Commit 1480ce37 authored by Petar Maymounkov's avatar Petar Maymounkov Committed by Russ Cox

http: unified body transfer (read & write) logic in http.Request/Response.

Compliance issue addressed here: POST requests carrying form data are required
to use "identity" transfer encoding by common nginx and apache server configurations,
e.g. wordpress.com (and many others). So, Request needed to be able to send
non-chunked encodings.

Thus, Request is extended to support identity and chunked encodings, like
Response.  Since the Read() and Write() logic are shared by both (and are
quite long), it is exported in a separate file transfer.go.

R=rsc
CC=golang-dev
https://golang.org/cl/217048
parent 8d9c2b2e
......@@ -16,6 +16,7 @@ GOFILES=\
response.go\
server.go\
status.go\
transfer.go\
url.go\
include ../../Make.pkg
......@@ -60,6 +60,7 @@ var reqTests = []reqTest{
"Content-Length": "7",
},
Close: false,
ContentLength: 7,
Host: "www.techcrunch.com",
Referer: "",
UserAgent: "Fake",
......
......@@ -49,7 +49,14 @@ type badStringError struct {
func (e *badStringError) String() string { return fmt.Sprintf("%s %q", e.what, e.str) }
var reqExcludeHeader = map[string]int{"Host": 0, "User-Agent": 0, "Referer": 0}
var reqExcludeHeader = map[string]int{
"Host": 0,
"User-Agent": 0,
"Referer": 0,
"Content-Length": 0,
"Transfer-Encoding": 0,
"Trailer": 0,
}
// A Request represents a parsed HTTP request header.
type Request struct {
......@@ -84,6 +91,15 @@ type Request struct {
// The message body.
Body io.ReadCloser
// ContentLength records the length of the associated content.
// The value -1 indicates that the length is unknown.
// Values >= 0 indicate that the given number of bytes may be read from Body.
ContentLength int64
// TransferEncoding lists the transfer encodings from outermost to innermost.
// An empty list denotes the "identity" encoding.
TransferEncoding []string
// Whether to close the connection after replying to this request.
Close bool
......@@ -109,6 +125,11 @@ type Request struct {
// The parsed form. Only available after ParseForm is called.
Form map[string][]string
// Trailer maps trailer keys to values. Like for Header, if the
// response has multiple trailer lines with the same key, they will be
// concatenated, delimited by commas.
Trailer map[string]string
}
// ProtoAtLeast returns whether the HTTP protocol used
......@@ -152,16 +173,22 @@ func (req *Request) Write(w io.Writer) os.Error {
}
fmt.Fprintf(w, "%s %s HTTP/1.1\r\n", valueOrDefault(req.Method, "GET"), uri)
// Header lines
fmt.Fprintf(w, "Host: %s\r\n", host)
fmt.Fprintf(w, "User-Agent: %s\r\n", valueOrDefault(req.UserAgent, defaultUserAgent))
if req.Referer != "" {
fmt.Fprintf(w, "Referer: %s\r\n", req.Referer)
}
if req.Body != nil {
// Force chunked encoding
req.Header["Transfer-Encoding"] = "chunked"
// Process Body,ContentLength,Close,Trailer
tw, err := newTransferWriter(req)
if err != nil {
return err
}
err = tw.WriteHeader(w)
if err != nil {
return err
}
// TODO: split long values? (If so, should share code with Conn.Write)
......@@ -171,29 +198,17 @@ func (req *Request) Write(w io.Writer) os.Error {
// from Request, and introduce Request methods along the lines of
// Response.{GetHeader,AddHeader} and string constants for "Host",
// "User-Agent" and "Referer".
err := writeSortedKeyValue(w, req.Header, reqExcludeHeader)
err = writeSortedKeyValue(w, req.Header, reqExcludeHeader)
if err != nil {
return err
}
io.WriteString(w, "\r\n")
if req.Body != nil {
cw := NewChunkedWriter(w)
if _, err := io.Copy(cw, req.Body); err != nil {
return err
}
if err := cw.Close(); err != nil {
return err
}
// TODO(petar): Write trailer here and append \r\n. For now, we
// simply send the final \r\n:
if _, err := fmt.Fprint(w, "\r\n"); err != nil {
return err
}
if err := req.Body.Close(); err != nil {
return err
}
// Write body and trailer
err = tw.WriteBody(w)
if err != nil {
return err
}
return nil
......@@ -507,26 +522,7 @@ func ReadRequest(b *bufio.Reader) (req *Request, err os.Error) {
req.Header["Host"] = "", false
}
// RFC2616: Should treat
// Pragma: no-cache
// like
// Cache-Control: no-cache
if v, present := req.Header["Pragma"]; present && v == "no-cache" {
if _, presentcc := req.Header["Cache-Control"]; !presentcc {
req.Header["Cache-Control"] = "no-cache"
}
}
// Determine whether to hang up after sending the reply.
if req.ProtoMajor < 1 || (req.ProtoMajor == 1 && req.ProtoMinor < 1) {
req.Close = true
} else if v, present := req.Header["Connection"]; present {
// TODO: Should split on commas, toss surrounding white space,
// and check each field.
if v == "close" {
req.Close = true
}
}
fixPragmaCacheControl(req.Header)
// Pull out useful fields as a convenience to clients.
if v, present := req.Header["Referer"]; present {
......@@ -564,17 +560,9 @@ func ReadRequest(b *bufio.Reader) (req *Request, err os.Error) {
// Via
// Warning
// A message body exists when either Content-Length or Transfer-Encoding
// headers are present. Transfer-Encoding trumps Content-Length.
if v, present := req.Header["Transfer-Encoding"]; present && v == "chunked" {
req.Body = &body{Reader: newChunkedReader(b), hdr: req, r: b, closing: req.Close}
} else if v, present := req.Header["Content-Length"]; present {
length, err := strconv.Btoi64(v, 10)
if err != nil {
return nil, &badStringError{"invalid Content-Length", v}
}
// TODO: limit the Content-Length. This is an easy DoS vector.
req.Body = &body{Reader: io.LimitReader(b, length), closing: req.Close}
err = readTransfer(req, b)
if err != nil {
return nil, err
}
return req, nil
......
......@@ -70,9 +70,10 @@ var reqWriteTests = []reqWriteTest{
Path: "/search",
},
ProtoMajor: 1,
ProtoMinor: 0,
ProtoMinor: 1,
Header: map[string]string{},
Body: nopCloser{bytes.NewBufferString("abcdef")},
TransferEncoding: []string{"chunked"},
},
"GET /search HTTP/1.1\r\n" +
......@@ -83,9 +84,6 @@ var reqWriteTests = []reqWriteTest{
},
}
// FIXME(petar): The write order of keys in Request.Header depends on the
// map[string]string iterator. Since this isn't defined in Go's semantics, we
// should eventually fix Request.Write()
func TestRequestWrite(t *testing.T) {
for i := range reqWriteTests {
tt := &reqWriteTests[i]
......
......@@ -16,7 +16,11 @@ import (
"strings"
)
var respExcludeHeader = map[string]int{}
var respExcludeHeader = map[string]int{
"Content-Length": 0,
"Transfer-Encoding": 0,
"Trailer": 0,
}
// Response represents the response from an HTTP request.
//
......@@ -116,78 +120,14 @@ func ReadResponse(r *bufio.Reader, requestMethod string) (resp *Response, err os
fixPragmaCacheControl(resp.Header)
// Transfer encoding, content length
resp.TransferEncoding, err = fixTransferEncoding(resp.Header)
if err != nil {
return nil, err
}
resp.ContentLength, err = fixLength(resp.StatusCode, resp.RequestMethod,
resp.Header, resp.TransferEncoding)
if err != nil {
return nil, err
}
// Closing
resp.Close = shouldClose(resp.ProtoMajor, resp.ProtoMinor, resp.Header)
// Trailer
resp.Trailer, err = fixTrailer(resp.Header, resp.TransferEncoding)
err = readTransfer(resp, r)
if err != nil {
return nil, err
}
// Prepare body reader. ContentLength < 0 means chunked encoding
// or close connection when finished, since multipart is not supported yet
switch {
case chunked(resp.TransferEncoding):
resp.Body = &body{Reader: newChunkedReader(r), hdr: resp, r: r, closing: resp.Close}
case resp.ContentLength >= 0:
resp.Body = &body{Reader: io.LimitReader(r, resp.ContentLength), closing: resp.Close}
default:
resp.Body = &body{Reader: r, closing: resp.Close}
}
return resp, nil
}
// body turns a Reader into a ReadCloser.
// Close ensures that the body has been fully read
// and then reads the trailer if necessary.
type body struct {
io.Reader
hdr interface{} // non-nil (Response or Request) value means read trailer
r *bufio.Reader // underlying wire-format reader for the trailer
closing bool // is the connection to be closed after reading body?
}
func (b *body) Close() os.Error {
if b.hdr == nil && b.closing {
// no trailer and closing the connection next.
// no point in reading to EOF.
return nil
}
trashBuf := make([]byte, 1024) // local for thread safety
for {
_, err := b.Read(trashBuf)
if err == nil {
continue
}
if err == os.EOF {
break
}
return err
}
if b.hdr == nil { // not reading trailer
return nil
}
// TODO(petar): Put trailer reader code here
return nil
}
// RFC2616: Should treat
// Pragma: no-cache
// like
......@@ -200,147 +140,6 @@ func fixPragmaCacheControl(header map[string]string) {
}
}
// Parse the trailer header
func fixTrailer(header map[string]string, te []string) (map[string]string, os.Error) {
raw, present := header["Trailer"]
if !present {
return nil, nil
}
header["Trailer"] = "", false
trailer := make(map[string]string)
keys := strings.Split(raw, ",", 0)
for _, key := range keys {
key = CanonicalHeaderKey(strings.TrimSpace(key))
switch key {
case "Transfer-Encoding", "Trailer", "Content-Length":
return nil, &badStringError{"bad trailer key", key}
}
trailer[key] = ""
}
if len(trailer) == 0 {
return nil, nil
}
if !chunked(te) {
// Trailer and no chunking
return nil, ErrUnexpectedTrailer
}
return trailer, nil
}
// Sanitize transfer encoding
func fixTransferEncoding(header map[string]string) ([]string, os.Error) {
raw, present := header["Transfer-Encoding"]
if !present {
return nil, nil
}
header["Transfer-Encoding"] = "", false
encodings := strings.Split(raw, ",", 0)
te := make([]string, 0, len(encodings))
// TODO: Even though we only support "identity" and "chunked"
// encodings, the loop below is designed with foresight. One
// invariant that must be maintained is that, if present,
// chunked encoding must always come first.
for _, encoding := range encodings {
encoding = strings.ToLower(strings.TrimSpace(encoding))
// "identity" encoding is not recored
if encoding == "identity" {
break
}
if encoding != "chunked" {
return nil, &badStringError{"unsupported transfer encoding", encoding}
}
te = te[0 : len(te)+1]
te[len(te)-1] = encoding
}
if len(te) > 1 {
return nil, &badStringError{"too many transfer encodings", strings.Join(te, ",")}
}
if len(te) > 0 {
// Chunked encoding trumps Content-Length. See RFC 2616
// Section 4.4. Currently len(te) > 0 implies chunked
// encoding.
header["Content-Length"] = "", false
return te, nil
}
return nil, nil
}
func noBodyExpected(requestMethod string) bool {
return requestMethod == "HEAD"
}
// Determine the expected body length, using RFC 2616 Section 4.4. This
// function is not a method, because ultimately it should be shared by
// ReadResponse and ReadRequest.
func fixLength(status int, requestMethod string, header map[string]string, te []string) (int64, os.Error) {
// Logic based on response type or status
if noBodyExpected(requestMethod) {
return 0, nil
}
if status/100 == 1 {
return 0, nil
}
switch status {
case 204, 304:
return 0, nil
}
// Logic based on Transfer-Encoding
if chunked(te) {
return -1, nil
}
// Logic based on Content-Length
if cl, present := header["Content-Length"]; present {
cl = strings.TrimSpace(cl)
if cl != "" {
n, err := strconv.Atoi64(cl)
if err != nil || n < 0 {
return -1, &badStringError{"bad Content-Length", cl}
}
return n, nil
} else {
header["Content-Length"] = "", false
}
}
// Logic based on media type. The purpose of the following code is just
// to detect whether the unsupported "multipart/byteranges" is being
// used. A proper Content-Type parser is needed in the future.
if ct, present := header["Content-Type"]; present {
ct = strings.ToLower(ct)
if strings.Index(ct, "multipart/byteranges") >= 0 {
return -1, ErrNotSupported
}
}
// Logic based on close
return -1, nil
}
// Determine whether to hang up after sending a request and body, or
// receiving a response and body
func shouldClose(major, minor int, header map[string]string) bool {
if major < 1 || (major == 1 && minor < 1) {
return true
} else if v, present := header["Connection"]; present {
// TODO: Should split on commas, toss surrounding white space,
// and check each field.
if v == "close" {
return true
}
}
return false
}
// Checks whether chunked is part of the encodings stack
func chunked(te []string) bool { return len(te) > 0 && te[0] == "chunked" }
// AddHeader adds a value under the given key. Keys are not case sensitive.
func (r *Response) AddHeader(key, value string) {
key = CanonicalHeaderKey(key)
......@@ -397,70 +196,18 @@ func (resp *Response) Write(w io.Writer) os.Error {
io.WriteString(w, strconv.Itoa(resp.ProtoMinor)+" ")
io.WriteString(w, strconv.Itoa(resp.StatusCode)+" "+text+"\r\n")
// Sanitize the field triple (Body, ContentLength, TransferEncoding)
contentLength := resp.ContentLength
if noBodyExpected(resp.RequestMethod) {
resp.Body = nil
resp.TransferEncoding = nil
// resp.ContentLength is expected to hold Content-Length
if contentLength < 0 {
return ErrMissingContentLength
}
} else {
if !resp.ProtoAtLeast(1, 1) || resp.Body == nil {
resp.TransferEncoding = nil
}
if chunked(resp.TransferEncoding) {
contentLength = -1
} else if resp.Body == nil { // no chunking, no body
contentLength = 0
}
}
// Write Content-Length and/or Transfer-Encoding whose values are a
// function of the sanitized field triple (Body, ContentLength,
// TransferEncoding)
if chunked(resp.TransferEncoding) {
io.WriteString(w, "Transfer-Encoding: chunked\r\n")
} else {
if contentLength > 0 || resp.RequestMethod == "HEAD" {
io.WriteString(w, "Content-Length: ")
io.WriteString(w, strconv.Itoa64(contentLength)+"\r\n")
}
}
if resp.Header != nil {
resp.Header["Content-Length"] = "", false
resp.Header["Transfer-Encoding"] = "", false
}
// Sanitize Trailer
if !chunked(resp.TransferEncoding) {
resp.Trailer = nil
} else if resp.Trailer != nil {
// TODO: At some point, there should be a generic mechanism for
// writing long headers, using HTTP line splitting
io.WriteString(w, "Trailer: ")
needComma := false
for k, _ := range resp.Trailer {
k = CanonicalHeaderKey(k)
switch k {
case "Transfer-Encoding", "Trailer", "Content-Length":
return &badStringError{"invalid Trailer key", k}
}
if needComma {
io.WriteString(w, ",")
}
io.WriteString(w, k)
needComma = true
}
io.WriteString(w, "\r\n")
// Process Body,ContentLength,Close,Trailer
tw, err := newTransferWriter(resp)
if err != nil {
return err
}
if resp.Header != nil {
resp.Header["Trailer"] = "", false
err = tw.WriteHeader(w)
if err != nil {
return err
}
// Rest of header
err := writeSortedKeyValue(w, resp.Header, respExcludeHeader)
err = writeSortedKeyValue(w, resp.Header, respExcludeHeader)
if err != nil {
return err
}
......@@ -468,30 +215,10 @@ func (resp *Response) Write(w io.Writer) os.Error {
// End-of-header
io.WriteString(w, "\r\n")
// Write body
if resp.Body != nil {
var err os.Error
if chunked(resp.TransferEncoding) {
cw := NewChunkedWriter(w)
_, err = io.Copy(cw, resp.Body)
if err == nil {
err = cw.Close()
}
} else {
_, err = io.Copy(w, io.LimitReader(resp.Body, contentLength))
}
if err != nil {
return err
}
if err = resp.Body.Close(); err != nil {
return err
}
}
// TODO(petar): Place trailer writer code here.
if chunked(resp.TransferEncoding) {
// Last chunk, empty trailer
io.WriteString(w, "\r\n")
// Write body and trailer
err = tw.WriteBody(w)
if err != nil {
return err
}
// Success
......
This diff is collapsed.
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment