 
+ * @update	gess5/18/98
+ * @param 
+ * @return
+ */
+PRInt32 CRtfDTD::DidBuildModel(PRInt32 anErrorCode){
+  PRInt32 result=0;
+
+  return result;
+}
+
+/**
+ * 
+ *  
+ *  @update  gess 3/25/98
+ *  @param   
+ *  @return 
+ */
+void CRtfDTD::SetParser(nsIParser* aParser) {
+  mParser=(nsParser*)aParser;
+}
+
+/**
+ *  This method gets called in order to set the content
+ *  sink for this parser to dump nodes to.
+ *  
+ *  @update  gess 3/25/98
+ *  @param   nsIContentSink interface for node receiver
+ *  @return  
+ */
+nsIContentSink* CRtfDTD::SetContentSink(nsIContentSink* aSink) {
+  return 0;
+}
+
+
+/*******************************************************************
+  These methods use to be hidden in the tokenizer-delegate. 
+  That file merged with the DTD, since the separation wasn't really
+  buying us anything.
+ *******************************************************************/
+
+/**
+ *  This gets called when we've just read a '\' char. 
+ *  It means we need to read an RTF control word.
+ *  
+ *  @update gess 3/25/98
+ *  @param  aToken ptr-ref to new token we create
+ *  @return error code -- preferably kNoError
+ */
+PRInt32 CRtfDTD::ConsumeControlWord(CToken*& aToken){
+  PRInt32 result=kNoError;
+  CRTFControlWord* cw=new CRTFControlWord("");
+  if(cw){
+    CScanner* theScanner=mParser->GetScanner();
+    cw->Consume(*theScanner);
+    aToken=cw;
+  }
+  return result;
+}
+
+/**
+ *  This gets called when we've just read a '{' or '}' group char. 
+ *  It means we have just started or ended an RTF group.
+ *  
+ *  @update gess 3/25/98
+ *  @param  aToken ptr-ref to new token we create
+ *  @return error code -- preferably kNoError
+ */
+static char* keys[] = {"}","{"};
+
+PRInt32 CRtfDTD::ConsumeGroupTag(CToken*& aToken,PRBool aStartGroup){
+  PRInt32 result=kNoError;
+  CRTFGroup* cw=new CRTFGroup(keys[PR_TRUE==aStartGroup],aStartGroup);
+  if(cw){
+    CScanner* theScanner=mParser->GetScanner();
+    cw->Consume(*theScanner);
+    aToken=cw;
+  }
+  return result;
+}
+
+/**
+ *  This gets called when we've just read plain text char.
+ *  It means we need to read RTF content.
+ *  
+ *  @update gess 3/25/98
+ *  @param  aToken ptr-ref to new token we create
+ *  @return error code -- preferably kNoError
+ */
+PRInt32 CRtfDTD::ConsumeContent(PRUnichar aChar,CToken*& aToken){
+  PRInt32 result=kNoError;
+  PRUnichar buffer[2] = {0,0};
+  buffer[0]=aChar;
+  CRTFContent * cw=new CRTFContent(buffer);
+  if(cw){
+    CScanner* theScanner=mParser->GetScanner();
+    cw->Consume(*theScanner);
+    aToken=cw;
+  }
+  return result;
+}
+
+/**
+ *  This method repeatedly called by the tokenizer. 
+ *  Each time, we determine the kind of token were about to 
+ *  read, and then we call the appropriate method to handle
+ *  that token type.
+ *  
+ *  @update gess 3/25/98
+ *  @param  aChar: last char read
+ *  @param  aScanner: see nsScanner.h
+ *  @param  anErrorCode: arg that will hold error condition
+ *  @return new token or null 
+ */
+PRInt32 CRtfDTD::ConsumeToken(CToken*& aToken){
+  
+  CScanner* theScanner=mParser->GetScanner();
+
+  PRUnichar aChar;
+  PRInt32   result=theScanner->GetChar(aChar);
+
+  switch(result) {
+    case kEOF:
+      break;
+
+    case kInterrupted:
+      theScanner->RewindToMark();
+      break; 
+
+    case kNoError:
+    default:
+      switch(aChar) {
+        case kLeftBrace:
+          result=ConsumeGroupTag(aToken,PR_TRUE); break;
+
+        case kRightBrace:
+          result=ConsumeGroupTag(aToken,PR_FALSE); break;
+
+        case kBackSlash:
+          result=ConsumeControlWord(aToken); break;
+
+        case kCR:
+        case kSpace:
+        case kTab:
+        case kLF:
+          break;
+
+        default:
+          result=ConsumeContent(aChar,aToken); break;
+      } //switch
+      break; 
+  } //switch
+  if(kNoError==result)
+    result=theScanner->Eof();
+  return result;
+}
+
+/**
+ * 
+ * @update	gess5/18/98
+ * @param 
+ * @return
+ */
+void CRtfDTD::WillResumeParse(void){
+  return;
+}
+
+/**
+ * 
+ * @update	gess5/18/98
+ * @param 
+ * @return
+ */
+void CRtfDTD::WillInterruptParse(void){
+  return;
+}
+
+/**
+ * 
+ * @update	jevering6/23/98
+ * @param 
+ * @return
+ */
+void CRtfDTD::SetDTDDebug(nsIDTDDebug * aDTDDebug) {
+}
+
+/**
+ *  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 -- int tag of parent container
+ *  @param   aChild -- int tag of child container
+ *  @return  PR_TRUE if parent can contain child
+ */
+PRBool CRtfDTD::CanContain(PRInt32 aParent,PRInt32 aChild){
+  PRBool result=PR_FALSE;
+  return result;
+}
+
+/**
+ *  
+ *  @update  gess 3/25/98
+ *  @param   aToken -- token object to be put into content model
+ *  @return  0 if all is well; non-zero is an error
+ */
+PRInt32 CRtfDTD::HandleGroup(CToken* aToken){
+  PRInt32 result=kNoError;
+  return result;
+}
+
+/**
+ *  
+ *  @update  gess 3/25/98
+ *  @param   aToken -- token object to be put into content model
+ *  @return  0 if all is well; non-zero is an error
+ */
+PRInt32 CRtfDTD::HandleControlWord(CToken* aToken){
+  PRInt32 result=kNoError;
+  return result;
+}
+
+/**
+ *  
+ *  @update  gess 3/25/98
+ *  @param   aToken -- token object to be put into content model
+ *  @return  0 if all is well; non-zero is an error
+ */
+PRInt32 CRtfDTD::HandleContent(CToken* aToken){
+  PRInt32 result=kNoError;
+  return result;
+}
+
+/**
+ *  
+ *  @update  gess 3/25/98
+ *  @param   aToken -- token object to be put into content model
+ *  @return  0 if all is well; non-zero is an error
+ */
+PRInt32 CRtfDTD::HandleToken(CToken* aToken) {
+  PRInt32 result=0;
+
+  if(aToken) {
+    eRTFTokenTypes theType=eRTFTokenTypes(aToken->GetTokenType());
+    
+    switch(theType) {
+      case eRTFToken_group:
+        result=HandleGroup(aToken); break;
+
+      case eRTFToken_controlword:
+        result=HandleControlWord(aToken); break;
+
+      case eRTFToken_content:
+        result=HandleContent(aToken); break;
+
+      default:
+        break;
+      //everything else is just text or attributes of controls...
+    }
+
+  }//if
+
+  return result;
+}
+
+
+
+/***************************************************************
+  Heres's the RTFControlWord subclass...
+ ***************************************************************/
+
+CRTFControlWord::CRTFControlWord(char* aKey) : CToken(aKey), mArgument("") {
+}
+
+
+PRInt32 CRTFControlWord::GetTokenType() {
+  return eRTFToken_controlword;
+}
+
+
+PRInt32 CRTFControlWord::Consume(CScanner& aScanner){
+  PRInt32 result=aScanner.ReadWhile(mTextValue,gAlphaChars,PR_FALSE);
+  if(kNoError==result) {
+    //ok, now look for an option parameter...
+    PRUnichar ch;
+    result=aScanner.Peek(ch);
+
+    switch(ch) {
+      case '0': case '1': case '2': case '3': case '4': 
+      case '5': case '6': case '7': case '8': case '9': 
+      case kMinus:
+        result=aScanner.ReadWhile(mArgument,gDigits,PR_FALSE);
+        break;
+
+      case kSpace:
+      default:
+        break;
+    }
+  }
+  if(kNoError==result)
+    result=aScanner.SkipWhitespace();
+  return result;
+}
+
+
+/***************************************************************
+  Heres's the RTFGroup subclass...
+ ***************************************************************/
+
+CRTFGroup::CRTFGroup(char* aKey,PRBool aStartGroup) : CToken(aKey) {
+  mStart=aStartGroup;
+}
+
+
+PRInt32 CRTFGroup::GetTokenType() {
+  return eRTFToken_group;
+}
+
+void CRTFGroup::SetGroupStart(PRBool aFlag){
+  mStart=aFlag;
+}
+
+PRBool CRTFGroup::IsGroupStart(){
+  return mStart;
+}
+
+PRInt32 CRTFGroup::Consume(CScanner& aScanner){
+  PRInt32 result=kNoError;
+  if(PR_FALSE==mStart)
+    result=aScanner.SkipWhitespace();
+  return result;
+}
+
+
+/***************************************************************
+  Heres's the RTFContent subclass...
+ ***************************************************************/
+
+CRTFContent::CRTFContent(PRUnichar* aKey) : CToken(aKey) {
+}
+
+
+PRInt32 CRTFContent::GetTokenType() {
+  return eRTFToken_content;
+}
+
+
+
+/**
+ * We're supposed to read text until we encounter one
+ * of the RTF control characters: \.{,}.
+ * @update	gess7/9/98
+ * @param 
+ * @return
+ */
+static nsString textTerminators("\\{}");
+PRInt32 CRTFContent::Consume(CScanner& aScanner){
+  PRInt32 result=aScanner.ReadUntil(mTextValue,textTerminators,PR_FALSE);
+  return result;
+}
+
+
+

mozilla/parser/htmlparser/src/CRtfDTD.h

@@ -0,0 +1,339 @@
+
+/* -*- Mode: C++; tab-width: 2; indent-tabs-mode: nil; c-basic-offset: 2 -*- */
+/*
+ * The contents of this file are subject to the Netscape Public License
+ * Version 1.0 (the "NPL"); you may not use this file except in
+ * compliance with the NPL.  You may obtain a copy of the NPL at
+ * http://www.mozilla.org/NPL/
+ *
+ * Software distributed under the NPL is distributed on an "AS IS" basis,
+ * WITHOUT WARRANTY OF ANY KIND, either express or implied. See the NPL
+ * for the specific language governing rights and limitations under the
+ * NPL.
+ *
+ * The Initial Developer of this code under the NPL is Netscape
+ * Communications Corporation.  Portions created by Netscape are
+ * Copyright (C) 1998 Netscape Communications Corporation.  All Rights
+ * Reserved.
+ */
+
+/**
+ * MODULE NOTES:
+ * @update  gess 4/8/98
+ * 
+ *         
+ */
+
+#ifndef NS_NAVHTMLDTD__
+#define NS_NAVHTMLDTD__
+
+#include "nsIDTD.h"
+#include "nsISupports.h"
+#include "nsHTMLTokens.h"
+#include "nshtmlpars.h"
+#include "nsVoidArray.h"
+#include "nsDeque.h"
+
+#define NS_RTF_DTD_IID      \
+  {0xa39c6bfc, 0x15f0,  0x11d2,  \
+  {0x80, 0x41, 0x00,    0x10, 0x4b, 0x98, 0x3f, 0xd4}}
+
+class nsIParserNode;
+class nsParser;
+
+
+enum eRTFTokenTypes {
+  eRTFToken_unknown=0,
+  eRTFToken_group,        // '{'
+  eRTFToken_controlword,  // '\'
+  eRTFToken_content,      // contains all document content
+    //everything else is just text or attributes of controls...
+  eRTFToken_last //make sure this stays the last token...
+};
+
+
+enum eRTFTags {
+  eRTFCtrl_unknown=0,
+  eRTFCtrl_quote,
+  eRTFCtrl_star,   
+  eRTFCtrl_linefeed,
+  eRTFCtrl_return,
+  eRTFCtrl_begincontrol,
+  eRTFCtrl_bold, 
+  eRTFCtrl_bin,
+  eRTFCtrl_blue,
+  eRTFCtrl_cols,
+  eRTFCtrl_comment,
+  eRTFCtrl_italic,
+  eRTFCtrl_font,
+  eRTFCtrl_fonttable,
+  eRTFCtrl_green,
+  eRTFCtrl_bottommargin,
+  eRTFCtrl_leftmargin,
+  eRTFCtrl_rightmargin,
+  eRTFCtrl_topmargin,
+  eRTFCtrl_par,
+  eRTFCtrl_pard,
+  eRTFCtrl_plain,
+  eRTFCtrl_justified,
+  eRTFCtrl_fulljustified,
+  eRTFCtrl_leftjustified,
+  eRTFCtrl_rightjustified,
+  eRTFCtrl_rdoublequote,
+  eRTFCtrl_red,
+  eRTFCtrl_rtf,
+  eRTFCtrl_tab,
+  eRTFCtrl_title,
+  eRTFCtrl_underline,
+  eRTFCtrl_startgroup,
+  eRTFCtrl_endgroup,
+  eRTFCtrl_last //make sure this stays the last token...
+};
+
+
+
+/**
+ * 
+ * @update	gess7/8/98
+ * @param 
+ * @return
+ */
+class CRTFControlWord : public CToken {
+public:
+                  CRTFControlWord(char* aKey);
+  virtual PRInt32 GetTokenType();
+  virtual PRInt32 Consume(CScanner& aScanner);
+protected:
+  nsString  mArgument;
+};
+
+
+/**
+ * 
+ * @update	gess7/8/98
+ * @param 
+ * @return
+ */
+class CRTFGroup: public CToken {
+public:
+                  CRTFGroup(char* aKey,PRBool aStartGroup);
+  virtual PRInt32 GetTokenType();
+  virtual void    SetGroupStart(PRBool aFlag);
+  virtual PRBool  IsGroupStart();
+  virtual PRInt32 Consume(CScanner& aScanner);
+protected:
+          PRBool  mStart;
+};
+
+
+/**
+ * 
+ * @update	gess7/8/98
+ * @param 
+ * @return
+ */
+class CRTFContent: public CToken {
+public:
+                  CRTFContent(PRUnichar* aValue);
+  virtual PRInt32 GetTokenType();
+  virtual PRInt32 Consume(CScanner& aScanner);
+};
+
+
+class CRtfDTD : public nsIDTD {
+            
+  public:
+
+    NS_DECL_ISUPPORTS
+
+
+    /**
+     *  
+     *  
+     *  @update  gess 4/9/98
+     *  @param   
+     *  @return  
+     */
+    CRtfDTD();
+
+    /**
+     *  
+     *  
+     *  @update  gess 4/9/98
+     *  @param   
+     *  @return  
+     */
+    virtual ~CRtfDTD();
+
+    /**
+     * This method is called to determine if the given DTD can parse
+     * a document in a given source-type. 
+     * NOTE: Parsing always assumes that the end result will involve
+     *       storing the result in the main content model.
+     * @update	gess6/24/98
+     * @param   
+     * @return  TRUE if this DTD can satisfy the request; FALSE otherwise.
+     */
+    virtual PRBool CanParse(nsString& aContentType, PRInt32 aVersion);
+
+    /**
+     * 
+     * @update	gess7/7/98
+     * @param 
+     * @return
+     */
+    virtual eAutoDetectResult AutoDetectContentType(nsString& aBuffer,nsString& aType);
+
+    /**
+     * 
+     * @update	gess5/18/98
+     * @param 
+     * @return
+     */
+    virtual PRInt32 WillBuildModel(const char* aFilename=0);
+
+    /**
+     * 
+     * @update	gess5/18/98
+     * @param 
+     * @return
+     */
+    virtual PRInt32 DidBuildModel(PRInt32 anErrorCode);
+
+    /**
+     *  
+     *  @update  gess 3/25/98
+     *  @param   aToken -- token object to be put into content model
+     *  @return  0 if all is well; non-zero is an error
+     */
+    virtual PRInt32 HandleGroup(CToken* aToken);
+
+    /**
+     *  
+     *  @update  gess 3/25/98
+     *  @param   aToken -- token object to be put into content model
+     *  @return  0 if all is well; non-zero is an error
+     */
+    virtual PRInt32 HandleControlWord(CToken* aToken);
+
+    /**
+     *  
+     *  @update  gess 3/25/98
+     *  @param   aToken -- token object to be put into content model
+     *  @return  0 if all is well; non-zero is an error
+     */
+    virtual PRInt32 HandleContent(CToken* aToken);
+
+    /**
+     *  
+     *  @update  gess 3/25/98
+     *  @param   aToken -- token object to be put into content model
+     *  @return  0 if all is well; non-zero is an error
+     */
+    virtual PRInt32 HandleToken(CToken* aToken);
+
+    /**
+     * 
+     *  
+     *  @update  gess 3/25/98
+     *  @param   
+     *  @return 
+     */
+    virtual void SetParser(nsIParser* aParser);
+
+    /**
+     *  Cause the tokenizer to consume the next token, and 
+     *  return an error result.
+     *  
+     *  @update  gess 3/25/98
+     *  @param   anError -- ref to error code
+     *  @return  new token or null
+     */
+    virtual PRInt32 ConsumeControlWord(CToken*& aToken);
+
+    /**
+     *  Cause the tokenizer to consume the next token, and 
+     *  return an error result.
+     *  
+     *  @update  gess 3/25/98
+     *  @param   anError -- ref to error code
+     *  @return  new token or null
+     */
+    virtual PRInt32 ConsumeGroupTag(CToken*& aToken,PRBool aStartTag);
+
+    /**
+     *  Cause the tokenizer to consume the next token, and 
+     *  return an error result.
+     *  
+     *  @update  gess 3/25/98
+     *  @param   anError -- ref to error code
+     *  @return  new token or null
+     */
+    virtual PRInt32 ConsumeContent(PRUnichar aChar,CToken*& aToken);
+
+    /**
+     *  Cause the tokenizer to consume the next token, and 
+     *  return an error result.
+     *  
+     *  @update  gess 3/25/98
+     *  @param   anError -- ref to error code
+     *  @return  new token or null
+     */
+    virtual PRInt32 ConsumeToken(CToken*& aToken);
+
+
+    /**
+     * 
+     * @update	gess5/18/98
+     * @param 
+     * @return
+     */
+    virtual void WillResumeParse(void);
+
+    /**
+     * 
+     * @update	gess5/18/98
+     * @param 
+     * @return
+     */
+    virtual void WillInterruptParse(void);
+
+   /**
+     * Select given content sink into parser for parser output
+     * @update	gess5/11/98
+     * @param   aSink is the new sink to be used by parser
+     * @return  old sink, or NULL
+     */
+    virtual nsIContentSink* SetContentSink(nsIContentSink* aSink);
+
+    /**
+     * 
+     * @update	jevering6/23/98
+     * @param 
+     * @return
+     */
+	  virtual void SetDTDDebug(nsIDTDDebug * aDTDDebug);
+
+    /**
+     *  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 -- int tag of parent container
+     *  @param   aChild -- int tag of child container
+     *  @return  PR_TRUE if parent can contain child
+     */
+    virtual PRBool CanContain(PRInt32 aParent,PRInt32 aChild);
+	
+protected:
+    
+    nsParser*           mParser;
+    char*               mFilename;
+};
+
+extern NS_HTMLPARS nsresult NS_NewRTF_DTD(nsIDTD** aInstancePtrResult);
+
+#endif 
+
+
+

mozilla/parser/htmlparser/src/MANIFEST

@@ -5,6 +5,7 @@
 nshtmlpars.h
 nsIContentSink.h
 nsIHTMLContentSink.h
+nsHTMLContentSinkStream.h
 nsHTMLTokens.h
 nsIParserNode.h
 nsIParser.h
@@ -12,3 +13,4 @@ nsToken.h
 nsIDTDDebug.h
 nsIDTD.h
 CNavDTD.h
+CRTFDTD.h

mozilla/parser/htmlparser/src/Makefile

@@ -30,16 +30,21 @@ CPPSRCS =     \
   nsTokenHandler.cpp \
   CNavDTD.cpp \
   COtherDTD.cpp \
+  CRtfDTD.cpp \
+  nsValidDTD.cpp \
+  nsWellFormedDTD.cpp \
   nsParser.cpp  \
   nsHTMLTokens.cpp  \
   nsHTMLTags.cpp \
   prstrm.cpp \
+  nsHTMLContentSinkStream.cpp \
   $(NULL)
 
 EXPORTS =     \
   nshtmlpars.h    \
   nsIContentSink.h  \
   nsIHTMLContentSink.h  \
+  nsHTMLContentSinkStream.h \
   nsHTMLTokens.h    \
   nsIParserNode.h   \
   nsIParser.h   \
@@ -48,6 +53,9 @@ EXPORTS =     \
   nsToken.h   \
   CNavDTD.h \
   COtherDTD.h \
+  nsWellFormedDTD.h \
+  nsValidDTD.h \
+  CRTFDTD.h \
   $(NULL)
 
 EXTRA_EXPORTS =					\

mozilla/parser/htmlparser/src/makefile.win

@@ -26,6 +26,9 @@ REQUIRES=xpcom raptor netlib
 CPPSRCS=					\
 	CNavDTD.obj				\
 	COtherDTD.obj				\
+	CRTFDTD.obj				\
+	nsWellFormedDTD.obj	 \
+	nsValidDTD.obj	 \
 	nsDTDDebug.obj				\
 	nsHTMLContentSink.obj			\
 	nsHTMLTags.obj				\
@@ -35,12 +38,16 @@ CPPSRCS=					\
 	nsScanner.obj				\
 	nsToken.obj				\
 	nsTokenHandler.obj			\
+	nsHTMLContentSinkStream.obj			\
 	prstrm.obj				\
 	$(NULL)
 
 CPP_OBJS=					\
 	.\$(OBJDIR)\CNavDTD.obj			\
+	.\$(OBJDIR)\CRTFDTD.obj			\
 	.\$(OBJDIR)\COtherDTD.obj		\
+	.\$(OBJDIR)\nsWellFormedDTD.obj		\
+	.\$(OBJDIR)\nsValidDTD.obj		\
 	.\$(OBJDIR)\nsDTDDebug.obj		\
 	.\$(OBJDIR)\nsHTMLContentSink.obj	\
 	.\$(OBJDIR)\nsHTMLTags.obj		\
@@ -50,6 +57,7 @@ CPP_OBJS=					\
 	.\$(OBJDIR)\nsScanner.obj		\
 	.\$(OBJDIR)\nsToken.obj			\
 	.\$(OBJDIR)\nsTokenHandler.obj		\
+	.\$(OBJDIR)\nsHTMLContentSinkStream.obj		\
 	.\$(OBJDIR)\prstrm.obj			\
 	$(NULL)
 
@@ -57,6 +65,7 @@ EXPORTS=					\
 	nshtmlpars.h				\
 	nsIContentSink.h			\
 	nsIHTMLContentSink.h			\
+	nsHTMLContentSinkStream.h			\
 	nsHTMLTokens.h				\
 	nsIParserNode.h				\
 	nsIParser.h				\
@@ -64,6 +73,9 @@ EXPORTS=					\
 	nsIDTDDebug.h				\
 	nsIParserFilter.h			\
 	CNavDTD.h				\
+	nsWellFormedDTD.h	\
+	nsValidDTD.h	\
+	CRTFDTD.h				\
 	COtherDTD.h				\
 	nsIDTD.h				\
 	$(NULL)

mozilla/parser/htmlparser/src/nsHTMLContentSinkStream.cpp

@@ -0,0 +1,646 @@
+/* -*- Mode: C++; tab-width: 2; indent-tabs-mode: nil; c-basic-offset: 2 -*- */
+/*
+ * The contents of this file are subject to the Netscape Public License
+ * Version 1.0 (the "NPL"); you may not use this file except in
+ * compliance with the NPL.  You may obtain a copy of the NPL at
+ * http://www.mozilla.org/NPL/
+ *
+ * Software distributed under the NPL is distributed on an "AS IS" basis,
+ * WITHOUT WARRANTY OF ANY KIND, either express or implied. See the NPL
+ * for the specific language governing rights and limitations under the
+ * NPL.
+ *
+ * The Initial Developer of this code under the NPL is Netscape
+ * Communications Corporation.  Portions created by Netscape are
+ * Copyright (C) 1998 Netscape Communications Corporation.  All Rights
+ * Reserved.
+ */
+
+/**
+ * MODULE NOTES:
+ * @update  gess 4/1/98
+ * 
+ * This file declares the concrete HTMLContentSink class.
+ * This class is used during the parsing process as the
+ * primary interface between the parser and the content
+ * model.
+ */
+
+
+#include "nsHTMLContentSinkStream.h"
+#include "nsHTMLTokens.h"
+#include <iostream.h>
+#include "nsstring.h"
+#include "nsParserTypes.h"
+
+static NS_DEFINE_IID(kISupportsIID,   NS_ISUPPORTS_IID);                 
+static NS_DEFINE_IID(kIContentSinkIID,NS_IHTMLCONTENTSINK_IID);
+static NS_DEFINE_IID(kClassIID,       NS_HTMLCONTENTSINK_STREAM_IID); 
+
+
+static char*    gHeaderComment = "<!-- This page was created by the NGLayout output system. -->";
+static char*    gDocTypeHeader = "<!DOCTYPE HTML PUBLIC \"-//W3C//DTD HTML 3.2//EN\">";
+const  int      gTabSize=2;
+static char     gBuffer[500];
+
+/**
+ *  This method gets called as part of our COM-like interfaces.
+ *  Its purpose is to create an interface to parser object
+ *  of some type.
+ *  
+ *  @update   gess 4/8/98
+ *  @param    nsIID  id of object to discover
+ *  @param    aInstancePtr ptr to newly discovered interface
+ *  @return   NS_xxx result code
+ */
+nsresult CHTMLContentSinkStream::QueryInterface(const nsIID& aIID, void** aInstancePtr)  
+{                                                                        
+  if (NULL == aInstancePtr) {                                            
+    return NS_ERROR_NULL_POINTER;                                        
+  }                                                                      
+
+  if(aIID.Equals(kISupportsIID))    {  //do IUnknown...
+    *aInstancePtr = (nsIContentSink*)(this);                                        
+  }
+  else if(aIID.Equals(kIContentSinkIID)) {  //do IParser base class...
+    *aInstancePtr = (nsIContentSink*)(this);                                        
+  }
+  else if(aIID.Equals(kClassIID)) {  //do this class...
+    *aInstancePtr = (CHTMLContentSinkStream*)(this);                                        
+  }                 
+  else {
+    *aInstancePtr=0;
+    return NS_NOINTERFACE;
+  }
+  ((nsISupports*) *aInstancePtr)->AddRef();
+  return NS_OK;                                                        
+}
+
+
+NS_IMPL_ADDREF(CHTMLContentSinkStream)
+NS_IMPL_RELEASE(CHTMLContentSinkStream)
+
+
+/**
+ *  This method is defined in nsIParser. It is used to 
+ *  cause the COM-like construction of an nsParser.
+ *  
+ *  @update  gess 4/8/98
+ *  @param   nsIParser** ptr to newly instantiated parser
+ *  @return  NS_xxx error result
+ */
+NS_HTMLPARS nsresult NS_New_HTML_ContentSinkStream(CHTMLContentSinkStream** aInstancePtrResult) {
+  CHTMLContentSinkStream* it = new CHTMLContentSinkStream();
+
+  if (it == 0) {
+    return NS_ERROR_OUT_OF_MEMORY;
+  }
+
+  return it->QueryInterface(kClassIID, (void **) aInstancePtrResult);
+}
+
+
+/**
+ * Construct a content sink stream.
+ * @update	gess7/7/98
+ * @param 
+ * @return
+ */
+CHTMLContentSinkStream::CHTMLContentSinkStream()  {
+  mOutput=&cout;
+}
+
+/**
+ * Construct a content sink stream.
+ * @update	gess7/7/98
+ * @param 
+ * @return
+ */
+CHTMLContentSinkStream::CHTMLContentSinkStream(ostream& aStream)  {
+  mOutput=&aStream;
+}
+
+
+/**
+ * 
+ * @update	gess7/7/98
+ * @param 
+ * @return
+ */
+CHTMLContentSinkStream::~CHTMLContentSinkStream() {
+  mOutput=0;  //we don't own the stream we're given; just forget it.
+}
+
+
+/**
+ * 
+ * @update	gess7/7/98
+ * @param 
+ * @return
+ */
+void CHTMLContentSinkStream::SetOutputStream(ostream& aStream) {
+  mOutput=&aStream;
+}
+
+
+/**
+ * 
+ * @update	gess7/7/98
+ * @param 
+ * @return
+ */
+void WriteAttributes(const nsIParserNode& aNode,ostream& aStream) {
+  int theCount=aNode.GetAttributeCount();
+  if(theCount) {
+    int i=0;
+    for(i=0;i<theCount;i++){
+      const nsString& key=aNode.GetKeyAt(i);
+      key.ToCString(gBuffer,sizeof(gBuffer)-1);
+      aStream << " " << gBuffer << char(kEqual);
+      const nsString& value=aNode.GetValueAt(i);
+      value.ToCString(gBuffer,sizeof(gBuffer)-1);
+      aStream << gBuffer;
+    }
+  }
+}
+
+
+/**
+ * 
+ * @update	gess7/5/98
+ * @param 
+ * @return
+ */
+void OpenTagWithAttributes(const char* theTag,const nsIParserNode& aNode,int tab,ostream& aStream,bool aNewline) {
+  int i=0;
+  for(i=0;i<tab*gTabSize;i++) 
+    aStream << " ";
+  aStream << (char)kLessThan << theTag;
+  WriteAttributes(aNode,aStream);
+  aStream << (char)kGreaterThan;
+  if(aNewline)
+    aStream << endl;
+}
+
+
+/**
+ * 
+ * @update	gess7/5/98
+ * @param 
+ * @return
+ */
+void OpenTag(const char* theTag,int tab,ostream& aStream,bool aNewline) {
+  int i=0;
+  for(i=0;i<tab*gTabSize;i++) 
+    aStream << " ";
+  aStream << (char)kLessThan << theTag << (char)kGreaterThan;
+  if(aNewline)
+    aStream << endl;
+}
+
+
+/**
+ * 
+ * @update	gess7/5/98
+ * @param 
+ * @return
+ */
+void CloseTag(const char* theTag,int tab,ostream& aStream) {
+  int i=0;
+  for(i=0;i<tab*gTabSize;i++) 
+    aStream << " ";
+  aStream << (char)kLessThan << (char)kForwardSlash << theTag << (char)kGreaterThan << endl;
+}
+
+
+/**
+ * 
+ * @update	gess7/5/98
+ * @param 
+ * @return
+ */
+void WritePair(eHTMLTags aTag,const nsString& theContent,int tab,ostream& aStream) {
+  const char* titleStr = GetTagName(aTag);
+  OpenTag(titleStr,tab,aStream,false);
+  theContent.ToCString(gBuffer,sizeof(gBuffer)-1);
+  aStream << gBuffer;
+  CloseTag(titleStr,0,aStream);
+}
+
+
+
+/**
+ * 
+ * @update	gess7/5/98
+ * @param 
+ * @return
+ */
+void WriteSingleton(eHTMLTags aTag,const nsString& theContent,int tab,ostream& aStream) {
+  const char* titleStr = GetTagName(aTag);
+  OpenTag(titleStr,tab,aStream,false);
+  if(theContent.Length()) {
+    theContent.ToCString(gBuffer,sizeof(gBuffer)-1);
+    aStream << gBuffer;
+  }
+  aStream << endl;
+}
+
+
+/**
+  * This method gets called by the parser when it encounters
+  * a title tag and wants to set the document title in the sink.
+  *
+  * @update 4/1/98 gess
+  * @param  nsString reference to new title value
+  * @return PR_TRUE if successful. 
+  */     
+PRBool CHTMLContentSinkStream::SetTitle(const nsString& aValue){
+  PRBool result=PR_FALSE;
+  if(mOutput) {
+    WritePair(eHTMLTag_title,aValue,1+mTabLevel,*mOutput);
+  }
+  return result;
+}
+
+
+/**
+  * This method is used to open the outer HTML container.
+  *
+  * @update 4/1/98 gess
+  * @param  nsIParserNode reference to parser node interface
+  * @return PR_TRUE if successful. 
+  */     
+PRInt32 CHTMLContentSinkStream::OpenHTML(const nsIParserNode& aNode){
+  PRInt32 result=0;
+  mTabLevel++;
+  if(mOutput) {
+    const char* theTag= GetTagName(eHTMLTag_html);
+    OpenTagWithAttributes(theTag,aNode,mTabLevel,*mOutput,true);
+  }
+  return result;
+}
+
+
+/**
+  * This method is used to close the outer HTML container.
+  *
+  * @update 4/1/98 gess
+  * @param  nsIParserNode reference to parser node interface
+  * @return PR_TRUE if successful. 
+  */     
+PRInt32 CHTMLContentSinkStream::CloseHTML(const nsIParserNode& aNode){
+  PRInt32 result=0;
+  if(mOutput) {
+    const char* theTag= GetTagName(eHTMLTag_html);
+    CloseTag(theTag,0,*mOutput);
+  }
+  mTabLevel--;
+  return result;
+}
+
+
+/**
+  * This method is used to open the only HEAD container.
+  *
+  * @update 4/1/98 gess
+  * @param  nsIParserNode reference to parser node interface
+  * @return PR_TRUE if successful. 
+  */     
+PRInt32 CHTMLContentSinkStream::OpenHead(const nsIParserNode& aNode){
+  PRInt32 result=0;
+  mTabLevel++;
+  if(mOutput) {
+    const char* theTag= GetTagName(eHTMLTag_head);
+    OpenTagWithAttributes(theTag,aNode,mTabLevel,*mOutput,true);
+  }
+  return result;
+}
+
+
+/**
+  * This method is used to close the only HEAD container.
+  *
+  * @update 4/1/98 gess
+  * @param  nsIParserNode reference to parser node interface
+  * @return PR_TRUE if successful. 
+  */     
+PRInt32 CHTMLContentSinkStream::CloseHead(const nsIParserNode& aNode){
+  PRInt32 result=0;
+  if(mOutput) {
+    const char* theTag= GetTagName(eHTMLTag_head);
+    CloseTag(theTag,mTabLevel,*mOutput);
+  }
+  mTabLevel--;
+  return result;
+}
+
+
+/**
+  * This method is used to open the main BODY container.
+  *
+  * @update 4/1/98 gess
+  * @param  nsIParserNode reference to parser node interface
+  * @return PR_TRUE if successful. 
+  */     
+PRInt32 CHTMLContentSinkStream::OpenBody(const nsIParserNode& aNode){
+  PRInt32 result=0;
+  mTabLevel++;
+  if(mOutput) {
+    const char* theTag= GetTagName(eHTMLTag_body);
+    OpenTagWithAttributes(theTag,aNode,mTabLevel,*mOutput,true);
+  }
+  return result;
+}
+
+
+/**
+  * This method is used to close the main BODY container.
+  *
+  * @update 4/1/98 gess
+  * @param  nsIParserNode reference to parser node interface
+  * @return PR_TRUE if successful. 
+  */     
+PRInt32 CHTMLContentSinkStream::CloseBody(const nsIParserNode& aNode){
+  PRInt32 result=0;
+  if(mOutput) {
+    const char* theTag= GetTagName(eHTMLTag_body);
+    CloseTag(theTag,mTabLevel,*mOutput);
+  }
+  mTabLevel--;
+  return result;
+}
+
+
+/**
+  * This method is used to open a new FORM container.
+  *
+  * @update 4/1/98 gess
+  * @param  nsIParserNode reference to parser node interface
+  * @return PR_TRUE if successful. 
+  */     
+PRInt32 CHTMLContentSinkStream::OpenForm(const nsIParserNode& aNode){
+  PRInt32 result=0;
+  mTabLevel++;
+  if(mOutput) {
+    const char* theTag= GetTagName(eHTMLTag_form);
+    OpenTagWithAttributes(theTag,aNode,mTabLevel,*mOutput,true);
+  }
+  return result;
+}
+
+
+/**
+  * This method is used to close the outer FORM container.
+  *
+  * @update 4/1/98 gess
+  * @param  nsIParserNode reference to parser node interface
+  * @return PR_TRUE if successful. 
+  */     
+PRInt32 CHTMLContentSinkStream::CloseForm(const nsIParserNode& aNode){
+  PRInt32 result=0;
+  if(mOutput) {
+    const char* theTag= GetTagName(eHTMLTag_form);
+    CloseTag(theTag,mTabLevel,*mOutput);
+  }
+  mTabLevel--;
+  return result;
+}
+
+    
+/**
+  * This method is used to open the FRAMESET container.
+  *
+  * @update 4/1/98 gess
+  * @param  nsIParserNode reference to parser node interface
+  * @return PR_TRUE if successful. 
+  */     
+PRInt32 CHTMLContentSinkStream::OpenFrameset(const nsIParserNode& aNode){
+  PRInt32 result=0;
+  mTabLevel++;
+  if(mOutput) {
+    const char* theTag= GetTagName(eHTMLTag_frameset);
+    OpenTagWithAttributes(theTag,aNode,mTabLevel,*mOutput,true);
+  }
+  return result;
+}
+
+
+/**
+  * This method is used to close the FRAMESET container.
+  *
+  * @update 4/1/98 gess
+  * @param  nsIParserNode reference to parser node interface
+  * @return PR_TRUE if successful. 
+  */     
+PRInt32 CHTMLContentSinkStream::CloseFrameset(const nsIParserNode& aNode){
+  PRInt32 result=0;
+  if(mOutput) {
+    const char* theTag= GetTagName(eHTMLTag_frameset);
+    CloseTag(theTag,mTabLevel,*mOutput);
+  }
+  mTabLevel--;
+  return result;
+}
+
+    
+/**
+  * This method is used to a general container. 
+  * This includes: OL,UL,DIR,SPAN,TABLE,H[1..6],etc.
+  *
+  * @update 4/1/98 gess
+  * @param  nsIParserNode reference to parser node interface
+  * @return PR_TRUE if successful. 
+  */     
+PRInt32 CHTMLContentSinkStream::OpenContainer(const nsIParserNode& aNode){
+  PRInt32 result=0;
+  mTabLevel++;
+  if(mOutput) {
+
+    eHTMLTags nodeType=(eHTMLTags)aNode.GetNodeType();
+
+    switch(nodeType) {
+
+      case eHTMLTag_html:
+        result=OpenHTML(aNode); break;
+
+      case eHTMLTag_body:
+        result=OpenBody(aNode); break;
+
+      case eHTMLTag_title:
+        {
+          const nsString& title=aNode.GetName();
+          result=SetTitle(title); 
+        }
+        break;
+
+      case eHTMLTag_head:
+        result=OpenHead(aNode); break;
+
+      case eHTMLTag_style:
+        break;
+
+      case eHTMLTag_form:
+        result=OpenForm(aNode); break;
+
+      default:
+        //this is the hard one. You have to decode aNode manually.
+        {
+          const nsString& name=aNode.GetName();
+          const char* tagName= GetTagName(nodeType);
+          OpenTagWithAttributes(tagName,aNode,mTabLevel,*mOutput,true);
+        }
+        break;
+    }
+
+
+  }
+  return result;
+}
+
+
+/**
+  * This method is used to close a generic container.
+  *
+  * @update 4/1/98 gess
+  * @param  nsIParserNode reference to parser node interface
+  * @return PR_TRUE if successful. 
+  */     
+PRInt32 CHTMLContentSinkStream::CloseContainer(const nsIParserNode& aNode){
+  PRInt32 result=0;
+  if(mOutput) {
+
+    eHTMLTags nodeType=(eHTMLTags)aNode.GetNodeType();
+  
+    switch(nodeType) {
+
+      case eHTMLTag_html:
+        result=CloseHTML(aNode); break;
+
+      case eHTMLTag_style:
+        break;
+
+      case eHTMLTag_head:
+        result=CloseHead(aNode); break; 
+
+      case eHTMLTag_body:
+        result=CloseBody(aNode); break;
+
+      case eHTMLTag_form:
+        result=CloseForm(aNode); break;
+
+      case eHTMLTag_title:
+//        result=SetTitle(aNode); break;
+
+      default:
+        {
+          const nsString& name=aNode.GetName();
+          const char* tagName= GetTagName(nodeType);
+          //(*mOutput) << endl;
+          CloseTag(tagName,mTabLevel,*mOutput);
+        }
+        break;
+    }
+  }
+  mTabLevel--;
+  return result;
+}
+
+
+/**
+  * This method is used to add a leaf to the currently 
+  * open container.
+  *
+  * @update 4/1/98 gess
+  * @param  nsIParserNode reference to parser node interface
+  * @return PR_TRUE if successful. 
+  */     
+PRInt32 CHTMLContentSinkStream::AddLeaf(const nsIParserNode& aNode){
+  PRInt32 result=0;
+  if(mOutput) {
+
+    eHTMLTags nodeType=(eHTMLTags)aNode.GetNodeType();
+
+    switch(nodeType) {
+      case eHTMLTag_hr:
+      case eHTMLTag_br:
+        {
+          nsAutoString empty;
+          WriteSingleton(nodeType,empty,1+mTabLevel,*mOutput);
+        }
+        break;
+
+      case eHTMLTag_style:
+        {
+          const nsString& skipped=aNode.GetSkippedContent();
+          WritePair(nodeType,skipped,1+mTabLevel,*mOutput);
+        }
+        break;
+
+      case eHTMLTag_newline:
+      case eHTMLTag_whitespace:
+        break;
+
+      default:
+        {
+          int i=0;
+          for(i=0;i<(1+mTabLevel)*gTabSize;i++) 
+            (*mOutput) << " ";
+    
+          const nsString& text=aNode.GetText();
+          text.ToCString(gBuffer,sizeof(gBuffer)-1);
+          (*mOutput) << gBuffer << endl;
+        }
+    }//switch
+  }
+  return result;
+}
+
+
+/**
+  * This method gets called when the parser begins the process
+  * of building the content model via the content sink.
+  *
+  * @update 5/7/98 gess
+  */     
+void CHTMLContentSinkStream::WillBuildModel(void){
+  mTabLevel=-1;
+  if(mOutput) {
+    (*mOutput) << gHeaderComment << endl;
+    (*mOutput) << gDocTypeHeader << endl;
+  }
+}
+
+
+/**
+  * This method gets called when the parser concludes the process
+  * of building the content model via the content sink.
+  *
+  * @param  aQualityLevel describes how well formed the doc was.
+  *         0=GOOD; 1=FAIR; 2=POOR;
+  * @update 5/7/98 gess
+  */     
+void CHTMLContentSinkStream::DidBuildModel(PRInt32 aQualityLevel) {
+}
+
+
+/**
+  * This method gets called when the parser gets i/o blocked,
+  * and wants to notify the sink that it may be a while before
+  * more data is available.
+  *
+  * @update 5/7/98 gess
+  */     
+void CHTMLContentSinkStream::WillInterrupt(void) {
+}
+
+
+/**
+  * This method gets called when the parser i/o gets unblocked,
+  * and we're about to start dumping content again to the sink.
+  *
+  * @update 5/7/98 gess
+  */     
+void CHTMLContentSinkStream::WillResume(void) {
+}
+
+

mozilla/parser/htmlparser/src/nsHTMLContentSinkStream.h

@@ -0,0 +1,259 @@
+/* -*- Mode: C++; tab-width: 2; indent-tabs-mode: nil; c-basic-offset: 2 -*- */
+/*
+ * The contents of this file are subject to the Netscape Public License
+ * Version 1.0 (the "NPL"); you may not use this file except in
+ * compliance with the NPL.  You may obtain a copy of the NPL at
+ * http://www.mozilla.org/NPL/
+ *
+ * Software distributed under the NPL is distributed on an "AS IS" basis,
+ * WITHOUT WARRANTY OF ANY KIND, either express or implied. See the NPL
+ * for the specific language governing rights and limitations under the
+ * NPL.
+ *
+ * The Initial Developer of this code under the NPL is Netscape
+ * Communications Corporation.  Portions created by Netscape are
+ * Copyright (C) 1998 Netscape Communications Corporation.  All Rights
+ * Reserved.
+ */
+
+/**
+ * MODULE NOTES:
+ * @update  gess 4/1/98
+ * 
+ * This file declares the concrete HTMLContentSink class.
+ * This class is used during the parsing process as the
+ * primary interface between the parser and the content
+ * model.
+ */
+
+#ifndef  NS_HTMLCONTENTSINK_STREAM
+#define  NS_ITMLCONTENTSINK_STREAM
+
+#include "nsIParserNode.h"
+#include "nsIHTMLContentSink.h"
+#include "nshtmlpars.h"
+
+#define NS_HTMLCONTENTSINK_STREAM_IID  \
+  {0xa39c6bff, 0x15f0, 0x11d2, \
+  {0x80, 0x41, 0x0, 0x10, 0x4b, 0x98, 0x3f, 0xd4}}
+
+
+
+class ostream;
+
+class CHTMLContentSinkStream : public nsIHTMLContentSink {
+  public:
+
+  NS_DECL_ISUPPORTS
+
+  /**
+   * 
+   * @update	gess7/7/98
+   * @param 
+   * @return
+   */
+  CHTMLContentSinkStream(); 
+
+  /**
+   * 
+   * @update	gess7/7/98
+   * @param 
+   * @return
+   */
+  CHTMLContentSinkStream(ostream& aStream); 
+
+  /**
+   * 
+   * @update	gess7/7/98
+   * @param 
+   * @return
+   */
+  virtual ~CHTMLContentSinkStream();
+
+
+  /**
+   * 
+   * @update	gess7/7/98
+   * @param 
+   * @return
+   */
+  void SetOutputStream(ostream& aStream);
+
+   /**
+    * This method gets called by the parser when it encounters
+    * a title tag and wants to set the document title in the sink.
+    *
+    * @update 4/1/98 gess
+    * @param  nsString reference to new title value
+    * @return PR_TRUE if successful. 
+    */     
+  virtual PRBool SetTitle(const nsString& aValue);
+
+   /**
+    * This method is used to open the outer HTML container.
+    *
+    * @update 4/1/98 gess
+    * @param  nsIParserNode reference to parser node interface
+    * @return PR_TRUE if successful. 
+    */     
+    virtual PRInt32 OpenHTML(const nsIParserNode& aNode);
+
+   /**
+    * This method is used to close the outer HTML container.
+    *
+    * @update 4/1/98 gess
+    * @param  nsIParserNode reference to parser node interface
+    * @return PR_TRUE if successful. 
+    */     
+    virtual PRInt32 CloseHTML(const nsIParserNode& aNode);
+
+   /**
+    * This method is used to open the only HEAD container.
+    *
+    * @update 4/1/98 gess
+    * @param  nsIParserNode reference to parser node interface
+    * @return PR_TRUE if successful. 
+    */     
+    virtual PRInt32 OpenHead(const nsIParserNode& aNode);
+
+   /**
+    * This method is used to close the only HEAD container.
+    *
+    * @update 4/1/98 gess
+    * @param  nsIParserNode reference to parser node interface
+    * @return PR_TRUE if successful. 
+    */     
+    virtual PRInt32 CloseHead(const nsIParserNode& aNode);
+  
+   /**
+    * This method is used to open the main BODY container.
+    *
+    * @update 4/1/98 gess
+    * @param  nsIParserNode reference to parser node interface
+    * @return PR_TRUE if successful. 
+    */     
+    virtual PRInt32 OpenBody(const nsIParserNode& aNode);
+
+   /**
+    * This method is used to close the main BODY container.
+    *
+    * @update 4/1/98 gess
+    * @param  nsIParserNode reference to parser node interface
+    * @return PR_TRUE if successful. 
+    */     
+    virtual PRInt32 CloseBody(const nsIParserNode& aNode);
+
+   /**
+    * This method is used to open a new FORM container.
+    *
+    * @update 4/1/98 gess
+    * @param  nsIParserNode reference to parser node interface
+    * @return PR_TRUE if successful. 
+    */     
+    virtual PRInt32 OpenForm(const nsIParserNode& aNode);
+
+   /**
+    * This method is used to close the outer FORM container.
+    *
+    * @update 4/1/98 gess
+    * @param  nsIParserNode reference to parser node interface
+    * @return PR_TRUE if successful. 
+    */     
+    virtual PRInt32 CloseForm(const nsIParserNode& aNode);
+        
+   /**
+    * This method is used to open the FRAMESET container.
+    *
+    * @update 4/1/98 gess
+    * @param  nsIParserNode reference to parser node interface
+    * @return PR_TRUE if successful. 
+    */     
+    virtual PRInt32 OpenFrameset(const nsIParserNode& aNode);
+
+   /**
+    * This method is used to close the FRAMESET container.
+    *
+    * @update 4/1/98 gess
+    * @param  nsIParserNode reference to parser node interface
+    * @return PR_TRUE if successful. 
+    */     
+    virtual PRInt32 CloseFrameset(const nsIParserNode& aNode);
+        
+   /**
+    * This method is used to a general container. 
+    * This includes: OL,UL,DIR,SPAN,TABLE,H[1..6],etc.
+    *
+    * @update 4/1/98 gess
+    * @param  nsIParserNode reference to parser node interface
+    * @return PR_TRUE if successful. 
+    */     
+    virtual PRInt32 OpenContainer(const nsIParserNode& aNode);
+    
+   /**
+    * This method is used to close a generic container.
+    *
+    * @update 4/1/98 gess
+    * @param  nsIParserNode reference to parser node interface
+    * @return PR_TRUE if successful. 
+    */     
+    virtual PRInt32 CloseContainer(const nsIParserNode& aNode);
+
+   /**
+    * This method is used to add a leaf to the currently 
+    * open container.
+    *
+    * @update 4/1/98 gess
+    * @param  nsIParserNode reference to parser node interface
+    * @return PR_TRUE if successful. 
+    */     
+    virtual PRInt32 AddLeaf(const nsIParserNode& aNode);
+
+   /**
+    * This method gets called when the parser begins the process
+    * of building the content model via the content sink.
+    *
+    * @update 5/7/98 gess
+    */     
+    virtual void WillBuildModel(void);
+
+   /**
+    * This method gets called when the parser concludes the process
+    * of building the content model via the content sink.
+    *
+    * @param  aQualityLevel describes how well formed the doc was.
+    *         0=GOOD; 1=FAIR; 2=POOR;
+    * @update 5/7/98 gess
+    */     
+    virtual void DidBuildModel(PRInt32 aQualityLevel);
+
+   /**
+    * This method gets called when the parser gets i/o blocked,
+    * and wants to notify the sink that it may be a while before
+    * more data is available.
+    *
+    * @update 5/7/98 gess
+    */     
+    virtual void WillInterrupt(void);
+
+   /**
+    * This method gets called when the parser i/o gets unblocked,
+    * and we're about to start dumping content again to the sink.
+    *
+    * @update 5/7/98 gess
+    */     
+    virtual void WillResume(void);
+
+protected:
+    ostream*  mOutput;
+    int       mTabLevel;
+
+};
+
+extern NS_HTMLPARS nsresult NS_New_HTML_ContentSinkStream(CHTMLContentSinkStream** aInstancePtrResult);
+
+
+#endif
+
+
+
+

mozilla/parser/htmlparser/src/nsHTMLTokens.cpp

@@ -247,7 +247,7 @@ PRInt32 CStartToken::Consume(PRUnichar aChar, CScanner& aScanner) {
    //and see if the next char is ">". If so, we have a complete
    //tag without attributes.
   if(kNoError==result) {  
-    result=aScanner.SkipWhite();
+    result=aScanner.SkipWhitespace();
     if(kNoError==result) {
       result=aScanner.GetChar(aChar);
       if(kNoError==result) {
@@ -678,7 +678,7 @@ PRInt32 ConsumeAttributeValueText(PRUnichar aChar,nsString& aString,CScanner& aS
  */
 PRInt32 CAttributeToken::Consume(PRUnichar aChar, CScanner& aScanner) {
 
-  aScanner.SkipWhite();             //skip leading whitespace                                      
+  aScanner.SkipWhitespace();             //skip leading whitespace                                      
   PRInt32 result=aScanner.Peek(aChar);
   if(kNoError==result) {
     if(kQuote==aChar) {               //if you're here, handle quoted key...
@@ -703,12 +703,12 @@ PRInt32 CAttributeToken::Consume(PRUnichar aChar, CScanner& aScanner) {
     }
 
       //now it's time to Consume the (optional) value...
-    if(!(result=aScanner.SkipWhite())) { 
+    if(!(result=aScanner.SkipWhitespace())) { 
       if(!(result=aScanner.Peek(aChar))) {
         if(kEqual==aChar){
           result=aScanner.GetChar(aChar);  //skip the equal sign...
           if(kNoError==result) {
-            result=aScanner.SkipWhite();     //now skip any intervening whitespace
+            result=aScanner.SkipWhitespace();     //now skip any intervening whitespace
             if(kNoError==result) {
               result=aScanner.GetChar(aChar);  //and grab the next char.    
               if(kNoError==result) {
@@ -722,7 +722,7 @@ PRInt32 CAttributeToken::Consume(PRUnichar aChar, CScanner& aScanner) {
                 } 
               }//if
               if(kNoError==result)
-                result=aScanner.SkipWhite();     
+                result=aScanner.SkipWhitespace();     
             }//if
           }//if
         }//if

mozilla/parser/htmlparser/src/nsIContentSink.h

@@ -60,15 +60,6 @@ class nsIContentSink : public nsISupports {
     */     
   virtual PRInt32 CloseContainer(const nsIParserNode& aNode) = 0;
 
-   /**
-    * This method gets called by the parser when a the
-    * topmost container in the content sink needs to be closed.
-    *
-    * @update 4/1/98 gess
-    * @return PR_TRUE if successful. 
-    */     
-  virtual PRInt32 CloseTopmostContainer() = 0;
-
    /**
     * This gets called by the parser when you want to add
     * a leaf node to the current container in the content

mozilla/parser/htmlparser/src/nsIDTD.h

@@ -41,7 +41,6 @@ class nsIDTDDebug;
 class nsIURL;
 class nsString;
 
-enum eProcessType {eParsing, eConverting};
 enum eAutoDetectResult {eUnknownDetect, eValidDetect, eInvalidDetect};
 
 class nsIDTD : public nsISupports {
@@ -76,12 +75,23 @@ class nsIDTD : public nsISupports {
     virtual nsIContentSink* SetContentSink(nsIContentSink* aSink)=0;
 
     /**
-     * 
+     * This method is called to determine if the given DTD can parse
+     * a document in a given source-type. 
+     * NOTE: Parsing always assumes that the end result will involve
+     *       storing the result in the main content model.
      * @update	gess6/24/98
+     * @param   
+     * @return  TRUE if this DTD can satisfy the request; FALSE otherwise.
+     */
+    virtual PRBool CanParse(nsString& aContentType, PRInt32 aVersion)=0;
+
+    /**
+     * 
+     * @update	gess7/7/98
      * @param 
      * @return
      */
-    virtual PRBool IsCapableOf(eProcessType aProcessType, nsString& aContentType, PRInt32 aVersion)=0;
+    virtual eAutoDetectResult AutoDetectContentType(nsString& aBuffer,nsString& aType)=0;
 
     /**
      * 
@@ -136,13 +146,16 @@ class nsIDTD : public nsISupports {
     virtual void WillInterruptParse(void)=0;
 
     /**
-     * 
-     * @update	jevering 6/18/98
-     * @param  aParent  parent tag
-     * @param  aChild   child tag
-     * @return PR_TRUE if valid container
+     *  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 -- int tag of parent container
+     *  @param   aChild -- int tag of child container
+     *  @return  PR_TRUE if parent can contain child
      */
-    virtual PRBool CanContain(PRInt32 aParent, PRInt32 aChild) = 0;
+    virtual PRBool CanContain(PRInt32 aParent,PRInt32 aChild)=0;
+
 
     /**
      * 

mozilla/parser/htmlparser/src/nsIHTMLContentSink.h

@@ -62,7 +62,7 @@
  */
 
 #ifndef  NS_IHTMLCONTENTSINK
-#define  NS_ITMLCONTENTSINK
+#define  NS_IHTMLCONTENTSINK
 
 #include "nsIParserNode.h"
 #include "nsIContentSink.h"
@@ -194,16 +194,6 @@ class nsIHTMLContentSink : public nsIContentSink {
     */     
     virtual PRInt32 CloseContainer(const nsIParserNode& aNode)=0;
 
-   /**
-    * This method is used to close the topmost container, regardless
-    * of the type.
-    *
-    * @update 4/1/98 gess
-    * @param  nsIParserNode reference to parser node interface
-    * @return PR_TRUE if successful. 
-    */     
-    virtual PRInt32 CloseTopmostContainer()=0;
-
    /**
     * This method is used to add a leaf to the currently 
     * open container.

mozilla/parser/htmlparser/src/nsIParser.h

@@ -35,6 +35,7 @@ class nsString;
 class CToken;
 class nsIURL;
 class nsIDTDDebug;
+class fstream;
 
 /**
  *  This class defines the iparser interface. This XPCOM
@@ -49,7 +50,7 @@ class nsIParser : public nsISupports {
 
     virtual nsIContentSink* SetContentSink(nsIContentSink* aContentSink)=0;
 
-    virtual eAutoDetectResult AutoDetectContentType(nsString& aBuffer)=0;
+    virtual eAutoDetectResult AutoDetectContentType(nsString& aBuffer,nsString& aType)=0;
 
     /**
      *  Cause the tokenizer to consume the next token, and 
@@ -61,26 +62,15 @@ class nsIParser : public nsISupports {
      */
     virtual PRInt32 ConsumeToken(CToken*& aToken)=0;
 
-    virtual PRInt32 BeginParse(nsIURL* aURL,
-                               nsIStreamObserver* aListener = nsnull,
-                               nsIDTDDebug * aDTDDebug = 0) = 0;
-
     /******************************************************************************************
      *  Parse methods always begin with an input source, and perform conversions 
      *  until you wind up with HTML in your actual content model.
      ******************************************************************************************/
     virtual PRInt32 Parse(nsIURL* aURL,nsIStreamObserver* aListener = nsnull,nsIDTDDebug * aDTDDebug = 0) = 0;
     virtual PRInt32 Parse(const char* aFilename)=0;
+    virtual PRInt32 Parse(fstream& aStream)=0;
     virtual PRInt32 Parse(nsString& anHTMLString,PRBool appendTokens)=0;
 
-    /******************************************************************************************
-     *  Convert methods start input source (of known or unknown form), and perform conversions 
-     *  until you wind up with a <i>stream</i> in your target form.
-     *  The internal content model is never effected.
-     ******************************************************************************************/
-    virtual PRInt32 Convert(nsIURL* aURL,char* aSourceForm,char* aTargetForm,nsIStreamListener* aListener) = 0;
-    virtual PRInt32 Convert(const char* aFilename,char* aSourceForm,char* aTargetForm)=0;
-    virtual PRInt32 Convert(nsString& anHTMLString,char* aSourceForm,char* aTargetForm,PRBool appendTokens)=0;
 
     virtual PRInt32 ResumeParse(void)=0;
 

mozilla/parser/htmlparser/src/nsParser.cpp

@@ -31,6 +31,13 @@
 #include "nsIDTDDebug.h"
 #include "nshtmlpars.h"
 
+#define rickgdebug 1
+#ifdef  rickgdebug
+#include "CRtfDTD.h"
+#include "nsWellFormedDTD.h"
+#endif
+
+
 static NS_DEFINE_IID(kISupportsIID, NS_ISUPPORTS_IID);                 
 static NS_DEFINE_IID(kClassIID, NS_PARSER_IID); 
 static NS_DEFINE_IID(kIParserIID, NS_IPARSER_IID);
@@ -136,7 +143,8 @@ CSharedParserObjects gSharedParserObjects;
  */
 nsParser::nsParser() : 
   mTokenDeque(gTokenDeallocator),
-  mContentType()  
+  mSourceType(),
+  mTargetType()
 {
   NS_INIT_REFCNT();
   mDTDDebug = 0;
@@ -270,6 +278,15 @@ nsIContentSink* nsParser::SetContentSink(nsIContentSink* aSink) {
  *  @return  nothing.
  */
 void nsParser::RegisterDTD(nsIDTD* aDTD){
+
+#ifdef rickgdebug
+  nsIDTD* rv=0;
+  NS_NewRTF_DTD(&rv);
+  gSharedParserObjects.RegisterDTD(rv);
+  NS_NewWellFormed_DTD(&rv);
+  gSharedParserObjects.RegisterDTD(rv);
+#endif
+
   gSharedParserObjects.RegisterDTD(aDTD);
 }
 
@@ -309,28 +326,31 @@ eParseMode DetermineParseMode() {
  *  @param   
  *  @return  
  */
-PRBool FindSuitableDTD(eProcessType aProcessType,eParseMode aMode,nsString& aContentType, nsIDTD*& aDefaultDTD) {
- PRBool result=PR_FALSE;
+PRBool FindSuitableDTD( eParseMode aMode,
+                        nsString& aSourceType, 
+                        nsString& aTargetType, 
+                        nsIDTD*& aDefaultDTD) {
 
-  if(aDefaultDTD) {
-    if(aDefaultDTD->IsCapableOf(aProcessType,aContentType,0)) {
-      result=PR_TRUE;
-    }
-  }
-  else {
-    nsDequeIterator b=gSharedParserObjects.mDTDDeque.Begin(); 
-    nsDequeIterator e=gSharedParserObjects.mDTDDeque.End(); 
+    //Let's start by tring the defaultDTD, if one exists...
+  if(aDefaultDTD && (aDefaultDTD->CanParse(aSourceType,0)))
+    return PR_TRUE;
 
-    while(b<e){
-      nsIDTD* theDTD=(nsIDTD*)b.GetCurrent();
-      if(theDTD && theDTD->IsCapableOf(aProcessType,aContentType,0)) {
+  PRBool  result=PR_FALSE;
+  nsDequeIterator b=gSharedParserObjects.mDTDDeque.Begin(); 
+  nsDequeIterator e=gSharedParserObjects.mDTDDeque.End(); 
+
+  while(b<e){
+    nsIDTD* theDTD=(nsIDTD*)b.GetCurrent();
+    if(theDTD) {
+      result=theDTD->CanParse(aSourceType,0); 
+      if(result){
         aDefaultDTD=theDTD;
-        result=PR_TRUE;
         break;
       }
-      ++b;
     }
-  }
+    b++;
+  } 
+
   return result;
 }
 
@@ -341,22 +361,30 @@ PRBool FindSuitableDTD(eProcessType aProcessType,eParseMode aMode,nsString& aCon
  *
  * @update	gess6/22/98
  * @param   aBuffer -- nsString containing sample data to be analyzed.
+ * @param   aType -- may hold typename given from netlib; will hold result given by DTD's.
  * @return  auto-detect result: eValid, eInvalid, eUnknown
  */
-eAutoDetectResult nsParser::AutoDetectContentType(nsString& aBuffer) {
+eAutoDetectResult nsParser::AutoDetectContentType(nsString& aBuffer,nsString& aType) {
 
     //The process:
     //  You should go out and ask each DTD if they
     //  recognize the content in the scanner.
     //  Somebody should say yes, or we can't continue.
 
-    //This method may change mContentType and mDTD.
+    //This method may change mSourceType and mDTD.
     //It absolutely changes mAutoDetectStatus
 
+  nsDequeIterator b=gSharedParserObjects.mDTDDeque.Begin(); 
+  nsDequeIterator e=gSharedParserObjects.mDTDDeque.End(); 
+
   mAutoDetectStatus=eUnknownDetect;
-  if(PR_TRUE==mContentType.Equals(kHTMLTextContentType)) {
-    mAutoDetectStatus=eValidDetect;
-  }
+  while((b<e) && (eUnknownDetect==mAutoDetectStatus)){
+    nsIDTD* theDTD=(nsIDTD*)b.GetCurrent();
+    if(theDTD) {
+      mAutoDetectStatus=theDTD->AutoDetectContentType(aBuffer,aType);
+    }
+    b++;
+  } 
 
   return mAutoDetectStatus;  
 }
@@ -373,13 +401,13 @@ eAutoDetectResult nsParser::AutoDetectContentType(nsString& aBuffer) {
  * @param 
  * @return
  */
-PRInt32 nsParser::WillBuildModel(eProcessType aProcessType, const char* aFilename){
+PRInt32 nsParser::WillBuildModel(const char* aFilename){
 
   mMajorIteration=-1;
   mMinorIteration=-1;
 
   mParseMode=DetermineParseMode();  
-  if(PR_TRUE==FindSuitableDTD(aProcessType,mParseMode,mContentType,mDTD)) {
+  if(PR_TRUE==FindSuitableDTD(mParseMode,mSourceType,mTargetType,mDTD)) {
     mDTD->SetParser(this);
     mDTD->SetContentSink(mSink);
     mDTD->WillBuildModel(aFilename);
@@ -439,11 +467,12 @@ PRBool nsParser::Parse(const char* aFilename){
   if(aFilename) {
 
     //ok, time to create our tokenizer and begin the process
+    mTargetType=kHTMLTextContentType;
     mScanner=new CScanner(aFilename,mParseMode);
     if(mScanner) {
       mScanner->Eof();
-      if(eValidDetect==AutoDetectContentType(mScanner->GetBuffer())) {
-        WillBuildModel(eParsing,aFilename);
+      if(eValidDetect==AutoDetectContentType(mScanner->GetBuffer(),mSourceType)) {
+        WillBuildModel(aFilename);
         status=ResumeParse();
         DidBuildModel(status);
       }
@@ -452,6 +481,18 @@ PRBool nsParser::Parse(const char* aFilename){
   return status;
 }
 
+/**
+ * Cause parser to parse input from given stream 
+ * @update	gess5/11/98
+ * @param   aStream is the i/o source
+ * @return  TRUE if all went well -- FALSE otherwise
+ */
+PRInt32 nsParser::Parse(fstream& aStream){
+  PRInt32 result=0;
+  return result;
+}
+
+
 /**
  *  This is the main controlling routine in the parsing process. 
  *  Note that it may get called multiple times for the same scanner, 
@@ -467,54 +508,6 @@ PRBool nsParser::Parse(const char* aFilename){
  *  @return  PR_TRUE if parse succeeded, PR_FALSE otherwise.
  */
 PRInt32 nsParser::Parse(nsIURL* aURL,nsIStreamObserver* aListener, nsIDTDDebug * aDTDDebug) {
-  PRInt32 status;
-
-  status = BeginParse(aURL, aListener, aDTDDebug);
-  if (NS_OK == status) {
-    status=mURL->Open(this);
-  }
-  return status;
-}
-
-
-/**
- * Call this method if all you want to do is parse 1 string full of HTML text.
- *
- * @update	gess5/11/98
- * @param   anHTMLString contains a string-full of real HTML
- * @param   appendTokens tells us whether we should insert tokens inline, or append them.
- * @return  TRUE if all went well -- FALSE otherwise
- */
-PRInt32 nsParser::Parse(nsString& aSourceBuffer,PRBool appendTokens){
-  PRInt32 result=kNoError;
-
-  mScanner=new CScanner(); 
-  mScanner->Append(aSourceBuffer);
-  
-  if(eValidDetect==AutoDetectContentType(aSourceBuffer)) {
-    WillBuildModel(eParsing,"from-string");
-    result=ResumeParse();
-    DidBuildModel(result);
-  }
-  
-  return result;
-}
-
-/**
- *  This is the main controlling routine in the parsing process. 
- *  Note that it may get called multiple times for the same scanner, 
- *  since this is a pushed based system, and all the tokens may 
- *  not have been consumed by the scanner during a given invocation 
- *  of this method. 
- *
- *  NOTE: We don't call willbuildmodel here, because it will happen
- *        as a result of calling OnStartBinding later on.
- *
- *  @update  gess 3/25/98
- *  @param   aFilename -- const char* containing file to be parsed.
- *  @return  PR_TRUE if parse succeeded, PR_FALSE otherwise.
- */
-PRInt32 nsParser::BeginParse(nsIURL* aURL,nsIStreamObserver* aListener, nsIDTDDebug * aDTDDebug) {
   NS_PRECONDITION(0!=aURL,kNullURL);
 
   PRInt32 status=kBadURL;
@@ -537,36 +530,27 @@ PRInt32 nsParser::BeginParse(nsIURL* aURL,nsIStreamObserver* aListener, nsIDTDDe
   return status;
 }
 
-/**
- * 
- * @update	gess6/23/98
- * @param 
- * @return
- */
-PRInt32 nsParser::Convert(nsIURL* aURL,char* aSourceForm,char* aTargetForm,nsIStreamListener* aListener) {
-  PRInt32 result=0;
-  return result;
-}
 
 /**
- * 
- * @update	gess6/23/98
- * @param 
- * @return
+ * Call this method if all you want to do is parse 1 string full of HTML text.
+ *
+ * @update	gess5/11/98
+ * @param   anHTMLString contains a string-full of real HTML
+ * @param   appendTokens tells us whether we should insert tokens inline, or append them.
+ * @return  TRUE if all went well -- FALSE otherwise
  */
-PRInt32 nsParser::Convert(const char* aFilename,char* aSourceForm,char* aTargetForm) {
-  PRInt32 result=0;
-  return result;
-}
+PRInt32 nsParser::Parse(nsString& aSourceBuffer,PRBool appendTokens){
+  PRInt32 result=kNoError;
 
-/** 
- * 
- * @update	gess6/23/98
- * @param 
- * @return
- */
-PRInt32 nsParser::Convert(nsString& anHTMLString,char* aSourceForm,char* aTargetForm,PRBool appendTokens){
-  PRInt32 result=0;
+  mTargetType=kHTMLTextContentType;
+  mScanner=new CScanner(); 
+  mScanner->Append(aSourceBuffer);  
+  if(eValidDetect==AutoDetectContentType(aSourceBuffer,mSourceType)) {
+    WillBuildModel("");
+    result=ResumeParse();
+    DidBuildModel(result);
+  }
+  
   return result;
 }
 
@@ -722,14 +706,14 @@ nsParser::OnProgress(PRInt32 aProgress, PRInt32 aProgressMax,
  *  @param   
  *  @return  
  */
-nsresult nsParser::OnStartBinding(const char *aContentType){
+nsresult nsParser::OnStartBinding(const char *aSourceType){
   if (nsnull != mObserver) {
-    mObserver->OnStartBinding(aContentType);
+    mObserver->OnStartBinding(aSourceType);
   }
   mAutoDetectStatus=eUnknownDetect;
   mDTD=0;
 
-  mContentType=aContentType;
+  mSourceType=aSourceType;
   return kNoError;
 }
 
@@ -773,8 +757,8 @@ nsresult nsParser::OnDataAvailable(nsIInputStream *pIStream, PRInt32 length){
         mScanner->Append(mTransferBuffer,len);
 
         if(eUnknownDetect==mAutoDetectStatus) {
-          if(eValidDetect==AutoDetectContentType(mScanner->GetBuffer())) {
-            nsresult result=WillBuildModel(eParsing,mURL->GetSpec());
+          if(eValidDetect==AutoDetectContentType(mScanner->GetBuffer(),mSourceType)) {
+            nsresult result=WillBuildModel(mURL->GetSpec());
           } //if
         }
       } //if

mozilla/parser/htmlparser/src/nsParser.h

@@ -74,7 +74,7 @@ class nsIDTD;
 class nsIDTDDebug;
 class CScanner;
 class nsIParserFilter;
-
+class fstream;
 
 class nsParser : public nsIParser, public nsIStreamListener {
             
@@ -118,10 +118,6 @@ friend class CTokenHandler;
      */
     virtual CScanner* GetScanner(void);
 
-    virtual PRInt32 BeginParse(nsIURL* aURL,
-                               nsIStreamObserver* aListener, 
-                               nsIDTDDebug * aDTDDebug = 0);
-
     /**
      * Cause parser to parse input from given URL in given mode
      * @update	gess5/11/98
@@ -136,11 +132,18 @@ friend class CTokenHandler;
      * Cause parser to parse input from given file in given mode
      * @update	gess5/11/98
      * @param   aFilename is a path for file document
-     * @param   aMode is the desired parser mode (Nav, other, etc.)
      * @return  TRUE if all went well -- FALSE otherwise
      */
     virtual PRInt32 Parse(const char* aFilename);
 
+    /**
+     * Cause parser to parse input from given stream 
+     * @update	gess5/11/98
+     * @param   aStream is the i/o source
+     * @return  TRUE if all went well -- FALSE otherwise
+     */
+    virtual PRInt32 Parse(fstream& aStream);
+
     /**
      * @update	gess5/11/98
      * @param   anHTMLString contains a string-full of real HTML
@@ -149,15 +152,6 @@ friend class CTokenHandler;
      */
     virtual PRInt32 Parse(nsString& anHTMLString,PRBool appendTokens);
 
-    /******************************************************************************************
-     *  Convert methods start input source (of known or unknown form), and perform conversions 
-     *  until you wind up with a <i>stream</i> in your target form.
-     *  The internal content model is never effected.
-     ******************************************************************************************/
-    virtual PRInt32 Convert(nsIURL* aURL,char* aSourceForm, char* aTargetForm,nsIStreamListener* aListener);
-    virtual PRInt32 Convert(const char* aFilename,char* aSourceForm,char* aTargetForm);
-    virtual PRInt32 Convert(nsString& anHTMLString,char* aSourceForm,char* aTargetForm,PRBool appendTokens);
-
     /**
      * This method gets called (automatically) during incremental parsing
      * @update	gess5/11/98
@@ -214,7 +208,7 @@ protected:
      * @param 
      * @return
      */
-    PRInt32 WillBuildModel(eProcessType theProcessType=eParsing,const char* aFilename=0);
+    PRInt32 WillBuildModel(const char* aFilename=0);
 
     /**
      * 
@@ -309,7 +303,7 @@ private:
      * @param 
      * @return  TRUE if we figured it out.
      */
-    eAutoDetectResult AutoDetectContentType(nsString& aBuffer);
+    eAutoDetectResult AutoDetectContentType(nsString& aBuffer,nsString& aType);
 
 
 protected:
@@ -335,7 +329,8 @@ protected:
     CScanner*           mScanner;
     nsIURL*             mURL;
 	  nsIDTDDebug*		    mDTDDebug;
-    nsString            mContentType;
+    nsString            mSourceType;
+    nsString            mTargetType;
     eAutoDetectResult   mAutoDetectStatus;
 };
 

mozilla/parser/htmlparser/src/nsParserTypes.h

@@ -42,40 +42,41 @@ enum  eParseMode {
 };
 
 
-const PRInt32   kEOF          = 1000000L;
-const PRInt32   kUnknownError = -1000;
-const PRInt32   kCantPropagate = -6;
-const PRInt32   kContextMismatch = -5;
-const PRInt32   kBadFilename  = -4;
-const PRInt32   kBadURL       = -3;
-const PRInt32   kInterrupted  = -2;
-const PRInt32   kNotFound     = -1;
-const PRInt32   kNoError      = 0;
+const PRInt32   kEOF              = 1000000L;
+const PRInt32   kUnknownError     = -1000;
+const PRInt32   kCantPropagate    = -6;
+const PRInt32   kContextMismatch  = -5;
+const PRInt32   kBadFilename      = -4;
+const PRInt32   kBadURL           = -3;
+const PRInt32   kInterrupted      = -2;
+const PRInt32   kNotFound         = -1;
+const PRInt32   kNoError          = 0;
 
-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  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  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  kForwardSlash     = '/';
+const PRUint32  kBackSlash        = '\\';
+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       = '}';
 
 
 #endif

mozilla/parser/htmlparser/src/nsScanner.cpp

@@ -47,6 +47,7 @@ CScanner::CScanner(eParseMode aMode) : mBuffer("") {
   mNetStream=0;
   mFileStream=0;
   mIncremental=PR_TRUE;
+  mOwnsStream=PR_TRUE;
 }
 
 /**
@@ -64,6 +65,7 @@ CScanner::CScanner(const char* aFilename,eParseMode aMode) : mBuffer("") {
   mParseMode=aMode;
   mNetStream=0;
   mIncremental=PR_FALSE;
+  mOwnsStream=PR_TRUE;
 #if defined(XP_UNIX) && (defined(IRIX) || defined(MKLINUX))
   /* XXX: IRIX does not support ios::binary */
   mFileStream=new fstream(aFilename,ios::in);
@@ -72,6 +74,24 @@ CScanner::CScanner(const char* aFilename,eParseMode aMode) : mBuffer("") {
 #endif
 }
 
+/**
+ *  Use this constructor if you want i/o to be file based.
+ *
+ *  @update  gess 5/12/98
+ *  @param   aMode represents the parser mode (nav, other)
+ *  @return  
+ */
+CScanner::CScanner(fstream& aStream,eParseMode aMode) : mBuffer("") {
+  mOffset=0;
+  mMarkPos=-1;
+  mTotalRead=0;
+  mParseMode=aMode;
+  mNetStream=0;
+  mIncremental=PR_FALSE;
+  mOwnsStream=PR_FALSE;
+  mFileStream=&aStream;
+}
+
 /**
  *  Use this constructor if you want i/o to be based on a
  *  non-incremental netstream.
@@ -91,6 +111,7 @@ CScanner::CScanner(nsIURL* aURL,eParseMode aMode) : mBuffer("") {
   mIncremental=PR_FALSE;
   mNetStream=aURL->Open(&error);
   gURLRef=aURL->GetSpec();
+  mOwnsStream=PR_FALSE;
 }
 
 
@@ -104,7 +125,8 @@ CScanner::CScanner(nsIURL* aURL,eParseMode aMode) : mBuffer("") {
 CScanner::~CScanner() {
   if(mFileStream) {
     mFileStream->close();
-    delete mFileStream;
+    if(mOwnsStream)
+      delete mFileStream;
   }
   else if(mNetStream) {
     mNetStream->Close();
@@ -354,7 +376,7 @@ PRInt32 CScanner::PutBack(PRUnichar aChar) {
  *  @param   
  *  @return  error status
  */
-PRInt32 CScanner::SkipWhite(void) {
+PRInt32 CScanner::SkipWhitespace(void) {
   static nsAutoString chars(" \n\r\t");
   return SkipOver(chars);
 }

mozilla/parser/htmlparser/src/nsScanner.h

@@ -25,7 +25,7 @@
  * how to consume characters out of an (internal) stream.
  * This class also offers a series of utility methods
  * that most tokenizers want, such as readUntil(), 
- * readWhile() and SkipWhite().
+ * readWhile() and SkipWhitespace().
  */
 
 
@@ -73,6 +73,16 @@ class CScanner {
        */
       CScanner(const char* aFilename,eParseMode aMode=eParseMode_navigator);
 
+      /**
+       *  Use this constructor if you want i/o to be stream based.
+       *
+       *  @update  gess 5/12/98
+       *  @param   aMode represents the parser mode (nav, other)
+       *  @return  
+       */
+      CScanner(fstream& aStream,eParseMode aMode=eParseMode_navigator);
+
+
       ~CScanner();
 
       /**
@@ -136,7 +146,7 @@ class CScanner {
        *  @update  gess 3/25/98
        *  @return  error status
        */
-      PRInt32 SkipWhite(void);
+      PRInt32 SkipWhitespace(void);
 
       /**
        *  Determine if the scanner has reached EOF.
@@ -265,6 +275,7 @@ class CScanner {
       PRInt32         mTotalRead;
       eParseMode      mParseMode;
       PRBool          mIncremental;
+      PRBool          mOwnsStream;
 };
 
 #endif

mozilla/parser/htmlparser/src/nsValidDTD.cpp

@@ -0,0 +1,334 @@
+/* -*- Mode: C++; tab-width: 2; indent-tabs-mode: nil; c-basic-offset: 2 -*- */
+/*
+ * The contents of this file are subject to the Netscape Public License
+ * Version 1.0 (the "NPL"); you may not use this file except in
+ * compliance with the NPL.  You may obtain a copy of the NPL at
+ * http://www.mozilla.org/NPL/
+ *
+ * Software distributed under the NPL is distributed on an "AS IS" basis,
+ * WITHOUT WARRANTY OF ANY KIND, either express or implied. See the NPL
+ * for the specific language governing rights and limitations under the
+ * NPL.
+ *
+ * The Initial Developer of this code under the NPL is Netscape
+ * Communications Corporation.  Portions created by Netscape are
+ * Copyright (C) 1998 Netscape Communications Corporation.  All Rights
+ * Reserved.
+ */
+
+/**
+ * MODULE NOTES:
+ * @update  gess 4/8/98
+ * 
+ *         
+ */
+
+/**
+ * TRANSIENT STYLE-HANDLING NOTES:
+ * @update  gess 6/15/98
+ * 
+ * ...add comments here about transient style stack.
+ *         
+   */
+
+#include "nsIDTDDebug.h"
+#include "nsValidDTD.h"
+#include "nsCRT.h"
+#include "nsParser.h"
+#include "nsScanner.h"
+#include "nsParserTypes.h"
+#include "nsTokenHandler.h"
+
+#include "prenv.h"  //this is here for debug reasons...
+#include "prtypes.h"  //this is here for debug reasons...
+#include "prio.h"
+#include "plstr.h"
+
+#ifdef XP_PC
+#include <direct.h> //this is here for debug reasons...
+#endif
+#include "prmem.h"
+
+
+static NS_DEFINE_IID(kISupportsIID, NS_ISUPPORTS_IID);                 
+static NS_DEFINE_IID(kIDTDIID,      NS_IDTD_IID);
+static NS_DEFINE_IID(kClassIID,     NS_VALID_DTD_IID); 
+
+
+static const char* kNullURL = "Error: Null URL given";
+static const char* kNullFilename= "Error: Null filename given";
+static const char* kNullTokenizer = "Error: Unable to construct tokenizer";
+static const char* kNullToken = "Error: Null token given";
+static const char* kInvalidTagStackPos = "Error: invalid tag stack position";
+static const char* kRTFTextContentType = "text/xml";
+
+static nsAutoString gEmpty;
+
+
+/**
+ *  This method gets called as part of our COM-like interfaces.
+ *  Its purpose is to create an interface to parser object
+ *  of some type.
+ *  
+ *  @update   gess 4/8/98
+ *  @param    nsIID  id of object to discover
+ *  @param    aInstancePtr ptr to newly discovered interface
+ *  @return   NS_xxx result code
+ */
+nsresult CValidDTD::QueryInterface(const nsIID& aIID, void** aInstancePtr)  
+{                                                                        
+  if (NULL == aInstancePtr) {                                            
+    return NS_ERROR_NULL_POINTER;                                        
+  }                                                                      
+
+  if(aIID.Equals(kISupportsIID))    {  //do IUnknown...
+    *aInstancePtr = (nsIDTD*)(this);                                        
+  }
+  else if(aIID.Equals(kIDTDIID)) {  //do IParser base class...
+    *aInstancePtr = (nsIDTD*)(this);                                        
+  }
+  else if(aIID.Equals(kClassIID)) {  //do this class...
+    *aInstancePtr = (CValidDTD*)(this);                                        
+  }                 
+  else {
+    *aInstancePtr=0;
+    return NS_NOINTERFACE;
+  }
+  ((nsISupports*) *aInstancePtr)->AddRef();
+  return NS_OK;                                                        
+}
+
+/**
+ *  This method is defined in nsIParser. It is used to 
+ *  cause the COM-like construction of an nsParser.
+ *  
+ *  @update  gess 4/8/98
+ *  @param   nsIParser** ptr to newly instantiated parser
+ *  @return  NS_xxx error result
+ */
+NS_HTMLPARS nsresult NS_NewValid_DTD(nsIDTD** aInstancePtrResult)
+{
+  CValidDTD* it = new CValidDTD();
+
+  if (it == 0) {
+    return NS_ERROR_OUT_OF_MEMORY;
+  }
+
+  return it->QueryInterface(kClassIID, (void **) aInstancePtrResult);
+}
+
+
+NS_IMPL_ADDREF(CValidDTD)
+NS_IMPL_RELEASE(CValidDTD)
+
+
+/**
+ *  Default constructor
+ *  
+ *  @update  gess 4/9/98
+ *  @param   
+ *  @return  
+ */
+CValidDTD::CValidDTD() : nsIDTD() {
+  NS_INIT_REFCNT();
+  mParser=0;
+  mFilename=0;
+}
+
+/**
+ *  Default destructor
+ *  
+ *  @update  gess 4/9/98
+ *  @param   
+ *  @return  
+ */
+CValidDTD::~CValidDTD(){
+}
+
+/**
+ * This method is called to determine if the given DTD can parse
+ * a document in a given source-type. 
+ * NOTE: Parsing always assumes that the end result will involve
+ *       storing the result in the main content model.
+ * @update	gess6/24/98
+ * @param   
+ * @return  TRUE if this DTD can satisfy the request; FALSE otherwise.
+ */
+PRBool CValidDTD::CanParse(nsString& aContentType, PRInt32 aVersion){
+  PRBool result=aContentType.Equals(kRTFTextContentType);
+  return result;
+}
+
+
+/**
+ * This method is called to determine if the given DTD can perform
+ * a document conversion from a given source-type to a given target-type.
+ * NOTE: conversion always assumes an input stream and an outputstream.
+ * @update	gess6/24/98
+ * @param   
+ * @return  TRUE if this DTD can satisfy the request; FALSE otherwise.
+ */
+PRBool CValidDTD::CanConvert(nsString& aSourceType, nsString& aTargetType, PRInt32 aVersion){
+  PRBool result=PR_TRUE;
+  return result;
+}
+
+/**
+ * 
+ * @update	gess7/7/98
+ * @param 
+ * @return
+ */
+eAutoDetectResult CValidDTD::AutoDetectContentType(nsString& aBuffer,nsString& aType){
+  eAutoDetectResult result=eUnknownDetect;
+  return result;
+}
+
+/**
+ * 
+ * @update	gess5/18/98
+ * @param 
+ * @return
+ */
+PRInt32 CValidDTD::WillBuildModel(const char* aFilename){
+  PRInt32 result=0;
+  return result;
+}
+
+/**
+ * 
+ * @update	gess5/18/98
+ * @param 
+ * @return
+ */
+PRInt32 CValidDTD::DidBuildModel(PRInt32 anErrorCode){
+  PRInt32 result=0;
+
+  return result;
+}
+
+/**
+ * 
+ *  
+ *  @update  gess 3/25/98
+ *  @param   
+ *  @return 
+ */
+void CValidDTD::SetParser(nsIParser* aParser) {
+  mParser=(nsParser*)aParser;
+}
+
+/**
+ *  This method gets called in order to set the content
+ *  sink for this parser to dump nodes to.
+ *  
+ *  @update  gess 3/25/98
+ *  @param   nsIContentSink interface for node receiver
+ *  @return  
+ */
+nsIContentSink* CValidDTD::SetContentSink(nsIContentSink* aSink) {
+  return 0;
+}
+
+
+/*******************************************************************
+  These methods use to be hidden in the tokenizer-delegate. 
+  That file merged with the DTD, since the separation wasn't really
+  buying us anything.
+ *******************************************************************/
+
+/**
+ *  This method repeatedly called by the tokenizer. 
+ *  Each time, we determine the kind of token were about to 
+ *  read, and then we call the appropriate method to handle
+ *  that token type.
+ *  
+ *  @update gess 3/25/98
+ *  @param  aChar: last char read
+ *  @param  aScanner: see nsScanner.h
+ *  @param  anErrorCode: arg that will hold error condition
+ *  @return new token or null 
+ */
+PRInt32 CValidDTD::ConsumeToken(CToken*& aToken){
+  
+  CScanner* theScanner=mParser->GetScanner();
+
+  PRUnichar aChar;
+  PRInt32   result=theScanner->GetChar(aChar);
+
+  switch(result) {
+    case kEOF:
+      break;
+
+    case kInterrupted:
+      theScanner->RewindToMark();
+      break; 
+
+    case kNoError:
+    default:
+      switch(aChar) {
+        case kLessThan:
+        default:
+          break;
+      } //switch
+      break; 
+  } //switch
+  if(kNoError==result)
+    result=theScanner->Eof();
+  return result;
+}
+
+/**
+ * 
+ * @update	gess5/18/98
+ * @param 
+ * @return
+ */
+void CValidDTD::WillResumeParse(void){
+  return;
+}
+
+/**
+ * 
+ * @update	gess5/18/98
+ * @param 
+ * @return
+ */
+void CValidDTD::WillInterruptParse(void){
+  return;
+}
+
+/**
+ * 
+ * @update	jevering6/23/98
+ * @param 
+ * @return
+ */
+void CValidDTD::SetDTDDebug(nsIDTDDebug * aDTDDebug) {
+}
+
+/**
+ *  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 -- int tag of parent container
+ *  @param   aChild -- int tag of child container
+ *  @return  PR_TRUE if parent can contain child
+ */
+PRBool CValidDTD::CanContain(PRInt32 aParent,PRInt32 aChild){
+  PRBool result=PR_FALSE;
+  return result;
+}
+
+/**
+ *  
+ *  @update  gess 3/25/98
+ *  @param   aToken -- token object to be put into content model
+ *  @return  0 if all is well; non-zero is an error
+ */
+PRInt32 CValidDTD::HandleToken(CToken* aToken) {
+  PRInt32 result=0;
+  return result;
+}
+

mozilla/parser/htmlparser/src/nsValidDTD.h

@@ -0,0 +1,205 @@
+
+/* -*- Mode: C++; tab-width: 2; indent-tabs-mode: nil; c-basic-offset: 2 -*- */
+/*
+ * The contents of this file are subject to the Netscape Public License
+ * Version 1.0 (the "NPL"); you may not use this file except in
+ * compliance with the NPL.  You may obtain a copy of the NPL at
+ * http://www.mozilla.org/NPL/
+ *
+ * Software distributed under the NPL is distributed on an "AS IS" basis,
+ * WITHOUT WARRANTY OF ANY KIND, either express or implied. See the NPL
+ * for the specific language governing rights and limitations under the
+ * NPL.
+ *
+ * The Initial Developer of this code under the NPL is Netscape
+ * Communications Corporation.  Portions created by Netscape are
+ * Copyright (C) 1998 Netscape Communications Corporation.  All Rights
+ * Reserved.
+ */
+
+/**
+ * MODULE NOTES:
+ * @update  gess 4/8/98
+ * 
+ *         
+ */
+
+#ifndef __NS_VALID_DTD
+#define __NS_VALID_DTD
+
+#include "nsIDTD.h"
+#include "nsISupports.h"
+#include "nsHTMLTokens.h"
+#include "nshtmlpars.h"
+#include "nsVoidArray.h"
+#include "nsDeque.h"
+
+#define NS_VALID_DTD_IID      \
+  {0xa39c6bfe, 0x15f0,    0x11d2, \
+  {0x80, 0x41, 0x0, 0x10, 0x4b, 0x98, 0x3f, 0xd4}}
+
+
+class nsIHTMLContentSink;
+class nsIDTDDebug;
+class nsIParserNode;
+class CITokenHandler;
+class nsParser;
+
+
+
+class CValidDTD : public nsIDTD {
+            
+  public:
+
+    NS_DECL_ISUPPORTS
+
+
+    /**
+     *  
+     *  
+     *  @update  gess 4/9/98
+     *  @param   
+     *  @return  
+     */
+    CValidDTD();
+
+    /**
+     *  
+     *  
+     *  @update  gess 4/9/98
+     *  @param   
+     *  @return  
+     */
+    virtual ~CValidDTD();
+
+    /**
+     * This method is called to determine if the given DTD can parse
+     * a document in a given source-type. 
+     * NOTE: Parsing always assumes that the end result will involve
+     *       storing the result in the main content model.
+     * @update	gess6/24/98
+     * @param   
+     * @return  TRUE if this DTD can satisfy the request; FALSE otherwise.
+     */
+    virtual PRBool CanParse(nsString& aContentType, PRInt32 aVersion);
+
+
+    /**
+     * This method is called to determine if the given DTD can perform
+     * a document conversion from a given source-type to a given target-type.
+     * NOTE: conversion always assumes an input stream and an outputstream.
+     * @update	gess6/24/98
+     * @param   
+     * @return  TRUE if this DTD can satisfy the request; FALSE otherwise.
+     */
+    virtual PRBool CanConvert(nsString& aSourceType, nsString& aTargetType, PRInt32 aVersion);
+
+    /**
+     * 
+     * @update	gess7/7/98
+     * @param 
+     * @return
+     */
+    virtual eAutoDetectResult AutoDetectContentType(nsString& aBuffer,nsString& aType);
+
+    /**
+     * 
+     * @update	gess5/18/98
+     * @param 
+     * @return
+     */
+    virtual PRInt32 WillBuildModel(const char* aFilename=0);
+
+    /**
+     * 
+     * @update	gess5/18/98
+     * @param 
+     * @return
+     */
+    virtual PRInt32 DidBuildModel(PRInt32 anErrorCode);
+
+    /**
+     *  
+     *  @update  gess 3/25/98
+     *  @param   aToken -- token object to be put into content model
+     *  @return  0 if all is well; non-zero is an error
+     */
+    virtual PRInt32 HandleToken(CToken* aToken);
+
+    /**
+     * 
+     *  
+     *  @update  gess 3/25/98
+     *  @param   
+     *  @return 
+     */
+    virtual void SetParser(nsIParser* aParser);
+
+    /**
+     *  Cause the tokenizer to consume the next token, and 
+     *  return an error result.
+     *  
+     *  @update  gess 3/25/98
+     *  @param   anError -- ref to error code
+     *  @return  new token or null
+     */
+    virtual PRInt32 ConsumeToken(CToken*& aToken);
+
+
+    /**
+     * 
+     * @update	gess5/18/98
+     * @param 
+     * @return
+     */
+    virtual void WillResumeParse(void);
+
+    /**
+     * 
+     * @update	gess5/18/98
+     * @param 
+     * @return
+     */
+    virtual void WillInterruptParse(void);
+
+   /**
+     * Select given content sink into parser for parser output
+     * @update	gess5/11/98
+     * @param   aSink is the new sink to be used by parser
+     * @return  old sink, or NULL
+     */
+    virtual nsIContentSink* SetContentSink(nsIContentSink* aSink);
+
+    /**
+     * 
+     * @update	jevering6/23/98
+     * @param 
+     * @return
+     */
+	  virtual void SetDTDDebug(nsIDTDDebug * aDTDDebug);
+
+    /**
+     *  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 -- int tag of parent container
+     *  @param   aChild -- int tag of child container
+     *  @return  PR_TRUE if parent can contain child
+     */
+    virtual PRBool CanContain(PRInt32 aParent,PRInt32 aChild);
+	
+protected:
+
+    
+    nsParser*           mParser;
+    nsIHTMLContentSink* mSink;
+    char*               mFilename;
+};
+
+extern NS_HTMLPARS nsresult NS_NewValid_DTD(nsIDTD** aInstancePtrResult);
+
+#endif 
+
+
+

mozilla/parser/htmlparser/src/nsWellFormedDTD.cpp

@@ -0,0 +1,321 @@
+/* -*- Mode: C++; tab-width: 2; indent-tabs-mode: nil; c-basic-offset: 2 -*- */
+/*
+ * The contents of this file are subject to the Netscape Public License
+ * Version 1.0 (the "NPL"); you may not use this file except in
+ * compliance with the NPL.  You may obtain a copy of the NPL at
+ * http://www.mozilla.org/NPL/
+ *
+ * Software distributed under the NPL is distributed on an "AS IS" basis,
+ * WITHOUT WARRANTY OF ANY KIND, either express or implied. See the NPL
+ * for the specific language governing rights and limitations under the
+ * NPL.
+ *
+ * The Initial Developer of this code under the NPL is Netscape
+ * Communications Corporation.  Portions created by Netscape are
+ * Copyright (C) 1998 Netscape Communications Corporation.  All Rights
+ * Reserved.
+ */
+
+/**
+ * MODULE NOTES:
+ * @update  gess 4/8/98
+ * 
+ *         
+ */
+
+/**
+ * TRANSIENT STYLE-HANDLING NOTES:
+ * @update  gess 6/15/98
+ * 
+ * ...add comments here about transient style stack.
+ *         
+   */
+
+#include "nsIDTDDebug.h"
+#include "nsWellFormedDTD.h"
+#include "nsCRT.h"
+#include "nsParser.h"
+#include "nsScanner.h"
+#include "nsParserTypes.h"
+#include "nsTokenHandler.h"
+
+#include "prenv.h"  //this is here for debug reasons...
+#include "prtypes.h"  //this is here for debug reasons...
+#include "prio.h"
+#include "plstr.h"
+
+#ifdef XP_PC
+#include <direct.h> //this is here for debug reasons...
+#endif
+#include "prmem.h"
+
+
+static NS_DEFINE_IID(kISupportsIID, NS_ISUPPORTS_IID);                 
+static NS_DEFINE_IID(kIDTDIID,      NS_IDTD_IID);
+static NS_DEFINE_IID(kClassIID,     NS_WELLFORMED_DTD_IID); 
+
+
+static const char* kNullURL = "Error: Null URL given";
+static const char* kNullFilename= "Error: Null filename given";
+static const char* kNullTokenizer = "Error: Unable to construct tokenizer";
+static const char* kNullToken = "Error: Null token given";
+static const char* kInvalidTagStackPos = "Error: invalid tag stack position";
+static const char* kRTFTextContentType = "text/xml";
+
+static nsAutoString gEmpty;
+
+
+/**
+ *  This method gets called as part of our COM-like interfaces.
+ *  Its purpose is to create an interface to parser object
+ *  of some type.
+ *  
+ *  @update   gess 4/8/98
+ *  @param    nsIID  id of object to discover
+ *  @param    aInstancePtr ptr to newly discovered interface
+ *  @return   NS_xxx result code
+ */
+nsresult CWellFormedDTD::QueryInterface(const nsIID& aIID, void** aInstancePtr)  
+{                                                                        
+  if (NULL == aInstancePtr) {                                            
+    return NS_ERROR_NULL_POINTER;                                        
+  }                                                                      
+
+  if(aIID.Equals(kISupportsIID))    {  //do IUnknown...
+    *aInstancePtr = (nsIDTD*)(this);                                        
+  }
+  else if(aIID.Equals(kIDTDIID)) {  //do IParser base class...
+    *aInstancePtr = (nsIDTD*)(this);                                        
+  }
+  else if(aIID.Equals(kClassIID)) {  //do this class...
+    *aInstancePtr = (CWellFormedDTD*)(this);                                        
+  }                 
+  else {
+    *aInstancePtr=0;
+    return NS_NOINTERFACE;
+  }
+  ((nsISupports*) *aInstancePtr)->AddRef();
+  return NS_OK;                                                        
+}
+
+/**
+ *  This method is defined in nsIParser. It is used to 
+ *  cause the COM-like construction of an nsParser.
+ *  
+ *  @update  gess 4/8/98
+ *  @param   nsIParser** ptr to newly instantiated parser
+ *  @return  NS_xxx error result
+ */
+NS_HTMLPARS nsresult NS_NewWellFormed_DTD(nsIDTD** aInstancePtrResult)
+{
+  CWellFormedDTD* it = new CWellFormedDTD();
+
+  if (it == 0) {
+    return NS_ERROR_OUT_OF_MEMORY;
+  }
+
+  return it->QueryInterface(kClassIID, (void **) aInstancePtrResult);
+}
+
+
+NS_IMPL_ADDREF(CWellFormedDTD)
+NS_IMPL_RELEASE(CWellFormedDTD)
+
+
+/**
+ *  Default constructor
+ *  
+ *  @update  gess 4/9/98
+ *  @param   
+ *  @return  
+ */
+CWellFormedDTD::CWellFormedDTD() : nsIDTD() {
+  NS_INIT_REFCNT();
+  mParser=0;
+  mFilename=0;
+}
+
+/**
+ *  Default destructor
+ *  
+ *  @update  gess 4/9/98
+ *  @param   
+ *  @return  
+ */
+CWellFormedDTD::~CWellFormedDTD(){
+}
+
+/**
+ * This method is called to determine if the given DTD can parse
+ * a document in a given source-type. 
+ * NOTE: Parsing always assumes that the end result will involve
+ *       storing the result in the main content model.
+ * @update	gess6/24/98
+ * @param   
+ * @return  TRUE if this DTD can satisfy the request; FALSE otherwise.
+ */
+PRBool CWellFormedDTD::CanParse(nsString& aContentType, PRInt32 aVersion){
+  PRBool result=aContentType.Equals(kRTFTextContentType);
+  return result;
+}
+
+
+/**
+ * 
+ * @update	gess7/7/98
+ * @param 
+ * @return
+ */
+eAutoDetectResult CWellFormedDTD::AutoDetectContentType(nsString& aBuffer,nsString& aType){
+  eAutoDetectResult result=eUnknownDetect;
+  return result;
+}
+
+/**
+ * 
+ * @update	gess5/18/98
+ * @param 
+ * @return
+ */
+PRInt32 CWellFormedDTD::WillBuildModel(const char* aFilename){
+  PRInt32 result=0;
+  return result;
+}
+
+/**
+ * 
+ * @update	gess5/18/98
+ * @param 
+ * @return
+ */
+PRInt32 CWellFormedDTD::DidBuildModel(PRInt32 anErrorCode){
+  PRInt32 result=0;
+
+  return result;
+}
+
+/**
+ * 
+ *  
+ *  @update  gess 3/25/98
+ *  @param   
+ *  @return 
+ */
+void CWellFormedDTD::SetParser(nsIParser* aParser) {
+  mParser=(nsParser*)aParser;
+}
+
+/**
+ *  This method gets called in order to set the content
+ *  sink for this parser to dump nodes to.
+ *  
+ *  @update  gess 3/25/98
+ *  @param   nsIContentSink interface for node receiver
+ *  @return  
+ */
+nsIContentSink* CWellFormedDTD::SetContentSink(nsIContentSink* aSink) {
+  return 0;
+}
+
+
+/*******************************************************************
+  These methods use to be hidden in the tokenizer-delegate. 
+  That file merged with the DTD, since the separation wasn't really
+  buying us anything.
+ *******************************************************************/
+
+/**
+ *  This method repeatedly called by the tokenizer. 
+ *  Each time, we determine the kind of token were about to 
+ *  read, and then we call the appropriate method to handle
+ *  that token type.
+ *  
+ *  @update gess 3/25/98
+ *  @param  aChar: last char read
+ *  @param  aScanner: see nsScanner.h
+ *  @param  anErrorCode: arg that will hold error condition
+ *  @return new token or null 
+ */
+PRInt32 CWellFormedDTD::ConsumeToken(CToken*& aToken){
+  
+  CScanner* theScanner=mParser->GetScanner();
+
+  PRUnichar aChar;
+  PRInt32   result=theScanner->GetChar(aChar);
+
+  switch(result) {
+    case kEOF:
+      break;
+
+    case kInterrupted:
+      theScanner->RewindToMark();
+      break; 
+
+    case kNoError:
+    default:
+      switch(aChar) {
+        case kLessThan:
+        default:
+          break;
+      } //switch
+      break; 
+  } //switch
+  if(kNoError==result)
+    result=theScanner->Eof();
+  return result;
+}
+
+/**
+ * 
+ * @update	gess5/18/98
+ * @param 
+ * @return
+ */
+void CWellFormedDTD::WillResumeParse(void){
+  return;
+}
+
+/**
+ * 
+ * @update	gess5/18/98
+ * @param 
+ * @return
+ */
+void CWellFormedDTD::WillInterruptParse(void){
+  return;
+}
+
+/**
+ * 
+ * @update	jevering6/23/98
+ * @param 
+ * @return
+ */
+void CWellFormedDTD::SetDTDDebug(nsIDTDDebug * aDTDDebug) {
+}
+
+/**
+ *  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 -- int tag of parent container
+ *  @param   aChild -- int tag of child container
+ *  @return  PR_TRUE if parent can contain child
+ */
+PRBool CWellFormedDTD::CanContain(PRInt32 aParent,PRInt32 aChild){
+  PRBool result=PR_FALSE;
+  return result;
+}
+
+/**
+ *  
+ *  @update  gess 3/25/98
+ *  @param   aToken -- token object to be put into content model
+ *  @return  0 if all is well; non-zero is an error
+ */
+PRInt32 CWellFormedDTD::HandleToken(CToken* aToken) {
+  PRInt32 result=0;
+  return result;
+}
+

mozilla/parser/htmlparser/src/nsWellFormedDTD.h

@@ -0,0 +1,194 @@
+
+/* -*- Mode: C++; tab-width: 2; indent-tabs-mode: nil; c-basic-offset: 2 -*- */
+/*
+ * The contents of this file are subject to the Netscape Public License
+ * Version 1.0 (the "NPL"); you may not use this file except in
+ * compliance with the NPL.  You may obtain a copy of the NPL at
+ * http://www.mozilla.org/NPL/
+ *
+ * Software distributed under the NPL is distributed on an "AS IS" basis,
+ * WITHOUT WARRANTY OF ANY KIND, either express or implied. See the NPL
+ * for the specific language governing rights and limitations under the
+ * NPL.
+ *
+ * The Initial Developer of this code under the NPL is Netscape
+ * Communications Corporation.  Portions created by Netscape are
+ * Copyright (C) 1998 Netscape Communications Corporation.  All Rights
+ * Reserved.
+ */
+
+/**
+ * MODULE NOTES:
+ * @update  gess 4/8/98
+ * 
+ *         
+ */
+
+#ifndef __NS_WELLFORMED_DTD
+#define __NS_WELLFORMED_DTD
+
+#include "nsIDTD.h"
+#include "nsISupports.h"
+#include "nsHTMLTokens.h"
+#include "nshtmlpars.h"
+#include "nsVoidArray.h"
+#include "nsDeque.h"
+
+#define NS_WELLFORMED_DTD_IID      \
+  {0xa39c6bfd, 0x15f0,  0x11d2, \
+  {0x80, 0x41, 0x0, 0x10, 0x4b, 0x98, 0x3f, 0xd4}}
+
+
+class nsIHTMLContentSink;
+class nsIDTDDebug;
+class nsIParserNode;
+class CITokenHandler;
+class nsParser;
+
+
+
+class CWellFormedDTD : public nsIDTD {
+            
+  public:
+
+    NS_DECL_ISUPPORTS
+
+
+    /**
+     *  
+     *  
+     *  @update  gess 4/9/98
+     *  @param   
+     *  @return  
+     */
+    CWellFormedDTD();
+
+    /**
+     *  
+     *  
+     *  @update  gess 4/9/98
+     *  @param   
+     *  @return  
+     */
+    virtual ~CWellFormedDTD();
+
+    /**
+     * This method is called to determine if the given DTD can parse
+     * a document in a given source-type. 
+     * NOTE: Parsing always assumes that the end result will involve
+     *       storing the result in the main content model.
+     * @update	gess6/24/98
+     * @param   
+     * @return  TRUE if this DTD can satisfy the request; FALSE otherwise.
+     */
+    virtual PRBool CanParse(nsString& aContentType, PRInt32 aVersion);
+
+    /**
+     * 
+     * @update	gess7/7/98
+     * @param 
+     * @return
+     */
+    virtual eAutoDetectResult AutoDetectContentType(nsString& aBuffer,nsString& aType);
+
+    /**
+     * 
+     * @update	gess5/18/98
+     * @param 
+     * @return
+     */
+    virtual PRInt32 WillBuildModel(const char* aFilename=0);
+
+    /**
+     * 
+     * @update	gess5/18/98
+     * @param 
+     * @return
+     */
+    virtual PRInt32 DidBuildModel(PRInt32 anErrorCode);
+
+    /**
+     *  
+     *  @update  gess 3/25/98
+     *  @param   aToken -- token object to be put into content model
+     *  @return  0 if all is well; non-zero is an error
+     */
+    virtual PRInt32 HandleToken(CToken* aToken);
+
+    /**
+     * 
+     *  
+     *  @update  gess 3/25/98
+     *  @param   
+     *  @return 
+     */
+    virtual void SetParser(nsIParser* aParser);
+
+    /**
+     *  Cause the tokenizer to consume the next token, and 
+     *  return an error result.
+     *  
+     *  @update  gess 3/25/98
+     *  @param   anError -- ref to error code
+     *  @return  new token or null
+     */
+    virtual PRInt32 ConsumeToken(CToken*& aToken);
+
+
+    /**
+     * 
+     * @update	gess5/18/98
+     * @param 
+     * @return
+     */
+    virtual void WillResumeParse(void);
+
+    /**
+     * 
+     * @update	gess5/18/98
+     * @param 
+     * @return
+     */
+    virtual void WillInterruptParse(void);
+
+   /**
+     * Select given content sink into parser for parser output
+     * @update	gess5/11/98
+     * @param   aSink is the new sink to be used by parser
+     * @return  old sink, or NULL
+     */
+    virtual nsIContentSink* SetContentSink(nsIContentSink* aSink);
+
+    /**
+     * 
+     * @update	jevering6/23/98
+     * @param 
+     * @return
+     */
+	  virtual void SetDTDDebug(nsIDTDDebug * aDTDDebug);
+
+    /**
+     *  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 -- int tag of parent container
+     *  @param   aChild -- int tag of child container
+     *  @return  PR_TRUE if parent can contain child
+     */
+    virtual PRBool CanContain(PRInt32 aParent,PRInt32 aChild);
+	
+protected:
+
+    
+    nsParser*           mParser;
+    nsIHTMLContentSink* mSink;
+    char*               mFilename;
+};
+
+extern NS_HTMLPARS nsresult NS_NewWellFormed_DTD(nsIDTD** aInstancePtrResult);
+
+#endif 
+
+
+

mozilla/uriloader/base/nsDocLoader.cpp

@@ -107,6 +107,8 @@ nsDocFactoryImpl::nsDocFactoryImpl()
 NS_DEFINE_IID(kIDocumentLoaderFactoryIID, NS_IDOCUMENTLOADERFACTORY_IID);
 NS_IMPL_ISUPPORTS(nsDocFactoryImpl,kIDocumentLoaderFactoryIID);
 
+static char* gValidTypes[] = {"text/html","text/xml","application/rtf",0};
+
 NS_IMETHODIMP
 nsDocFactoryImpl::CreateInstance(nsIURL* aURL, 
                                  const char* aContentType, 
@@ -118,16 +120,15 @@ nsDocFactoryImpl::CreateInstance(nsIURL* aURL,
     nsIDocument* doc = nsnull;
     nsIWebWidget* ww = nsnull;
 
-    /*
-     * XXX: 
-     *    All of this code should be replaced by a registry and factories
-     *    for each content type...
-     */
-    if (0 != PL_strcmp("text/html", aContentType)) {
-        rv = NS_ERROR_FAILURE;
-        goto done;
+    int typeIndex=0;
+    while(gValidTypes[typeIndex]) {
+      if (0== PL_strcmp(gValidTypes[typeIndex++], aContentType)) {
+        goto nextstep;
+      }
     }
+    goto done;
 
+nextstep:
     /*
      * Create the HTML document...
      */

mozilla/webshell/src/nsDocLoader.cpp

@@ -107,6 +107,8 @@ nsDocFactoryImpl::nsDocFactoryImpl()
 NS_DEFINE_IID(kIDocumentLoaderFactoryIID, NS_IDOCUMENTLOADERFACTORY_IID);
 NS_IMPL_ISUPPORTS(nsDocFactoryImpl,kIDocumentLoaderFactoryIID);
 
+static char* gValidTypes[] = {"text/html","text/xml","application/rtf",0};
+
 NS_IMETHODIMP
 nsDocFactoryImpl::CreateInstance(nsIURL* aURL, 
                                  const char* aContentType, 
@@ -118,16 +120,15 @@ nsDocFactoryImpl::CreateInstance(nsIURL* aURL,
     nsIDocument* doc = nsnull;
     nsIWebWidget* ww = nsnull;
 
-    /*
-     * XXX: 
-     *    All of this code should be replaced by a registry and factories
-     *    for each content type...
-     */
-    if (0 != PL_strcmp("text/html", aContentType)) {
-        rv = NS_ERROR_FAILURE;
-        goto done;
+    int typeIndex=0;
+    while(gValidTypes[typeIndex]) {
+      if (0== PL_strcmp(gValidTypes[typeIndex++], aContentType)) {
+        goto nextstep;
+      }
     }
+    goto done;
 
+nextstep:
     /*
      * Create the HTML document...
      */