Node *Parser::parseDocument()
{
readChar( false );
Node *result = new Node;
result->setName( "" );
result->setNodeType( Document );
if( !parseProlog( result ) )
reportError( "Missing prolog" );
if( !parseElement( result ) )
reportError( "Missing content" );
while( parseMisc( result ) ) {}
return result;
}
void SGMLParser :: parse(const char * aSchemaFile, const char * aDocument) {
printf("Loading schema\n");
clock_t start = clock();
loadSchema(aSchemaFile);
clock_t end = clock();
printf("Time taken for loading the schema: %f\n", (double)(end - start)/CLOCKS_PER_SEC);
start = clock();
printf("Starting to scan the HTML document\n");
mScanner->setDocument(aDocument);
printf("Loaded the document\n");
// Assume the doctype is HTML.
mDocTypeName = "HTML";
ElementParser elementParser(mScanner, mSchema, mDocTypeName);
// See if we can scan a whole HTML document.
try {
mToken = mScanner->nextToken();
parseSStar();
printf("Got first token: %s\n", mScanner->getTokenText().c_str());
parseProlog();
while (mToken != EOF_SYM) {
switch (mToken) {
case ELEMENT_OPEN_SYM: {
// Kickstart the element parser.
TElementPtr element = elementParser.parseStartTag();
TDOMString name = element->getTagName();
ElementToken elmToken = ElementToken(START_TAG, name, element);
TElementDeclarationPtr declaration = mSchema->getDeclaration(mDocTypeName);
mToken = elementParser.parse(elmToken, declaration);
break;
}
case DECLARATION_SYM: {
mToken = mScanner->nextToken();
if (mToken == COMMENT_SYM) {
if (mCommentDeclParser == NULL)
mCommentDeclParser = new CommentDeclParser(mScanner, TSchemaPtr());
mToken = mCommentDeclParser->parse(mToken, ELEMENT_OPEN_SYM);
}
else
throw ReadException(mScanner->getLineNr(),
mScanner->getCharNr(),
"Expected comment sym",
GENERIC,
true);
break;
}
case DECLARATION_END_SYM: {
mToken = mScanner->nextToken(ELEMENT_OPEN_SYM);
break;
}
case TEXT_SYM: {
mToken = mScanner->nextToken();
break;
}
case SPACE_SYM: {
// Not doing anything with that right now.
mToken = mScanner->nextToken();
break;
}
default: {
printf("Found token: %s\n", mScanner->getTokenText().c_str());
mToken = mScanner->nextToken();
}
}
}
}
catch(ReadException r) {
printf(
"Found error: line: %i char %i message: %s\n",
r.getLineNr(),
r.getCharNr(),
r.getErrorMessage().c_str());
}
end = clock();
printf("Time taken: %f\n", (double)(end - start)/CLOCKS_PER_SEC);
TDocumentPtr document = elementParser.getDocument();
showTree(document, 0);
}
static UXMLNODE parseElement(struct UXMLPARSER *parser, UXMLNODE parent, int *type)
{
UXMLNODE node;
UXMLCHAR *name = parser->offset;
size_t nameLength;
int isEmpty;
if (!skipName(parser))
{
setError(parser, 0, "Unexpected end of input parsing stag name");
return NULL;
}
nameLength = (parser->offset - name);
switch (*name)
{
case '!':
{
if (stringCompare(name, "![CDATA[", 8))
{
*type = UXML_NT_CDATA;
return parseCDATA(parser, parent);
}
else if (stringCompare(name, "!--", 3))
{
*type = UXML_NT_COMMENT;
return parseComment(parser, parent);
}
else if (stringCompare(name, "!DOCTYPE", 8))
{
*type = UXML_NT_DOCTYPE;
return parseDocType(parser, parent);
}
else
{
setError(parser, 0, "Unexpected ! tag");
return NULL;
}
break;
}
case '?':
{
if (stringCompare(name, "?xml", 4))
{
*type = UXML_NT_PROLOG;
return parseProlog(parser, parent);
}
else
{
setError(parser, 0, "Unexpected prolog tag");
return NULL;
}
break;
}
default:
{
*type = UXML_NT_ELEMENT;
node = parser->createElement(parser, parent, name, nameLength);
if (!parseAttributes(parser, node, &isEmpty))
{
goto ERROR_DESTROY_NODE;
}
if (isEmpty)
{
parser->offset ++;
if (*(parser->offset) != '>')
{
setError(parser, 0, "Unexpected end of input parsing stag name");
goto ERROR_DESTROY_NODE;
}
parser->offset ++;
}
else
{
parser->offset ++;
if (!parseContent(parser, node, name, nameLength))
{
goto ERROR_DESTROY_NODE;
}
}
break;
}
}
return node;
ERROR_DESTROY_NODE:
parser->destroyNode(parser, node);
return NULL;
}