cpl_minixml.cpp

用于读取TAB、MIF、SHP文件的类

    else if( psContext->bInElement && chNext == '\'' )
    {
        psContext->eTokenType = TString;

        while( (chNext = ReadChar(psContext)) != '\'' 
               && chNext != '\0' )
            AddToToken( psContext, chNext );
        
        if( chNext != '\'' )
        {
            psContext->eTokenType = TNone;
            CPLError( CE_Failure, CPLE_AppDefined, 
                  "Parse error on line %d, reached EOF before closing quote.", 
                      psContext->nInputLine );
        }

        /* Do we need to unescape it? */
        if( strchr(psContext->pszToken,'&') != NULL )
        {
            int  nLength;
            char *pszUnescaped = CPLUnescapeString( psContext->pszToken, 
                                                    &nLength, CPLES_XML );
            strcpy( psContext->pszToken, pszUnescaped );
            CPLFree( pszUnescaped );
            psContext->nTokenSize = strlen(psContext->pszToken );
        }
    }

/* -------------------------------------------------------------------- */
/*      Collect an unquoted string, terminated by a open angle          */
/*      bracket.                                                        */
/* -------------------------------------------------------------------- */
    else if( !psContext->bInElement )
    {
        psContext->eTokenType = TString;

        AddToToken( psContext, chNext );
        while( (chNext = ReadChar(psContext)) != '<' 
               && chNext != '\0' )
            AddToToken( psContext, chNext );
        UnreadChar( psContext, chNext );

        /* Do we need to unescape it? */
        if( strchr(psContext->pszToken,'&') != NULL )
        {
            int  nLength;
            char *pszUnescaped = CPLUnescapeString( psContext->pszToken, 
                                                    &nLength, CPLES_XML );
            strcpy( psContext->pszToken, pszUnescaped );
            CPLFree( pszUnescaped );
            psContext->nTokenSize = strlen(psContext->pszToken );
        }
    }
    
/* -------------------------------------------------------------------- */
/*      Collect a regular token terminated by white space, or           */
/*      special character(s) like an equal sign.                        */
/* -------------------------------------------------------------------- */
    else
    {
        psContext->eTokenType = TToken;

        /* add the first character to the token regardless of what it is */
        AddToToken( psContext, chNext );

        for( chNext = ReadChar(psContext); 
             (chNext >= 'A' && chNext <= 'Z')
                 || (chNext >= 'a' && chNext <= 'z')
                 || chNext == '-'
                 || chNext == '_'
                 || chNext == '.'
                 || chNext == ':'
                 || (chNext >= '0' && chNext <= '9');
             chNext = ReadChar(psContext) ) 
        {
            AddToToken( psContext, chNext );
        }

        UnreadChar(psContext, chNext);
    }
    
    return psContext->eTokenType;
}
    
/************************************************************************/
/*                              PushNode()                              */
/************************************************************************/

static void PushNode( ParseContext *psContext, CPLXMLNode *psNode )

{
    if( psContext->nStackMaxSize <= psContext->nStackSize )
    {
        psContext->nStackMaxSize += 10;
        psContext->papsStack = (CPLXMLNode **)
            CPLRealloc(psContext->papsStack, 
                       sizeof(CPLXMLNode*) * psContext->nStackMaxSize);
    }

    psContext->papsStack[psContext->nStackSize++] = psNode;
}
    
/************************************************************************/
/*                             AttachNode()                             */
/*                                                                      */
/*      Attach the passed node as a child of the current node.          */
/*      Special handling exists for adding siblings to psFirst if       */
/*      there is nothing on the stack.                                  */
/************************************************************************/

static void AttachNode( ParseContext *psContext, CPLXMLNode *psNode )

{
    if( psContext->psFirstNode == NULL )
        psContext->psFirstNode = psNode;
    else if( psContext->nStackSize == 0 )
    {
        CPLXMLNode *psSibling;

        psSibling = psContext->psFirstNode;
        while( psSibling->psNext != NULL )
            psSibling = psSibling->psNext;
        psSibling->psNext = psNode;
    }
    else if( psContext->papsStack[psContext->nStackSize-1]->psChild == NULL )
    {
        psContext->papsStack[psContext->nStackSize-1]->psChild = psNode;
    }
    else
    {
        CPLXMLNode *psSibling;

        psSibling = psContext->papsStack[psContext->nStackSize-1]->psChild;
        while( psSibling->psNext != NULL )
            psSibling = psSibling->psNext;
        psSibling->psNext = psNode;
    }
}

/************************************************************************/
/*                         CPLParseXMLString()                          */
/************************************************************************/

/**
 * \brief Parse an XML string into tree form.
 *
 * The passed document is parsed into a CPLXMLNode tree representation. 
 * If the document is not well formed XML then NULL is returned, and errors
 * are reported via CPLError().  No validation beyond wellformedness is
 * done.  The CPLParseXMLFile() convenience function can be used to parse
 * from a file. 
 *
 * The returned document tree is is owned by the caller and should be freed
 * with CPLDestroyXMLNode() when no longer needed.
 *
 * If the document has more than one "root level" element then those after the 
 * first will be attached to the first as siblings (via the psNext pointers)
 * even though there is no common parent.  A document with no XML structure
 * (no angle brackets for instance) would be considered well formed, and 
 * returned as a single CXT_Text node.  
 * 
 * @param pszString the document to parse. 
 *
 * @return parsed tree or NULL on error. 
 */

CPLXMLNode *CPLParseXMLString( const char *pszString )

{
    ParseContext sContext;

    CPLErrorReset();

    if( pszString == NULL )
    {
        CPLError( CE_Failure, CPLE_AppDefined, 
                  "CPLParseXMLString() called with NULL pointer." );
        return NULL;
    }

/* -------------------------------------------------------------------- */
/*      Initialize parse context.                                       */
/* -------------------------------------------------------------------- */
    sContext.pszInput = pszString;
    sContext.nInputOffset = 0;
    sContext.nInputLine = 0;
    sContext.bInElement = FALSE;
    sContext.pszToken = NULL;
    sContext.nTokenMaxSize = 0;
    sContext.nTokenSize = 0;
    sContext.eTokenType = TNone;
    sContext.nStackMaxSize = 0;
    sContext.nStackSize = 0;
    sContext.papsStack = NULL;
    sContext.psFirstNode = NULL;

    /* ensure token is initialized */
    AddToToken( &sContext, ' ' );
    
/* ==================================================================== */
/*      Loop reading tokens.                                            */
/* ==================================================================== */
    while( ReadToken( &sContext ) != TNone )
    {
/* -------------------------------------------------------------------- */
/*      Create a new element.                                           */
/* -------------------------------------------------------------------- */
        if( sContext.eTokenType == TOpen )
        {
            CPLXMLNode *psElement;

            if( ReadToken(&sContext) != TToken )
            {
                CPLError( CE_Failure, CPLE_AppDefined, 
                          "Line %d: Didn't find element token after open angle bracket.",
                          sContext.nInputLine );
                break;
            }

            if( sContext.pszToken[0] != '/' )
            {
                psElement = CPLCreateXMLNode( NULL, CXT_Element,
                                              sContext.pszToken );
                AttachNode( &sContext, psElement );
                PushNode( &sContext, psElement );
            }
            else 
            {
                if( sContext.nStackSize == 0
                    || !EQUAL(sContext.pszToken+1,
                         sContext.papsStack[sContext.nStackSize-1]->pszValue) )
                {
                    CPLError( CE_Failure, CPLE_AppDefined, 
                              "Line %d: <%.500s> doesn't have matching <%.500s>.",
                              sContext.nInputLine,
                              sContext.pszToken, sContext.pszToken+1 );
                    break;
                }
                else
                {
                    if( ReadToken(&sContext) != TClose )
                    {
                        CPLError( CE_Failure, CPLE_AppDefined, 
                                  "Line %d: Missing close angle bracket after <%.500s.",
                                  sContext.nInputLine,
                                  sContext.pszToken );
                        break;
                    }

                    /* pop element off stack */
                    sContext.nStackSize--;
                }
            }
        }

/* -------------------------------------------------------------------- */
/*      Add an attribute to a token.                                    */
/* -------------------------------------------------------------------- */
        else if( sContext.eTokenType == TToken )
        {
            CPLXMLNode *psAttr;

            psAttr = CPLCreateXMLNode(NULL, CXT_Attribute, sContext.pszToken);
            AttachNode( &sContext, psAttr );
            
            if( ReadToken(&sContext) != TEqual )
            {
                CPLError( CE_Failure, CPLE_AppDefined, 
                          "Line %d: Didn't find expected '=' for value of attribute '%.500s'.",
                          sContext.nInputLine, psAttr->pszValue );
                break;
            }

            if( ReadToken(&sContext) != TString 
                && sContext.eTokenType != TToken )
            {
                CPLError( CE_Failure, CPLE_AppDefined, 
                          "Line %d: Didn't find expected attribute value.",
                          sContext.nInputLine );
                break;
            }

            CPLCreateXMLNode( psAttr, CXT_Text, sContext.pszToken );
        }

/* -------------------------------------------------------------------- */
/*      Close the start section of an element.                          */
/* -------------------------------------------------------------------- */
        else if( sContext.eTokenType == TClose )
        {
            if( sContext.nStackSize == 0 )
            {
                CPLError( CE_Failure, CPLE_AppDefined, 
                          "Line %d: Found unbalanced '>'.",
                          sContext.nInputLine );
                break;
            }
        }

/* -------------------------------------------------------------------- */
/*      Close the start section of an element, and pop it               */
/*      immediately.                                                    */
/* -------------------------------------------------------------------- */
        else if( sContext.eTokenType == TSlashClose )
        {
            if( sContext.nStackSize == 0 )
            {
                CPLError( CE_Failure, CPLE_AppDefined, 
                          "Line %d: Found unbalanced '/>'.",
                          sContext.nInputLine );
                break;
            }

            sContext.nStackSize--;
        }

/* -------------------------------------------------------------------- */
/*      Close the start section of a <?...?> element, and pop it        */
/*      immediately.                                                    */
/* -------------------------------------------------------------------- */
        else if( sContext.eTokenType == TQuestionClose )
        {
            if( sContext.nStackSize == 0 )
            {
                CPLError( CE_Failure, CPLE_AppDefined, 
                          "Line %d: Found unbalanced '?>'.",
                          sContext.nInputLine );
                break;
            }
            else if( sContext.papsStack[sContext.nStackSize-1]->pszValue[0] != '?' )
            {
                CPLError( CE_Failure, CPLE_AppDefined, 
                          "Line %d: Found '?>' without matching '<?'.",
                          sContext.nInputLine );
                break;
            }

            sContext.nStackSize--;
        }

/* -------------------------------------------------------------------- */
/*      Handle comments.  They are returned as a whole token with the     */
/*      prefix and postfix omitted.  No processing of white space       */
/*      will be done.                                                   */
/* -------------------------------------------------------------------- */
        else if( sContext.eTokenType == TComment )
        {
            CPLXMLNode *psValue;

            psValue = CPLCreateXMLNode(NULL, CXT_Comment, sContext.pszToken);
            AttachNode( &sContext, psValue );
        }

/* -------------------------------------------------------------------- */
/*      Handle literals.  They are returned without processing.         */
/* -------------------------------------------------------------------- */
        else if( sContext.eTokenType == TLiteral )
        {
            CPLXMLNode *psValue;

            psValue = CPLCreateXMLNode(NULL, CXT_Literal, sContext.pszToken);
            AttachNode( &sContext, psValue );
        }

/* -------------------------------------------------------------------- */
/*      Add a text value node as a child of the current element.        */
/* -------------------------------------------------------------------- */
        else if( sContext.eTokenType == TString && !sContext.bInElement )
        {
            CPLXMLNode *psValue;

            psValue = CPLCreateXMLNode(NULL, CXT_Text, sContext.pszToken);
            AttachNode( &sContext, psValue );
        }