tree
xml_html_parser::parse_doctype () {
s += 9;
tree dt= tuple ("doctype");
skip_space ();
dt << parse_name ();
skip_space ();
if (test (s, "SYSTEM")) dt << parse_system ();
else if (test (s, "PUBLIC")) dt << parse_public ();
skip_space ();
if (test (s, "[")) {
s += 1;
while (s) {
skip_space ();
if (test (s, "]")) { s += 1; break; }
else if (test (s, "<!ELEMENT")) dt << parse_element ();
else if (test (s, "<!ATTLIST")) dt << parse_cdata ();
else if (test (s, "<!ENTITY")) parse_entity_decl ();
else if (test (s, "<!NOTATION")) a << parse_notation ();
else if (test (s, "<?")) dt << parse_pi ();
else if (test (s, "<!--")) dt << parse_comment ();
else if (s[0] == '&' || s[0] == '%') (void) parse_entity ();
else s += 1;
}
}
skip_space ();
if (test (s, ">")) s += 1;
return dt;
}
void
xml_html_parser::parse () {
string r;
while (s) {
if (s[0] == '<') {
if (N(r) != 0) { a << tree (r); }
if (test (s, "</")) a << parse_closing ();
else if (test (s, "<?")) a << parse_pi ();
else if (test (s, "<!--")) a << parse_comment ();
else if (test (s, "<![CDATA[")) a << parse_cdata ();
else if (test (s, "<!DOCTYPE")) a << parse_doctype ();
else if (test (s, "<!")) a << parse_misc ();
else a << parse_opening ();
r= "";
}
else if (s[0] == '&') r << parse_entity ();
else r << s->read (1);
}
if (N(r) != 0) a << tree (r);
}
static int parse_node(xmlNodePtr node, simple_binary_t **sibxml)
{
int status = 0;
/* Call for the parser function of the node type. */
switch (node->type) {
case XML_ELEMENT_NODE:
status = parse_element(node, sibxml);
break;
case XML_TEXT_NODE:
status = parse_text(node, sibxml);
break;
case XML_CDATA_SECTION_NODE:
status = parse_cdata(node, sibxml);
break;
case XML_COMMENT_NODE:
case XML_PI_NODE:
/* Comments and PIs are ignored. */
break;
/*
* XML has also many other node types, these are not needed with
* SI. Therefore they are assumed to be an error.
*/
default:
error(0, "SI compiler: Unknown XML node in the SI source.");
return -1;
break;
}
/*
* If node is an element with content, it will need an end tag after it's
* children. The status for it is returned by parse_element.
*/
switch (status) {
case 0:
if (node->children != NULL)
if (parse_node(node->children, sibxml) == -1)
return -1;
break;
case 1:
if (node->children != NULL)
if (parse_node(node->children, sibxml) == -1)
return -1;
parse_end(sibxml);
break;
case -1: /* Something went wrong in the parsing. */
return -1;
default:
warning(0,"SI compiler: undefined return value in a parse function.");
return -1;
break;
}
if (node->next != NULL)
if (parse_node(node->next, sibxml) == -1)
return -1;
return 0;
}
bool xml_istream::process()
{
stack<string> tag_stack;
stack< map<string, string> > attribute_stack;
bool ok = true;
while (ok != false && good() != false)
{
string tag, cdata, doctype, s;
map<string, string> attributes;
xml_t t = get_xml_node(s, tag, attributes);
switch (t)
{
case OPEN_TAG:
tag_stack.push(tag);
attribute_stack.push(attributes);
break;
case EMPTY_TAG:
{
callback pFunc = _callback_map[tag];
if (pFunc != NULL)
{
pFunc(s, attributes);
}
}
break;
case CLOSE_TAG:
if (tag_stack.size() > 0 && tag_stack.top() == tag) /* Ensure opening tag matches the closing tag. */
{
tag_stack.pop();
attribute_stack.pop();
}
else
{
ok = false;
}
break;
case TEXT:
if (s.empty() == false && tag_stack.size() > 0)
{
callback pFunc = _callback_map[tag_stack.top()];
if (pFunc != NULL)
{
if (_entity_translation != false)
{
string ts;
entity_translation(s, ts);
s = ts;
}
if (_collapse_whitespace != false)
{
string ts;
collapse_whitespace(s, ts);
s = ts;
}
pFunc(s, attribute_stack.top());
}
}
else
{
ok = false;
}
break;
case CDATA:
ok = parse_cdata(s, cdata);
if (cdata.empty() == false && tag_stack.size() > 0)
{
callback pFunc = _callback_map[tag_stack.top()];
if (pFunc != NULL)
{
pFunc(cdata, attribute_stack.top());
}
}
else
{
ok = false;
}
break;
case UNKNOWN:
ok = false;
break;
}
}
if (tag_stack.size() != 0 || attribute_stack.size() != 0)
{
ok = false;
}
return ok;
}
