/*

* Copyright (c) 2005 Nokia Corporation and/or its subsidiary(-ies).

* All rights reserved.

* This component and the accompanying materials are made available

* under the terms of the License "Eclipse Public License v1.0"

* which accompanies this distribution, and is available

* at the URL "http://www.eclipse.org/legal/epl-v10.html".

*

* Initial Contributors:

* Nokia Corporation - initial contribution.

*

* Contributors:

*

* Description:  Misc. libxml2 related utilities.

*

*/

#include <charconv.h>

#include <EscapeUtils.h>

#include <utf.h>

#include "CleanupLibXml2.h"  

#include "Logger.h"

#include "XmlEntity.h"

// Local data-types

struct ErrorContext

    {

    TInt   error;

    TBool  retry;

    };

// Static variables

static CXmlEntity*  sXmlEntity = NULL;

// -----------------------------------------------------------------------------

// EntityResolverL

//

// A local function to resolve entities found in the buffer being parsed. 

// -----------------------------------------------------------------------------

//

static xmlEntityPtr EntityResolverL(void* /*aContext*/, const xmlChar *aName) 

    {

    xmlEntityPtr  entity = NULL;

    // Use the standard resolver.

    entity = xmlGetPredefinedEntity(aName);

    // Resolve other entities too.

    if ((sXmlEntity != NULL) && (entity == NULL))

        {

        entity = const_cast<xmlEntityPtr>(sXmlEntity->ResolveL(aName));

        }

    return entity;

    }

// -----------------------------------------------------------------------------

// StructuredErrorReporter

//

// A local function to collect errors found while parsing the buffer.

// See http://www.xmlsoft.org/html/libxml-xmlerror.html.

// -----------------------------------------------------------------------------

//

static void StructuredErrorReporter(void* aContext, xmlErrorPtr error)

    {

    _LIT(KErr, "error");

    TBool          ignore = EFalse;

    TPtrC          type(KErr);

    ErrorContext*  context = static_cast<ErrorContext*>(aContext);

    if (context != NULL)

        {        

        context->error = error->code; 

        switch (error->code)

            {

            case XML_ERR_UNSUPPORTED_ENCODING:          // 32

                // Try to reparse the buffer with another char-encoding.

                context->retry = ETrue;

                ignore = ETrue;

                break;

            case XML_ERR_UNDECLARED_ENTITY:             // 26

            case XML_WAR_UNDECLARED_ENTITY:             // 27

                type.Set(_L("warning"));

                break;

            case XML_ERR_INTERNAL_ERROR:                // 1

            case XML_ERR_DOCUMENT_EMPTY:                // 4

            case XML_ERR_DOCUMENT_END:                  // 5

            case XML_ERR_INVALID_CHAR:                  // 9

            case XML_ERR_INVALID_DEC_CHARREF:           // 7

            case XML_ERR_ENTITYREF_SEMICOL_MISSING:     // 23

            case XML_NS_ERR_QNAME:                      // 38

            case XML_ERR_LT_IN_ATTRIBUTE:               // 39

            case XML_ERR_ATTRIBUTE_NOT_STARTED:         // 41

            case XML_ERR_ATTRIBUTE_WITHOUT_VALUE:       // 42

            case XML_ERR_ATTRIBUTE_REDEFINED:           // 43

            case XML_ERR_LITERAL_NOT_STARTED:           // 46

            case XML_ERR_PI_NOT_STARTED:                // 47

            case XML_ERR_PI_NOT_FINISHED:               // 61

            case XML_ERR_DOCTYPE_NOT_FINISHED:          // 62

            case XML_ERR_MISPLACED_CDATA_END:           // 63

            case XML_ERR_RESERVED_XML_NAME:             // 64

            case XML_ERR_SPACE_REQUIRED:                // 65

            case XML_ERR_NAME_REQUIRED:                 // 68

            case XML_ERR_URI_REQUIRED:                  // 70

            case XML_ERR_GT_REQUIRED:                   // 73

            case XML_ERR_LTSLASH_REQUIRED:              // 74

            case XML_ERR_HYPHEN_IN_COMMENT:             // 80

            case XML_ERR_TAG_NAME_MISMATCH:             // 76

            case XML_ERR_TAG_NOT_FINISHED:              // 77

            case XML_NS_ERR_UNDEFINED_NAMESPACE:        // 201

                // Just log these as errors and continue if possible.

                break;

            default:

                break;

            }

        }

    if (ignore == EFalse)

        {

        FEED_LOG2(_L("Feeds"), _L("Feeds.log"), 

                EFileLoggingModeAppend, _L("FEEDS: parser %S!!! Code:%d"), &type, error->code);

        }

    }

// -----------------------------------------------------------------------------

// CXmlUtils::NewL

//

// Two-phased constructor.

// -----------------------------------------------------------------------------

//

CXmlUtils* CXmlUtils::NewL()

    {

    CXmlUtils* self = new (ELeave) CXmlUtils();

    CleanupStack::PushL(self);

    self->ConstructL();

    CleanupStack::Pop();

    return self;

    }

// -----------------------------------------------------------------------------

// CXmlUtils::CXmlUtils

//

// C++ default constructor.

// -----------------------------------------------------------------------------

//

CXmlUtils::CXmlUtils():

        iLeakTracker(CLeakTracker::ELibXml2)

    {

    }

// -----------------------------------------------------------------------------

// CXmlEncoding::ConstructL

//

// Symbian 2nd phase constructor can leave.

// -----------------------------------------------------------------------------

//

void CXmlUtils::ConstructL()

    {

    // will call XmlEngine::XmlEngineAttachL()

    iImpl.OpenL();

    // Create the XmlEncoding singleton.

    iXmlEncoding = CXmlEncoding::NewL();

    // Create the XmlEntity singleton.

    iXmlEntity = CXmlEntity::NewL();

    }

// -----------------------------------------------------------------------------

// CXmlUtils::~CXmlUtils

//

// Deconstructor.

// -----------------------------------------------------------------------------

//

CXmlUtils::~CXmlUtils()

    {

    delete iXmlEncoding;

    delete iXmlEntity;

    // will call XmlEngine::XmlEngineCleanup()

    iImpl.Close();

    }

// -----------------------------------------------------------------------------

// CXmlUtils::ParseBufferL

//

// Returns a libxml2 document from the provided buffer.

// -----------------------------------------------------------------------------

//

RXmlEngDocument CXmlUtils::ParseBufferL(const TDesC8& aBuffer, const TDesC& aCharSet) const

    {

    xmlDocPtr         document = NULL;

    xmlParserCtxtPtr  contxt = NULL;

    ErrorContext      errorContext;

    HBufC8*           utf8Buffer = NULL;

    const xmlChar*    parseBuffer = NULL;

    TInt              parseLength = 0;

    TUint             encoding;

    RXmlEngDocument   doc;

    if (aBuffer.Length() == 0)

        {

        User::Leave(KErrArgument);

        }

    // Prepare the parser's context.

  	User::LeaveIfNull(contxt = xmlNewParserCtxt());

    sXmlEntity = iXmlEntity;

    contxt->sax->getEntity = EntityResolverL;

    CleanupStack::PushL(TCleanupItem(&CleanupParseBuffer, contxt));

    // Set the error callback

    xmlSetStructuredErrorFunc(&errorContext, StructuredErrorReporter);

    // If need be convert the buffer to utf8.

    if (iXmlEncoding->DetermineCharEncodingL(aBuffer, aCharSet, encoding))

        {

        if (encoding != KCharacterSetIdentifierUtf8)

            {

            utf8Buffer = iXmlEncoding->ConvertToUtf8L(encoding, aBuffer);

            CleanupStack::PushL(utf8Buffer);

            // Skip anything before the xml prolog.

            parseBuffer = utf8Buffer->Ptr();

            parseLength = utf8Buffer->Length();

            SkipCharsBeforeXmlProlog(&parseBuffer, parseLength);

            }

        }

    // Otherwise its already utf8 so use the provided buffer.

    if (utf8Buffer == NULL)

        {

        // Skip anything before the xml prolog.

        parseBuffer = aBuffer.Ptr();

        parseLength = aBuffer.Length();

        SkipCharsBeforeXmlProlog(&parseBuffer, parseLength);

        }

    // Parse the document.

    errorContext.error = XML_ERR_OK;

    errorContext.retry = EFalse;        

    document = xmlSAXParseMemory(contxt->sax, (char*) parseBuffer, parseLength, FALSE);

    if (document == NULL)

        {

        User::Leave(KErrCorrupt);

        }

    if (utf8Buffer != NULL)

        {

        CleanupStack::PopAndDestroy(utf8Buffer);

        }

    CleanupStack::PopAndDestroy(/*contxt*/);    

    // Wrap and return the document.

    CleanupStack::PushL(document);

    doc.OpenL( (RXmlEngDOMImplementation&)iImpl, document );

    CleanupStack::Pop(/*document*/);

    return doc;

    }

// -----------------------------------------------------------------------------

// CXmlUtils::CleanupParseBuffer

// 

// The cleanup method for ParseBufferL

// -----------------------------------------------------------------------------

//

void CXmlUtils::CleanupParseBuffer(TAny *aPtr)

    {

    xmlResetLastError();

    xmlSetGenericErrorFunc(NULL, NULL);

    xmlFreeParserCtxt(static_cast<xmlParserCtxtPtr>(aPtr));

    }

// -----------------------------------------------------------------------------

// CXmlUtils::SkipCharsBeforeXmlProlog

// 

// Skip any chars before the XML-prolog.

// -----------------------------------------------------------------------------

//

void CXmlUtils::SkipCharsBeforeXmlProlog(const TUint8** aString, TInt& aLen) const

    {

    _LIT8(KStart, "<?xml");

    const xmlChar*  origStr = *aString;

    TInt            origLen = aLen;

    while (aLen > 0)

        {

        if ((aLen >= 5) && (xmlStrncasecmp(*aString, KStart().Ptr(), 5) == 0))

            {

            break;

            }

        else

            {

            (*aString)++;

            aLen--;

            }

        }

    // If the xml prolog is missing then do nothing.

    if (aLen == 0)

        {

        *aString = origStr;

        aLen = origLen;

        }

    }

// -----------------------------------------------------------------------------

// CXmlUtils::CleanupUrlL

// 

// Resolves any entities and escaped chars in the given url.  

// -----------------------------------------------------------------------------

//

void CXmlUtils::CleanupUrlL(TDes& aUrl) const

    {

    HBufC*  temp = NULL;

    // Resolve entities.

    (void) ResolveEntitiesL(aUrl);

    // Unescape the url.

    temp = EscapeUtils::EscapeDecodeL(aUrl);

    CleanupStack::PushL(temp);

    // Trim leading and trailing whitespace.

    temp->Des().Trim();

    // Copy temp back to aUrl.  This is always safe because temp->Des().Length()

    // is always <= aUrl.Length().

    aUrl.Copy(*temp);

    CleanupStack::PopAndDestroy(temp);

    }    

// -----------------------------------------------------------------------------

// CXmlUtils::CleanupMarkupL

// 

// Resolves any html entities and removes any markup found in the given descriptor.

// -----------------------------------------------------------------------------

//

TBool CXmlUtils::CleanupMarkupL(TDes& aBuffer, TInt aNewLineChar) const

    {

    TBool   modified = EFalse;

    // Strip all markup.

    modified |= StripMarkupL(aBuffer, aNewLineChar);

    // Resolve entities.

    modified |= ResolveEntitiesL(aBuffer);

    // Strip the CDATA markers.

    modified |= StripCDataMarkers(aBuffer);

    // Strip all markup created after entities were resolved.

    modified |= StripMarkupL(aBuffer, aNewLineChar);

    // Strip leading and trailing whitespace.

    aBuffer.Trim();

    return modified;

    }

// -----------------------------------------------------------------------------

// CXmlUtils::ResolveEntitiesL

// 

// Resolves any entities found in aOrig.

// -----------------------------------------------------------------------------

//

TBool CXmlUtils::ResolveEntitiesL(TDes& aBuffer) const

	{

    const TUint  KEntityStart =	'&';

    const TUint  KEntityEnd = ';';

	TInt         entityStart = -1;

    TInt         currentPos =  0;    

    TBool        entityFound = EFalse;

	TBool        textHasChanged = EFalse;

    while (currentPos < aBuffer.Length())

        {

        // If this is the beginning of an entity...

        if (aBuffer[currentPos] == KEntityStart) 

            {

            entityStart = currentPos;

            entityFound = ETrue;

            currentPos++;

            }

        // If this is the end of an entity...

        else if ((aBuffer[currentPos] == KEntityEnd) && entityFound)

            {

            TUint16    ucs2Value;

            TInt       entityLength = currentPos - entityStart + 1;

            TPtrC      entityStr(aBuffer.Mid(entityStart + 1, entityLength - 2));

            TBuf16<1>  entityCode;

            // Resolve the entity inline.

            iXmlEntity->ResolveL(entityStr, ucs2Value);            

            entityCode.SetLength(1);

            entityCode[0] = ucs2Value;

            aBuffer.Replace(entityStart, entityLength, entityCode);

            currentPos = entityStart;

			textHasChanged = ETrue;

            entityStart = -1;

            entityFound = EFalse;

            }

        // Otherwise just move on.

        else

            {

            currentPos++;

            }

        }

	return textHasChanged;

	}

// -----------------------------------------------------------------------------

// CXmlUtils::StripMarkup

// 

// Remove any markup found in the given descriptor.

// -----------------------------------------------------------------------------

//

TBool CXmlUtils::StripMarkupL(TDes& aBuffer, TInt aNewLineChar) const

    {

	TInt    tagStart = -1;

    TInt    currentPos = 0;    

    TBool   tagStartFound = EFalse;

	TBool   textHasChanged = EFalse;

    // TODO: 1) Remove xml comments as well.

    while (currentPos < aBuffer.Length())

        {

        // It's at the beginning of a tag.

        if (aBuffer.Mid(currentPos, 1).Compare(_L("<")) == 0)

            {

            tagStartFound = ETrue;

            tagStart = currentPos;

            currentPos++;

            }

        // It's at the end of a tag.

        else if ((aBuffer.Mid(currentPos,1).Compare(_L(">")) == 0) && 

                tagStartFound)

            {

			TBuf<6>  tag;

			TInt     tagLength;

            // Extract enough of the tag to determine if it needs to insert a

            // newline.

            tagLength = Min(6, currentPos - tagStart + 1);

			tag.Copy(aBuffer.Mid(tagStart, tagLength));

			tag.LowerCase();

            // If it found a <p> or <br> tag insert a newline.

            // TODO: This fails if the tag includes attributes (i.e. style).

			if ((aNewLineChar != 0) && 

			        ((tag.Compare(_L("<br>")) == 0) ||

                    (tag.Compare(_L("<br/>")) == 0) ||

                    (tag.Compare(_L("<br />")) == 0) ||

                    (tag.Compare(_L("<p>")) == 0)))

				{

                // Delete the tag.

				aBuffer.Delete(tagStart, currentPos - tagStart + 1);

                // Insert the newline.

				TBuf<1> chr;

				_LIT(KFormat,"%c");

				chr.Format(KFormat, aNewLineChar);

				aBuffer.Insert(tagStart, chr);

				currentPos = tagStart + 1;

				textHasChanged = ETrue;

				}

            // Otherwise just delete the tag.

			else

				{

				aBuffer.Delete(tagStart, currentPos - tagStart + 1);

				currentPos = tagStart;

				textHasChanged = ETrue;

				}

            tagStart = -1;

			tagStartFound = EFalse;

            }

        // Otherwise just skip to the next char.

        else

            {

            currentPos++;

            }

        }