void KHTMLReader::completed()
{
kDebug(30503) << "KHTMLReader::completed";
qApp->exit_loop();
DOM::Document doc = _html->document(); // FIXME parse <HEAD> too
DOM::NodeList list = doc.getElementsByTagName("body");
DOM::Node docbody = list.item(0);
if (docbody.isNull()) {
kWarning(30503) << "no <BODY>, giving up";
_it_worked = false;
return;
}
parseNode(docbody);
list = doc.getElementsByTagName("head");
DOM::Node dochead = list.item(0);
if (!dochead.isNull())
parse_head(dochead);
else
kWarning(30503) << "WARNING: no html <HEAD> section";
_writer->cleanUpParagraph(state()->paragraph);
_it_worked = _writer->writeDoc();
}
bool Parser::parse() {
skip_newline(false);
while (1) {
ptr<Token> token = cur();
if (!token->type()) {
break;
}
eat();
if (*token == '\n') {
continue;
}
if (*token == TOKEN_INCLUDE) {
bool old_skip_newline = skip_newline();
skip_newline(false);
token = cur();
if (*token == TOKEN_CONST_STRING && look()->is_eol()) {
eat();
eat();
skip_newline(old_skip_newline);
ptr<Path> path = object<Path>(token->text());
if (_input.is_root()) {
_symbols.exportSymbol(object<IncludeTree>(path));
}
_input.load(path);
}
else {
log_expect(token->loc(), "string eol");
}
continue;
}
SegmentToken *seg = nullptr;
if (token->type() == TOKEN_SEGMENT) {
seg = static_cast<SegmentToken*>(token.get());
if (!seg->name()) {
bool old_skip_newline = skip_newline();
skip_newline(false);
if (!look()->is_eol()) {
log_expect(token->loc(), "eol");
}
skip_newline(old_skip_newline);
eat();
}
else {
seg = nullptr;
}
}
switch (_phase) {
case PARSE_PHASE_HEAD:
if (seg) {
_phase = PARSE_PHASE_BODY;
continue;
}
skip_newline(false);
parse_head(token);
break;
case PARSE_PHASE_BODY:
if (seg) {
_phase = PARSE_PHASE_TAIL;
continue;
}
skip_newline(true);
parse_body(token);
break;
case PARSE_PHASE_TAIL:
if (seg) {
log_error(token->loc(), "too more segment declear.");
}
skip_newline(false);
parse_tail(token);
break;
}
}
return true;
}
inline bool xml_element::parse_body(const char *&data) {
while(true) {
if(!*data) return false;
if(*data++ != '<') continue;
if(*data == '/') return false;
if(strbegin(data, "!DOCTYPE") == true) {
parse_doctype(data);
return true;
}
if(strbegin(data, "!--")) {
if(optional<unsigned> offset = strpos(data, "-->")) {
data += offset() + 3;
continue;
} else {
throw "...";
}
}
if(strbegin(data, "![CDATA[")) {
if(optional<unsigned> offset = strpos(data, "]]>")) {
data += offset() + 3;
continue;
} else {
throw "...";
}
}
optional<unsigned> offset = strpos(data, ">");
if(!offset) throw "...";
string tag = substr(data, 0, offset());
data += offset() + 1;
const char *content_begin = data;
bool self_terminating = false;
if(strend(tag, "?") == true) {
self_terminating = true;
tag.rtrim_once("?");
} else if(strend(tag, "/") == true) {
self_terminating = true;
tag.rtrim_once("/");
}
parse_head(tag);
if(self_terminating) return true;
while(*data) {
unsigned index = element.size();
xml_element node;
if(node.parse_body(data) == false) {
if(*data == '/') {
signed length = data - content_begin - 1;
if(length > 0) content = substr(content_begin, 0, length);
data++;
optional<unsigned> offset = strpos(data, ">");
if(!offset) throw "...";
tag = substr(data, 0, offset());
data += offset() + 1;
tag.replace("\t", " ");
tag.replace("\r", " ");
tag.replace("\n", " ");
while(strpos(tag, " ")) tag.replace(" ", " ");
tag.rtrim();
if(name != tag) throw "...";
return true;
}
} else {
element.append(node);
}
}
}
}