string ParseString(ref int pos)
{
text.Clear();
int p = pos;
bool esc = false;
for (;;)
{
if (p >= length - 1)
{
throw ParserEx;
}
int b = this[++p];
if (esc)
{
text.Add(b == '"' ? '"' :
b == '\\' ? '\\' :
b == 'b' ? '\b' :
b == 'f' ? '\f' :
b == 'n' ? '\n' :
b == 'r' ? '\r' :
b == 't' ? '\t' : (char)0);
esc = !esc;
}
else
{
if (b == '\\')
{
esc = !esc;
}
else if (b == '"')
{
pos = p;
return(text.ToString());
}
else
{
text.Accept(b);
}
}
}
}
string ParseName(ref int pos)
{
text.Clear();
int p = pos;
for (;;)
{
if (p >= length)
{
return(null);
}
int b = this[p++];
if (b == '=')
{
pos = p;
return(text.ToString());
}
else if (b == '+')
{
text.Accept(' ');
}
else if (b == '%') // percent-encoding %xy
{
if (p >= length)
{
throw ParserEx;
}
int x = this[p++];
if (p >= length)
{
throw ParserEx;
}
int y = this[p++];
text.Accept(Dv(x) << 4 | Dv(y));
}
else
{
text.Accept(b);
}
}
}
public Form Parse()
{
// locality for performance
byte[] bound_ = this.bound;
byte[] buffer_ = this.buffer;
int length_ = this.length;
// UTF-8 header builder
HeaderText hdr = new HeaderText(128);
Text text = new Text(128);
// keep local for speed
int boundlen = bound_.Length;
// shall init lately
Form frm = null;
int p = 0;
// skip first bound line whatever
for (;;)
{
if (buffer_[p++] == '\r' && buffer_[p++] == '\n')
{
break;
}
}
// parse parts
for (;;)
{
string name = null;
string filename = null;
string ctype = null;
// parse headers
for (;;)
{
hdr.Clear();
// parse a header line
for (;;)
{
if (p >= length_ - 2)
{
throw ParserEx;
}
byte b;
if ((b = buffer_[p++]) == '\r' && buffer_[p++] == '\n')
{
break;
}
hdr.Accept(b); // lineup the byte
}
if (hdr.Count == 0) // if empty line then quit header section
{
break;
}
if (name == null && hdr.Check("Content-Disposition"))
{
name = hdr.SeekParameter("name");
filename = hdr.SeekParameter("filename");
}
else if (ctype == null && hdr.Check("Content-Type"))
{
ctype = hdr.GetVvalue();
}
}
// get part's content
text.Clear();
bool plain = ctype == null || "text/plain".Equals(ctype);
int start = p; // mark down content start
int idx = 0; // index on bound
for (;;)
{
byte b = buffer_[p++];
if (b == bound_[idx])
{
idx++;
if (idx >= boundlen) // fully matched the bound accumulatively
{
if (frm == null)
{
frm = new Form(true)
{
Buffer = buffer_
}
}
;
if (plain)
{
frm.Add(name, text.ToString());
}
else
{
frm.Add(name, filename, start, p - start - boundlen);
}
// goto the ending CRLF/-- check
break;
}
}
else if (idx > 0) // if fail-match
{
if (plain) // re-add
{
for (int i = 0; i < idx; i++)
{
text.Accept(bound_[i]);
}
}
idx = 0; // reset
}
else
{
if (plain)
{
text.Accept(b);
}
}
}
// check if any more part
if (buffer_[p++] == '\r' && buffer_[p++] == '\n')
{
continue;
}
break;
} // parts
return(frm ?? Empty);
}
}
XElem ParseElem(ref int pos, int startchar)
{
int p = pos;
int b;
// parse element tag name
text.Clear();
text.Accept(startchar);
while (IsNameChar(b = this[++p]))
{
text.Accept(b); // to comprise start tag
}
string tag = text.ToString();
XElem elem = new XElem(tag);
// optionally parse attributes
while (IsWs(b))
{
while (IsWs(b = this[++p]))
{
} // skip ws
if (IsNameStartChar(b))
{
// attribute name
text.Clear();
text.Accept(b);
while ((b = this[++p]) != '=')
{
text.Accept(b);
}
string name = text.ToString();
// attribute value
if (this[++p] != '"')
{
throw ParserEx; // left quote
}
text.Clear();
while ((b = this[++p]) != '"') // till right quote
{
if (b == '&') // escape < > & "
{
int b1 = this[p + 1];
int b2 = this[p + 2];
int b3 = this[p + 3];
if (b1 == 'l' && b2 == 't' && b3 == ';')
{
b = '<';
p += 3;
}
else if (b1 == 'g' && b2 == 't' && b3 == ';')
{
b = '>';
p += 3;
}
else if (b1 == 'a' && b2 == 'm' && b3 == 'p' && this[p + 4] == ';')
{
b = '&';
p += 4;
}
else if (b1 == 'q' && b2 == 'u' && b3 == 'o' && this[p + 4] == 't' && this[p + 5] == ';')
{
b = '"';
p += 5;
}
}
text.Accept(b);
}
string value = text.ToString();
elem.AddAttr(name, value);
b = this[++p]; // step
}
} // end of attributes
if (b == '>') // a start tag just finished, expecting the ending-tag
{
for (;;) // child nodes iteration
{
while (IsWs(b = this[++p])) // skip ws
{
}
if (b == '<')
{
b = this[++p];
if (b == '/') // the ending tag
{
// consume
text.Clear();
while ((b = this[++p]) != '>')
{
text.Accept(b);
}
if (!text.Equals(tag))
{
throw ParserEx;
}
pos = p; // adjust current position
return(elem);
}
else if (b == '!') // CDATA section
{
if (this[p + 1] == '[' && this[p + 2] == 'C' && this[p + 3] == 'D' && this[p + 4] == 'A' && this[p + 5] == 'T' && this[p + 6] == 'A' && this[p + 7] == '[')
{
text.Clear();
p += 7;
while ((b = this[++p]) != ']' || this[p + 1] != ']' || this[p + 2] != '>')
{
text.Accept(b);
}
elem.Text = text.ToString();
p += 2; // skip ]>
}
}
else if (IsNameStartChar(b))
{
XElem child = ParseElem(ref p, b);
elem.Add(child);
}
}
else // text node
{
text.Clear();
while ((b = this[p]) != '<') // NOTE from the first char
{
if (b == '&') // escape < > & "
{
int b1 = this[p + 1];
int b2 = this[p + 2];
int b3 = this[p + 3];
if (b1 == 'l' && b2 == 't' && b3 == ';')
{
b = '<';
p += 3;
}
else if (b1 == 'g' && b2 == 't' && b3 == ';')
{
b = '>';
p += 3;
}
else if (b1 == 'a' && b2 == 'm' && b3 == 'p' && this[p + 4] == ';')
{
b = '&';
p += 4;
}
else if (b1 == 'q' && b2 == 'u' && b3 == 'o' && this[p + 4] == 't' && this[p + 5] == ';')
{
b = '"';
p += 5;
}
}
text.Accept(b);
++p;
}
if (text.Count > 0)
{
elem.Text = text.ToString();
}
// NOTE decrease in position to behave as other child nodes
--p;
}
} // child nodes iteration
}
if (b == '/' && this[++p] == '>') // empty-element
{
pos = p; // adjust current position
return(elem);
}
throw ParserEx;
}
