htmlparse.cpp

hl2 source code. Do not use it illegal.

/*----------------------------------------------------------------------
Copyright (c) 1998 Gipsysoft. All Rights Reserved.
Please see the file "licence.txt" for licencing details.
File:	htmlparse.cpp
Owner:	[e-mail]
Purpose:	Main HTML parser
					The CHTMLParse::Parse() function generates a document, the document
					contains all of the elements of the HTML page but broken into
					their parts.
					The document will then be used to create the display page.
----------------------------------------------------------------------*/
#include "stdafx.h"
#include <WinHelper.h>
#include <ImgLib.h>
#include "QHTM_Includes.h"
#include "AquireImage.h"
#include "DrawContext.h"
#include "defaults.h"
#include "smallstringhash.h"
#include "HTMLParse.h"

extern LPTSTR stristr( LPTSTR pszSource, LPCTSTR pcszSearch );
extern BYTE DecodeCharset( const CStaticString &strCharSet );

static CHTMLParse::Align knDefaultImageAlignment = CHTMLParse::algBottom;
static CHTMLParse::Align knDefaultHRAlignment = CHTMLParse::algLeft;
static CHTMLParse::Align knDefaultParagraphAlignment = CHTMLParse::algLeft;
// richg - 19990224 - Default table alignment changed form algLeft to algTop
static CHTMLParse::Align knDefaultTableAlignment = CHTMLParse::algMiddle;
// richg - 19990224 - Table Cells have their own default
static CHTMLParse::Align knDefaultTableCellAlignment = CHTMLParse::algLeft;

TCHAR g_cCarriageReturn = _T('\r');
static const CStaticString g_strTabSpaces( _T("        ") );

static MapClass< StringClass, bool> g_mapFontName;

CHTMLParse::CHTMLParse( LPCTSTR pcszStream, UINT uLength, HINSTANCE hInstLoadedFrom, LPCTSTR pcszFilePath, CDefaults *pDefaults )
	: CHTMLParseBase( pcszStream, uLength )
	, m_pProp( NULL )
	, m_pDocument( NULL )
	, m_hInstLoadedFrom( hInstLoadedFrom )
	, m_pcszFilePath( pcszFilePath )
	, m_pLastAnchor( NULL )
	, m_pDefaults( pDefaults )
{
	ASSERT( m_pDefaults );
}


CHTMLParse::~CHTMLParse()
{
	CleanupParse();
}


CHTMLDocument * CHTMLParse::Parse()
//
//	returns either a fully created document ready to create the display from or
//	 NULL in the event of failure.
{
	CleanupParse();

	//
	//	Create the first properties
	m_pProp = new Properties;
	m_pProp->m_crFore = m_pDefaults->m_crDefaultForeColour;
	_tcscpy( m_pProp->szFaceName, m_pDefaults->m_strFontName );
	m_pProp->nSize = m_pDefaults->m_nFontSize;
	m_stkProperties.Push( m_pProp );

	//
	//	Create the main document and a paragraph to add to it.
	m_pMasterDocument = m_pDocument = new CHTMLDocument( m_pDefaults );
	m_pDocument->m_crBack = m_pDefaults->m_crBackground;
	CreateNewParagraph( m_pDefaults->m_nParagraphLinesAbove, m_pDefaults->m_nParagraphLinesBelow, knDefaultParagraphAlignment );

	if( !ParseBase() )
	{
		if( m_pMasterDocument )
		{
			delete m_pMasterDocument;
			m_pDocument = NULL;
			m_pMasterDocument = NULL;
		}
	}

	return m_pMasterDocument;
}


void CHTMLParse::CleanupParse()
//
//	Cleanup anything left over from our previous parsing
{
	while( m_stkProperties.GetSize() )
		delete m_stkProperties.Pop();

	while( m_stkDocument.GetSize() )
		m_stkDocument.Pop();


	//	richg - 19990227 - Clean up the table stack as well
	while( m_stkTable.GetSize() )
		m_stkTable.Pop();

	while( m_stkInTableCell.GetSize() )
		m_stkInTableCell.Pop();

	// richg - 19990621 - Clean up list stack
	while( m_stkList.GetSize() )
		m_stkList.Pop();
	
	m_pLastAnchor = NULL;
}


void CHTMLParse::OnGotText( TCHAR ch )
//
//	Callback when some text has been interrupted with a tag or end tag.
{
	if( ch == _T('\t') )
	{
		m_strToken.Add( g_strTabSpaces, g_strTabSpaces.GetLength() );
	}
	else
	{
		m_strToken.Add( ch );
	}
}


void CHTMLParse::OnEndDoc()
{
	CreateNewTextObject();
}


void CHTMLParse::OnGotTag( const Token token, const CParameters &pList )
{
	switch( token )
	{
	case tokIgnore:
		break;

	case tokSub:
		CreateNewTextObject();
		CreateNewProperties();
		if( m_pProp->nSize > 1 )
			m_pProp->nSize--;
		m_pProp->m_nSub++;
		break;

	case tokSup:
		CreateNewTextObject();
		CreateNewProperties();

		if( m_pProp->nSize > 1 )
			m_pProp->nSize--;
		m_pProp->m_nSup++;
		break;

	case tokPre:
		CreateNewProperties();
		_tcscpy( m_pProp->szFaceName, m_pDefaults->m_strDefaultPreFontName );
		break;

	case tokBody:
		OnGotBody( pList );
		break;

	case tokFont:
		OnGotFont( pList );
		break;

	case tokBold:
		CreateNewTextObject();
		CreateNewProperties();
		m_pProp->bBold = true;
		break;

	case tokUnderline:
		CreateNewTextObject();
		CreateNewProperties();
		m_pProp->bUnderline = true;
		break;

	case tokItalic:
		CreateNewTextObject();
		CreateNewProperties();
		m_pProp->bItalic = true;
		break;

	case tokStrikeout:
		CreateNewTextObject();
		CreateNewProperties();
		m_pProp->bStrikeThrough = true;
		break;

	case tokImage:
		OnGotImage( pList );
		break;

	case tokTableHeading:
	case tokTableDef:
		OnGotTableCell( pList );
		break;

	case tokTableRow:
		OnGotTableRow( pList );
		break;

	case tokTable:
		CreateNewProperties();
		OnGotTable( pList );
		break;

	case tokHorizontalRule:
		OnGotHR( pList );
		break;

	case tokCenter:
		CreateNewProperties();
		m_pProp->nAlignment = algCentre;
		OnGotParagraph( pList );
		break;

	case tokDiv:
		OnGotParagraph( pList );
		break;

	case tokParagraph:
		OnGotParagraph( pList );
		break;

	case tokBreak:
		m_strToken.Add( g_cCarriageReturn );
		break;

	case tokAnchor:
		OnGotAnchor( pList );
		break;

	case tokH1:
	case tokH2:
	case tokH3:
	case tokH4:
	case tokH5:
	case tokH6:
		OnGotHeading( token, pList );
		break;

	case tokOrderedList:
		OnGotOrderedList( pList );
		break;

	case tokUnorderedList:
		OnGotUnorderedList( pList );
		break;

	case tokListItem:
		OnGotListItem( pList );
		break;

	case tokAddress:
		OnGotAddress( pList );
		break;

	case tokBlockQuote:
		OnGotBlockQuote( pList );
		break;

	case tokCode:
		CreateNewTextObject();
		CreateNewProperties();
		_tcscpy( m_pProp->szFaceName, m_pDefaults->m_strDefaultPreFontName );
		break;

	case tokMeta:
		OnGotMeta( pList );
		break;
	}
}


void CHTMLParse::OnGotEndTag( const Token token )
{
	switch( token )
	{
	case tokCode:
		CreateNewTextObject();
		PopPreviousProperties();
		break;

	case tokPre:
		CreateNewTextObject();
		PopPreviousProperties();
		CreateNewParagraph( 0, 0, m_pProp->nAlignment );
		break;

	case tokCenter:
		CreateNewTextObject();
		PopPreviousProperties();
		CreateNewParagraph( 0, 0, m_pProp->nAlignment );
		break;
		
	case tokParagraph:
		CreateNewTextObject();
		CreateNewParagraph( 0, 0, m_pProp->nAlignment );
		break;

	case tokImage:				//	Ignore end images
		//	There is no end image tag!
		ASSERT( FALSE );
		break;

	case tokDiv:
	case tokIgnore:
		CreateNewTextObject();
		CreateNewParagraph( 0, 0, m_pProp->nAlignment );
		break;

	case tokFont:
	case tokBold:
	case tokUnderline:
	case tokItalic:
	case tokStrikeout:
	case tokSub:
	case tokSup:
		CreateNewTextObject();
		PopPreviousProperties();
		break;

	case tokH1:
	case tokH2:
	case tokH3:
	case tokH4:
	case tokH5:
	case tokH6:
		CreateNewTextObject();
		PopPreviousProperties();
		CreateNewParagraph( 0, 0, m_pProp->nAlignment );
		break;

	case tokAnchor:
		OnGotEndAnchor();
		break;

	case tokTableHeading:
	case tokTableDef:
		OnGotEndTableCell();
		break;

	case tokTableRow:
		OnGotEndTableRow();
		break;

	//
	//	IE seems to ne okay about using an end BR (</br>) so we should too.
	case tokBreak:
		m_strToken.Add( g_cCarriageReturn );
		break;

	case tokTable:
		if( m_stkTable.GetSize() )
		{
			m_stkTable.Pop();
		}
		else
		{
			TRACE( _T("Got an table but no tables left in the stack\n") );
		}

		// Pop the InACell flag
		if (m_stkInTableCell.GetSize())
		{
			// If the cell is still open... close it!
			OnGotEndTableRow();		// Does the same thing.
			m_stkInTableCell.Pop();
		}
		else
		{
			TRACE( _T("Got an end table but in-cell stack empty.") );
		}
		PopPreviousProperties();
		CreateNewParagraph( 0, 0, m_pProp->nAlignment );
		break;

	case tokHorizontalRule:
		break;

	case tokOrderedList:
	case tokUnorderedList:
		OnGotEndList();
		break;

	case tokListItem:
		OnGotEndListItem();
		break;

	case tokAddress:
		OnGotEndAddress();
		break;

	case tokBlockQuote:
		OnGotEndBlockQuote();
		break;

	case tokDocumentTitle:
		m_strToken.Add( 0 );
		m_pMasterDocument->m_strTitle = m_strToken.GetData();
		m_strToken.SetSize( 0 );
		break;
	}
}


void CHTMLParse::OnGotBody( const CParameters &pList )
{
	const UINT uParamSize = pList.GetSize();
	for( UINT n = 0; n < uParamSize; n++ )
	{
		const CStaticString &strParam = pList[n].m_strValue;
		switch( pList[n].m_param )
		{
		case pBColor:
			m_pDocument->m_crBack = GetColourFromString( strParam, m_pDocument->m_crBack );
			break;
		case pLink:
			m_pDocument->m_crLink = GetColourFromString( strParam, RGB( 141, 7, 102 ) );
			break;
		case pALink:
			m_pDocument->m_crLinkHover = GetColourFromString( strParam, RGB( 29, 49, 149 ) );
			break;
		case pMarginTop:
			m_pDocument->m_nTopMargin = GetNumberParameter( strParam, m_pDocument->m_nTopMargin );
			break;
		case pMarginBottom:
			m_pDocument->m_nBottomMargin = GetNumberParameter( strParam, m_pDocument->m_nBottomMargin );
			break;
		case pMarginLeft:
			m_pDocument->m_nLeftMargin = GetNumberParameter( strParam, m_pDocument->m_nLeftMargin );
			break;
		case pMarginRight:
			m_pDocument->m_nRightMargin = GetNumberParameter( strParam, m_pDocument->m_nRightMargin );
			break;
		}
	}
}


void CHTMLParse::OnGotImage( const CParameters &pList )
{
	CreateNewTextObject();
	int nHeight = 0;
	int nWidth = 0;
	int nBorder = 0;
	CStaticString strFilename;
	Align alg = knDefaultImageAlignment;

	const UINT uParamSize = pList.GetSize();
	for( UINT n = 0; n < uParamSize; n++ )
	{
		const CStaticString &strParam = pList[n].m_strValue;
		switch( pList[n].m_param )
		{
		case pWidth:
			nWidth = GetNumberParameter( strParam, nWidth );
			break;