internal DataParseStatus ParseHeaders(byte[] buffer, int size, ref int unparsed, ref int totalResponseHeadersLength, int maximumResponseHeadersLength)
{
int headerNameStartOffset = -1;
int headerNameEndOffset = -1;
int headerValueStartOffset = -1;
int headerValueEndOffset = -1;
int numberOfLf = -1;
int index = unparsed;
bool spaceAfterLf;
string headerMultiLineValue;
string headerName;
string headerValue;
// we need this because this method is entered multiple times.
int localTotalResponseHeadersLength = totalResponseHeadersLength;
DataParseStatus parseStatus = DataParseStatus.Invalid;
GlobalLog.Enter("WebHeaderCollection::ParseHeaders(): size:" + size.ToString() + ", unparsed:" + unparsed.ToString() + " buffer:[" + Encoding.ASCII.GetString(buffer, unparsed, Math.Min(256, size - unparsed)) + "]");
//GlobalLog.Print("WebHeaderCollection::ParseHeaders(): advancing. index=" + index.ToString() + " current character='" + ((char)buffer[index]).ToString() + "':" + ((int)buffer[index]).ToString() + " next character='" + ((char)buffer[index+1]).ToString() + "':" + ((int)buffer[index+1]).ToString());
//GlobalLog.Print("WebHeaderCollection::ParseHeaders(): headerName:[" + headerName + "], headerValue:[" + headerValue + "], headerMultiLineValue:[" + headerMultiLineValue + "] numberOfLf=" + numberOfLf.ToString() + " index=" + index.ToString());
//GlobalLog.Print("WebHeaderCollection::ParseHeaders(): headerNameStartOffset=" + headerNameStartOffset + " headerNameEndOffset=" + headerNameEndOffset + " headerValueStartOffset=" + headerValueStartOffset + " headerValueEndOffset=" + headerValueEndOffset);
//
// according to RFC216 a header can have the following syntax:
//
// message-header = field-name ":" [ field-value ]
// field-name = token
// field-value = *( field-content | LWS )
// field-content = <the OCTETs making up the field-value and consisting of either *TEXT or combinations of token, separators, and quoted-string>
// TEXT = <any OCTET except CTLs, but including LWS>
// CTL = <any US-ASCII control character (octets 0 - 31) and DEL (127)>
// SP = <US-ASCII SP, space (32)>
// HT = <US-ASCII HT, horizontal-tab (9)>
// CR = <US-ASCII CR, carriage return (13)>
// LF = <US-ASCII LF, linefeed (10)>
// LWS = [CR LF] 1*( SP | HT )
// CHAR = <any US-ASCII character (octets 0 - 127)>
// token = 1*<any CHAR except CTLs or separators>
// separators = "(" | ")" | "<" | ">" | "@" | "," | ";" | ":" | "\" | <"> | "/" | "[" | "]" | "?" | "=" | "{" | "}" | SP | HT
// quoted-string = ( <"> *(qdtext | quoted-pair ) <"> )
// qdtext = <any TEXT except <">>
// quoted-pair = "\" CHAR
//
//
// At each iteration of the following loop we expect to parse a single HTTP header entirely.
//
for (;;)
{
//
// trim leading whitespaces (LWS) just for extra robustness, in fact if there are leading white spaces then:
// 1) it could be that after the status line we might have spaces. handle this.
// 2) this should have been detected to be a multiline header so there'll be no spaces and we'll spend some time here.
//
headerName = string.Empty;
headerValue = string.Empty;
spaceAfterLf = false;
headerMultiLineValue = null;
if (Count == 0)
{
//
// so, restrict this extra trimming only on the first header line
//
while (index < size && (buffer[index] == ' ' || buffer[index] == '\t'))
{
index++;
if (maximumResponseHeadersLength >= 0 && ++localTotalResponseHeadersLength >= maximumResponseHeadersLength)
{
parseStatus = DataParseStatus.DataTooBig;
goto quit;
}
}
if (index == size)
{
//
// we reached the end of the buffer. ask for more data.
//
parseStatus = DataParseStatus.NeedMoreData;
goto quit;
}
}
//
// what we have here is the beginning of a new header
//
headerNameStartOffset = index;
while (index < size && (buffer[index] != ':' && buffer[index] != '\n'))
{
if (buffer[index] > ' ')
{
//
// if thre's an illegal character we should return DataParseStatus.Invalid
// instead we choose to be flexible, try to trim it, but include it in the string
//
headerNameEndOffset = index;
}
index++;
if (maximumResponseHeadersLength >= 0 && ++localTotalResponseHeadersLength >= maximumResponseHeadersLength)
{
parseStatus = DataParseStatus.DataTooBig;
goto quit;
}
}
if (index == size)
{
//
// we reached the end of the buffer. ask for more data.
//
parseStatus = DataParseStatus.NeedMoreData;
goto quit;
}
startOfValue:
//
// skip all [' ','\t',':','\r','\n'] characters until HeaderValue starts
// if we didn't find any headers yet, we set numberOfLf to 1
// so that we take the '\n' from the status line into account
//
numberOfLf = (Count == 0 && headerNameEndOffset < 0) ? 1 : 0;
while (index < size && numberOfLf < 2 && (buffer[index] <= ' ' || buffer[index] == ':'))
{
if (buffer[index] == '\n')
{
numberOfLf++;
spaceAfterLf = index + 1 < size && (buffer[index + 1] == ' ' || buffer[index + 1] == '\t');
}
index++;
if (maximumResponseHeadersLength >= 0 && ++localTotalResponseHeadersLength >= maximumResponseHeadersLength)
{
parseStatus = DataParseStatus.DataTooBig;
goto quit;
}
}
if (numberOfLf == 2 || (numberOfLf == 1 && !spaceAfterLf))
{
//
// if we've counted two '\n' we got at the end of the headers even if we're past the end of the buffer
// if we've counted one '\n' and the first character after that was a ' ' or a '\t'
// no matter if we found a ':' or not, treat this as an empty header name.
//
goto addHeader;
}
if (index == size)
{
//
// we reached the end of the buffer. ask for more data.
//
parseStatus = DataParseStatus.NeedMoreData;
goto quit;
}
headerValueStartOffset = index;
while (index < size && buffer[index] != '\n')
{
if (buffer[index] > ' ')
{
headerValueEndOffset = index;
}
index++;
if (maximumResponseHeadersLength >= 0 && ++localTotalResponseHeadersLength >= maximumResponseHeadersLength)
{
parseStatus = DataParseStatus.DataTooBig;
goto quit;
}
}
if (index == size)
{
//
// we reached the end of the buffer. ask for more data.
//
parseStatus = DataParseStatus.NeedMoreData;
goto quit;
}
//
// at this point we found either a '\n' or the end of the headers
// hence we are at the end of the Header Line. 4 options:
// 1) need more data
// 2) if we find two '\n' => end of headers
// 3) if we find one '\n' and a ' ' or a '\t' => multiline header
// 4) if we find one '\n' and a valid char => next header
//
numberOfLf = 0;
while (index < size && numberOfLf < 2 && (buffer[index] == '\r' || buffer[index] == '\n'))
{
if (buffer[index] == '\n')
{
numberOfLf++;
}
index++;
if (maximumResponseHeadersLength >= 0 && ++localTotalResponseHeadersLength >= maximumResponseHeadersLength)
{
parseStatus = DataParseStatus.DataTooBig;
goto quit;
}
}
if (index == size && numberOfLf < 2)
{
//
// we reached the end of the buffer but not of the headers. ask for more data.
//
parseStatus = DataParseStatus.NeedMoreData;
goto quit;
}
addHeader:
if (headerValueStartOffset >= 0 && headerValueStartOffset > headerNameEndOffset && headerValueEndOffset >= headerValueStartOffset)
{
//
// Encoding fastest way to build the UNICODE string off the byte[]
//
headerValue = HeaderEncoding.GetString(buffer, headerValueStartOffset, headerValueEndOffset - headerValueStartOffset + 1);
}
//
// if we got here from the beginning of the for loop, headerMultiLineValue will be null
// otherwise it will contain the headerValue constructed for the multiline header
// add this line as well to it, separated by a single space
//
headerMultiLineValue = (headerMultiLineValue == null ? headerValue : headerMultiLineValue + " " + headerValue);
if (index < size && numberOfLf == 1 && (buffer[index] == ' ' || buffer[index] == '\t'))
{
//
// since we found only one Lf and the next header line begins with a Lws,
// this is the beginning of a multiline header.
// parse the next line into headerValue later it will be added to headerMultiLineValue
//
index++;
if (maximumResponseHeadersLength >= 0 && ++localTotalResponseHeadersLength >= maximumResponseHeadersLength)
{
parseStatus = DataParseStatus.DataTooBig;
goto quit;
}
goto startOfValue;
}
if (headerNameStartOffset >= 0 && headerNameEndOffset >= headerNameStartOffset)
{
//
// Encoding is the fastest way to build the UNICODE string off the byte[]
//
headerName = HeaderEncoding.GetString(buffer, headerNameStartOffset, headerNameEndOffset - headerNameStartOffset + 1);
}
//
// now it's finally safe to add the header if we have a name for it
//
if (headerName.Length > 0)
{
//
// the base clasee will check for pre-existing headerValue and append
// it using commas as indicated in the RFC
//
GlobalLog.Print("WebHeaderCollection::ParseHeaders() calling base.Add() key:[" + headerName + "], value:[" + headerMultiLineValue + "]");
base.Add(headerName, headerMultiLineValue);
}
//
// and update unparsed
//
totalResponseHeadersLength = localTotalResponseHeadersLength;
unparsed = index;
if (numberOfLf == 2)
{
parseStatus = DataParseStatus.Done;
goto quit;
}
} // for (;;)
quit:
GlobalLog.Leave("WebHeaderCollection::ParseHeaders() returning parseStatus:" + parseStatus.ToString());
return(parseStatus);
}
