curl/lib/http_chunks.c
Stefan Eissing d7b6ce64ce
lib: replace readwrite with write_resp
This clarifies the handling of server responses by folding the code for
the complicated protocols into their protocol handlers. This concerns
mainly HTTP and its bastard sibling RTSP.

The terms "read" and "write" are often used without clear context if
they refer to the connect or the client/application side of a
transfer. This PR uses "read/write" for operations on the client side
and "send/receive" for the connection, e.g. server side. If this is
considered useful, we can revisit renaming of further methods in another
PR.

Curl's protocol handler `readwrite()` method been changed:

```diff
-  CURLcode (*readwrite)(struct Curl_easy *data, struct connectdata *conn,
-                        const char *buf, size_t blen,
-                        size_t *pconsumed, bool *readmore);
+  CURLcode (*write_resp)(struct Curl_easy *data, const char *buf, size_t blen,
+                         bool is_eos, bool *done);
```

The name was changed to clarify that this writes reponse data to the
client side. The parameter changes are:

* `conn` removed as it always operates on `data->conn`
* `pconsumed` removed as the method needs to handle all data on success
* `readmore` removed as no longer necessary
* `is_eos` as indicator that this is the last call for the transfer
  response (end-of-stream).
* `done` TRUE on return iff the transfer response is to be treated as
  finished

This change affects many files only because of updated comments in
handlers that provide no implementation. The real change is that the
HTTP protocol handlers now provide an implementation.

The HTTP protocol handlers `write_resp()` implementation will get passed
**all** raw data of a server response for the transfer. The HTTP/1.x
formatted status and headers, as well as the undecoded response
body. `Curl_http_write_resp_hds()` is used internally to parse the
response headers and pass them on. This method is public as the RTSP
protocol handler also uses it.

HTTP/1.1 "chunked" transport encoding is now part of the general
*content encoding* writer stack, just like other encodings. A new flag
`CLIENTWRITE_EOS` was added for the last client write. This allows
writers to verify that they are in a valid end state. The chunked
decoder will check if it indeed has seen the last chunk.

The general response handling in `transfer.c:466` happens in function
`readwrite_data()`. This mainly operates now like:

```
static CURLcode readwrite_data(data, ...)
{
  do {
    Curl_xfer_recv_resp(data, buf)
    ...
    Curl_xfer_write_resp(data, buf)
    ...
  } while(interested);
  ...
}
```

All the response data handling is implemented in
`Curl_xfer_write_resp()`. It calls the protocol handler's `write_resp()`
implementation if available, or does the default behaviour.

All raw response data needs to pass through this function. Which also
means that anyone in possession of such data may call
`Curl_xfer_write_resp()`.

Closes #12480
2024-01-13 17:23:42 +01:00

456 lines
14 KiB
C

/***************************************************************************
* _ _ ____ _
* Project ___| | | | _ \| |
* / __| | | | |_) | |
* | (__| |_| | _ <| |___
* \___|\___/|_| \_\_____|
*
* Copyright (C) Daniel Stenberg, <daniel@haxx.se>, et al.
*
* This software is licensed as described in the file COPYING, which
* you should have received as part of this distribution. The terms
* are also available at https://curl.se/docs/copyright.html.
*
* You may opt to use, copy, modify, merge, publish, distribute and/or sell
* copies of the Software, and permit persons to whom the Software is
* furnished to do so, under the terms of the COPYING file.
*
* This software is distributed on an "AS IS" basis, WITHOUT WARRANTY OF ANY
* KIND, either express or implied.
*
* SPDX-License-Identifier: curl
*
***************************************************************************/
#include "curl_setup.h"
#ifndef CURL_DISABLE_HTTP
#include "urldata.h" /* it includes http_chunks.h */
#include "sendf.h" /* for the client write stuff */
#include "dynbuf.h"
#include "content_encoding.h"
#include "http.h"
#include "strtoofft.h"
#include "warnless.h"
/* The last #include files should be: */
#include "curl_memory.h"
#include "memdebug.h"
/*
* Chunk format (simplified):
*
* <HEX SIZE>[ chunk extension ] CRLF
* <DATA> CRLF
*
* Highlights from RFC2616 section 3.6 say:
The chunked encoding modifies the body of a message in order to
transfer it as a series of chunks, each with its own size indicator,
followed by an OPTIONAL trailer containing entity-header fields. This
allows dynamically produced content to be transferred along with the
information necessary for the recipient to verify that it has
received the full message.
Chunked-Body = *chunk
last-chunk
trailer
CRLF
chunk = chunk-size [ chunk-extension ] CRLF
chunk-data CRLF
chunk-size = 1*HEX
last-chunk = 1*("0") [ chunk-extension ] CRLF
chunk-extension= *( ";" chunk-ext-name [ "=" chunk-ext-val ] )
chunk-ext-name = token
chunk-ext-val = token | quoted-string
chunk-data = chunk-size(OCTET)
trailer = *(entity-header CRLF)
The chunk-size field is a string of hex digits indicating the size of
the chunk. The chunked encoding is ended by any chunk whose size is
zero, followed by the trailer, which is terminated by an empty line.
*/
void Curl_httpchunk_init(struct Curl_easy *data, struct Curl_chunker *ch,
bool ignore_body)
{
(void)data;
ch->hexindex = 0; /* start at 0 */
ch->state = CHUNK_HEX; /* we get hex first! */
ch->last_code = CHUNKE_OK;
Curl_dyn_init(&ch->trailer, DYN_H1_TRAILER);
ch->ignore_body = ignore_body;
}
void Curl_httpchunk_reset(struct Curl_easy *data, struct Curl_chunker *ch,
bool ignore_body)
{
(void)data;
ch->hexindex = 0; /* start at 0 */
ch->state = CHUNK_HEX; /* we get hex first! */
ch->last_code = CHUNKE_OK;
Curl_dyn_reset(&ch->trailer);
ch->ignore_body = ignore_body;
}
void Curl_httpchunk_free(struct Curl_easy *data, struct Curl_chunker *ch)
{
(void)data;
Curl_dyn_free(&ch->trailer);
}
bool Curl_httpchunk_is_done(struct Curl_easy *data, struct Curl_chunker *ch)
{
(void)data;
return ch->state == CHUNK_DONE;
}
static CURLcode httpchunk_readwrite(struct Curl_easy *data,
struct Curl_chunker *ch,
struct Curl_cwriter *cw_next,
const char *buf, size_t blen,
size_t *pconsumed)
{
CURLcode result = CURLE_OK;
size_t piece;
*pconsumed = 0; /* nothing's written yet */
/* first check terminal states that will not progress anywhere */
if(ch->state == CHUNK_DONE)
return CURLE_OK;
if(ch->state == CHUNK_FAILED)
return CURLE_RECV_ERROR;
/* the original data is written to the client, but we go on with the
chunk read process, to properly calculate the content length */
if(data->set.http_te_skip && !ch->ignore_body) {
if(cw_next)
result = Curl_cwriter_write(data, cw_next, CLIENTWRITE_BODY, buf, blen);
else
result = Curl_client_write(data, CLIENTWRITE_BODY, (char *)buf, blen);
if(result) {
ch->state = CHUNK_FAILED;
ch->last_code = CHUNKE_PASSTHRU_ERROR;
return result;
}
}
while(blen) {
switch(ch->state) {
case CHUNK_HEX:
if(ISXDIGIT(*buf)) {
if(ch->hexindex >= CHUNK_MAXNUM_LEN) {
failf(data, "chunk hex-length longer than %d", CHUNK_MAXNUM_LEN);
ch->state = CHUNK_FAILED;
ch->last_code = CHUNKE_TOO_LONG_HEX; /* longer than we support */
return CURLE_RECV_ERROR;
}
ch->hexbuffer[ch->hexindex++] = *buf;
buf++;
blen--;
}
else {
char *endptr;
if(0 == ch->hexindex) {
/* This is illegal data, we received junk where we expected
a hexadecimal digit. */
failf(data, "chunk hex-length char not a hex digit: 0x%x", *buf);
ch->state = CHUNK_FAILED;
ch->last_code = CHUNKE_ILLEGAL_HEX;
return CURLE_RECV_ERROR;
}
/* blen and buf are unmodified */
ch->hexbuffer[ch->hexindex] = 0;
if(curlx_strtoofft(ch->hexbuffer, &endptr, 16, &ch->datasize)) {
failf(data, "chunk hex-length not valid: '%s'", ch->hexbuffer);
ch->state = CHUNK_FAILED;
ch->last_code = CHUNKE_ILLEGAL_HEX;
return CURLE_RECV_ERROR;
}
ch->state = CHUNK_LF; /* now wait for the CRLF */
}
break;
case CHUNK_LF:
/* waiting for the LF after a chunk size */
if(*buf == 0x0a) {
/* we're now expecting data to come, unless size was zero! */
if(0 == ch->datasize) {
ch->state = CHUNK_TRAILER; /* now check for trailers */
}
else
ch->state = CHUNK_DATA;
}
buf++;
blen--;
break;
case CHUNK_DATA:
/* We expect 'datasize' of data. We have 'blen' right now, it can be
more or less than 'datasize'. Get the smallest piece.
*/
piece = blen;
if(ch->datasize < (curl_off_t)blen)
piece = curlx_sotouz(ch->datasize);
/* Write the data portion available */
if(!data->set.http_te_skip && !ch->ignore_body) {
if(cw_next)
result = Curl_cwriter_write(data, cw_next, CLIENTWRITE_BODY,
buf, piece);
else
result = Curl_client_write(data, CLIENTWRITE_BODY,
(char *)buf, piece);
if(result) {
ch->state = CHUNK_FAILED;
ch->last_code = CHUNKE_PASSTHRU_ERROR;
return result;
}
}
*pconsumed += piece;
ch->datasize -= piece; /* decrease amount left to expect */
buf += piece; /* move read pointer forward */
blen -= piece; /* decrease space left in this round */
if(0 == ch->datasize)
/* end of data this round, we now expect a trailing CRLF */
ch->state = CHUNK_POSTLF;
break;
case CHUNK_POSTLF:
if(*buf == 0x0a) {
/* The last one before we go back to hex state and start all over. */
Curl_httpchunk_reset(data, ch, ch->ignore_body);
}
else if(*buf != 0x0d) {
ch->state = CHUNK_FAILED;
ch->last_code = CHUNKE_BAD_CHUNK;
return CURLE_RECV_ERROR;
}
buf++;
blen--;
break;
case CHUNK_TRAILER:
if((*buf == 0x0d) || (*buf == 0x0a)) {
char *tr = Curl_dyn_ptr(&ch->trailer);
/* this is the end of a trailer, but if the trailer was zero bytes
there was no trailer and we move on */
if(tr) {
size_t trlen;
result = Curl_dyn_addn(&ch->trailer, (char *)STRCONST("\x0d\x0a"));
if(result) {
ch->state = CHUNK_FAILED;
ch->last_code = CHUNKE_OUT_OF_MEMORY;
return result;
}
tr = Curl_dyn_ptr(&ch->trailer);
trlen = Curl_dyn_len(&ch->trailer);
if(!data->set.http_te_skip) {
if(cw_next)
result = Curl_cwriter_write(data, cw_next,
CLIENTWRITE_HEADER|
CLIENTWRITE_TRAILER,
tr, trlen);
else
result = Curl_client_write(data,
CLIENTWRITE_HEADER|
CLIENTWRITE_TRAILER,
tr, trlen);
if(result) {
ch->state = CHUNK_FAILED;
ch->last_code = CHUNKE_PASSTHRU_ERROR;
return result;
}
}
Curl_dyn_reset(&ch->trailer);
ch->state = CHUNK_TRAILER_CR;
if(*buf == 0x0a)
/* already on the LF */
break;
}
else {
/* no trailer, we're on the final CRLF pair */
ch->state = CHUNK_TRAILER_POSTCR;
break; /* don't advance the pointer */
}
}
else {
result = Curl_dyn_addn(&ch->trailer, buf, 1);
if(result) {
ch->state = CHUNK_FAILED;
ch->last_code = CHUNKE_OUT_OF_MEMORY;
return result;
}
}
buf++;
blen--;
break;
case CHUNK_TRAILER_CR:
if(*buf == 0x0a) {
ch->state = CHUNK_TRAILER_POSTCR;
buf++;
blen--;
}
else {
ch->state = CHUNK_FAILED;
ch->last_code = CHUNKE_BAD_CHUNK;
return CURLE_RECV_ERROR;
}
break;
case CHUNK_TRAILER_POSTCR:
/* We enter this state when a CR should arrive so we expect to
have to first pass a CR before we wait for LF */
if((*buf != 0x0d) && (*buf != 0x0a)) {
/* not a CR then it must be another header in the trailer */
ch->state = CHUNK_TRAILER;
break;
}
if(*buf == 0x0d) {
/* skip if CR */
buf++;
blen--;
}
/* now wait for the final LF */
ch->state = CHUNK_STOP;
break;
case CHUNK_STOP:
if(*buf == 0x0a) {
blen--;
/* Record the length of any data left in the end of the buffer
even if there's no more chunks to read */
ch->datasize = blen;
ch->state = CHUNK_DONE;
return CURLE_OK;
}
else {
ch->state = CHUNK_FAILED;
ch->last_code = CHUNKE_BAD_CHUNK;
return CURLE_RECV_ERROR;
}
case CHUNK_DONE:
return CURLE_OK;
case CHUNK_FAILED:
return CURLE_RECV_ERROR;
}
}
return CURLE_OK;
}
static const char *Curl_chunked_strerror(CHUNKcode code)
{
switch(code) {
default:
return "OK";
case CHUNKE_TOO_LONG_HEX:
return "Too long hexadecimal number";
case CHUNKE_ILLEGAL_HEX:
return "Illegal or missing hexadecimal sequence";
case CHUNKE_BAD_CHUNK:
return "Malformed encoding found";
case CHUNKE_PASSTHRU_ERROR:
return "Error writing data to client";
case CHUNKE_BAD_ENCODING:
return "Bad content-encoding found";
case CHUNKE_OUT_OF_MEMORY:
return "Out of memory";
}
}
CURLcode Curl_httpchunk_read(struct Curl_easy *data,
struct Curl_chunker *ch,
char *buf, size_t blen,
size_t *pconsumed)
{
return httpchunk_readwrite(data, ch, NULL, buf, blen, pconsumed);
}
struct chunked_writer {
struct Curl_cwriter super;
struct Curl_chunker ch;
};
static CURLcode cw_chunked_init(struct Curl_easy *data,
struct Curl_cwriter *writer)
{
struct chunked_writer *ctx = (struct chunked_writer *)writer;
data->req.chunk = TRUE; /* chunks coming our way. */
Curl_httpchunk_init(data, &ctx->ch, FALSE);
return CURLE_OK;
}
static void cw_chunked_close(struct Curl_easy *data,
struct Curl_cwriter *writer)
{
struct chunked_writer *ctx = (struct chunked_writer *)writer;
Curl_httpchunk_free(data, &ctx->ch);
}
static CURLcode cw_chunked_write(struct Curl_easy *data,
struct Curl_cwriter *writer, int type,
const char *buf, size_t blen)
{
struct chunked_writer *ctx = (struct chunked_writer *)writer;
CURLcode result;
size_t consumed;
if(!(type & CLIENTWRITE_BODY))
return Curl_cwriter_write(data, writer->next, type, buf, blen);
consumed = 0;
result = httpchunk_readwrite(data, &ctx->ch, writer->next, buf, blen,
&consumed);
if(result) {
if(CHUNKE_PASSTHRU_ERROR == ctx->ch.last_code) {
failf(data, "Failed reading the chunked-encoded stream");
}
else {
failf(data, "%s in chunked-encoding",
Curl_chunked_strerror(ctx->ch.last_code));
}
return result;
}
blen -= consumed;
if(CHUNK_DONE == ctx->ch.state) {
/* chunks read successfully, download is complete */
data->req.download_done = TRUE;
if(blen) {
infof(data, "Leftovers after chunking: %zu bytes", blen);
}
}
else if((type & CLIENTWRITE_EOS) && !data->req.no_body) {
failf(data, "transfer closed with outstanding read data remaining");
return CURLE_PARTIAL_FILE;
}
return CURLE_OK;
}
/* HTTP chunked Transfer-Encoding decoder */
const struct Curl_cwtype Curl_httpchunk_unencoder = {
"chunked",
NULL,
cw_chunked_init,
cw_chunked_write,
cw_chunked_close,
sizeof(struct chunked_writer)
};
#endif /* CURL_DISABLE_HTTP */