public static DocumentNode ParseWellFormedXml(string text, List<long> tags)
{
var root = new DocumentNode();
ElementNode current = root;
foreach (var tag in tags)
{
int index = ((int)tag << 8) >> 8;
int length = (int)((tag << 16) >> (24 + 16));
ushort other = (ushort)(tag >> 64 - 16);
if (other == 0)
{
var node = new TextNode(text.Substring(index, length));
current.AppendChild(node);
}
else if (other == ushort.MaxValue)
{
var node = new CommentNode(text.SubstringTrim(index + 4, length - 7));
current.AppendChild(node);
}
else
{
bool isClosingTag = text[index + 1] == '/';
if (isClosingTag)
{
current = current.Parent;
}
else
{
bool isSelfClosingTag = text[index + length - 2] == '/';
string tagName = text.Substring(index + 1, other);
var node = new ElementNode(tagName, isSelfClosingTag);
int attributeStart = index + other + 1;
ParseAttributesFromWellFormedXml(node, text, attributeStart, length - (attributeStart - index) - 1);
current.AppendChild(node);
if (!isSelfClosingTag)
{
current = node;
}
}
}
}
return root;
}
public static DocumentNode ParseTagsToTree(TokenList tokens, List<TokenGroup> taglist)
{
// temporary variables
var root = new DocumentNode();
ElementNode current = root;
Node node = null;
// loop through final tag array and create the tree
// this is separate from setting the depth so that errors can
// be isolated in that section
foreach(var tag in taglist){
node = null;
if(tag.TreeDepth != -1) {
if(tag.Type == NodeType.ELEMENT_NODE) {
if(tag.IsClosing) {
current = current.Parent;
} else if(tag.IsEmpty) {
node = new ElementNode(tokens[tag.TokenStartIndex].Value, true);
current.AppendChild(node);
} else {
var elementNode = new ElementNode(tokens[tag.TokenStartIndex].Value, false);
node = elementNode;
current.AppendChild(node);
current = elementNode;
}
if(node != null) {
// add attributes
for(int j = tag.TokenStartIndex + 1; j <= tag.TokenEndIndex; j++) {
if(tokens[j].Type == TokenType.CDATA) {
var attribute = new NodeAttribute(tokens[j].Value, null);
++j;
if( j < tag.TokenEndIndex &&
tokens[j].Type == TokenType.EQUALS_SIGN ) {
++j;
if(tokens[j].Type == TokenType.START_QUOTE) {
++j;
if(j < tag.TokenEndIndex) {
if(tokens[j].Type == TokenType.END_QUOTE) {
// blank value
attribute.Value = "";
} else {
// standard quoted value
attribute.Value = tokens[j].Value;
++j;
}
} else {
// half-quoted value (trailing off end of document?)
//this->insertLogEntry(ERROR, ATTRIBUTE_FORMAT, MISSING_END_QUOTE, null);
attribute.Value = tokens[j].Value;
++j;
}
} else {
// unquoted value
attribute.Value = tokens[j].Value;
}
} else {
//debugval = Parser::reassembleTokenString(array_slice(tokens, j - 1, 1));
//this->insertLogEntry(WARNING, ATTRIBUTE_FORMAT, MISSING_VALUE, "after '".debugval."'");
--j;
}
(node as ElementNode).Attributes.Add(attribute.Name, attribute.Value);
} else if(tokens[j].Type == TokenType.FORWARD_SLASH) {
// ignore for now
} else {
//int t = j;
if(tokens[j].Type == TokenType.START_QUOTE) {
while(tokens[j].Type != TokenType.END_QUOTE) {
++j;
}
} else {
while(tokens[j].Type != TokenType.CDATA) {
++j;
}
}
--j;
//debugval = Parser::reassembleTokenString(array_slice(tokens, t, j - t + 1));
//this->insertLogEntry(ERROR, ATTRIBUTE_FORMAT, INVALID_TOKEN_POSITION, "token sequence '".debugval."' removed");
}
}
}
} else {
if(tag.Type == NodeType.TEXT_NODE) {
node = new TextNode(tokens[tag.TokenStartIndex].Value);
} else if(tag.Type == NodeType.COMMENT_NODE) {
node = new CommentNode(tokens[tag.TokenStartIndex].Value);
} else if(tag.Type == NodeType.CDATA_SECTION_NODE) {
node = new CDATASectionNode(tokens[tag.TokenStartIndex].Value);
} else if(tag.Type == NodeType.PROCESSING_INSTRUCTION_NODE) {
node = new ProcessingInstructionNode(tokens[tag.TokenStartIndex].Value);
} else if(tag.Type == NodeType.DOCUMENT_TYPE_NODE) {
// format: PUBLIC "public" "system"
// or SYSTEM "system"
var name = tokens[tag.TokenStartIndex].Value;
var type = tokens[tag.TokenStartIndex + 1].Value;
string publicId = null;
string systemId = null;
int j = tag.TokenStartIndex + 2;
if(String.Compare(type, "PUBLIC", true) == 0) {
if(tokens[j].Type == TokenType.START_QUOTE) {
publicId = tokens[j + 1].Value;
j += 3;
}
}
if(tokens[j].Type == TokenType.START_QUOTE) {
systemId = tokens[j + 1].Value;
}
node = new DocumentTypeNode(name, publicId, systemId);
} else if(tag.Type == NodeType.DTD_ENTITY_NODE) {
//node = new DTDEntityNode();
} else if(tag.Type == NodeType.DTD_ELEMENT_NODE) {
//node = new DTDElementNode();
} else if(tag.Type == NodeType.DTD_ATTLIST_NODE) {
//node = new DTDAttListNode();
}
if(node != null) {
current.AppendChild(node);
}
}
}
}
return root;
}
public static unsafe DocumentNode ParseTreeFromWellFormedXml(string text, List<long> tags)
{
var root = new DocumentNode();
ElementNode current = root;
fixed (char* pText = text)
{
for (int i = 0; i < tags.Count; i++)
{
long tag = tags[i];
if (tag == 0)
continue;
int index = (int)tag;
int length = (int)(tag >> 32);
if (text[index] != '<')
{
var node = new TextNode(text.Substring(index, length));
current.AppendChild(node);
}
//else if (other == ushort.MaxValue)
//{
// var node = new CommentNode(text.SubstringTrim(index + 4, length - 7));
// current.AppendChild(node);
//}
else
{
bool isClosingTag = text[index + 1] == '/';
if (isClosingTag)
{
current = current.Parent;
}
else
{
bool isSelfClosingTag = (text[index + length - 2] == '/');
char* p = pText + index + 1;
char* pEnd = p + length - 2;
while (p != pEnd && !char.IsWhiteSpace(*p))
++p;
int tagNameLength = (int)(p - (pText + index + 1));
string tagName = text.Substring(index + 1, tagNameLength);
var node = new ElementNode(tagName, isSelfClosingTag);
int attributeStart = index + tagNameLength + 1;
ParseAttributesFromWellFormedXml(node, text, attributeStart, length - (attributeStart - index) - 1);
current.AppendChild(node);
if (!isSelfClosingTag)
{
current = node;
}
}
}
}
}
return root;
}