2012-07-11 11:31:32 +02:00
|
|
|
/* Based on src/http/ngx_http_parse.c from NGINX copyright Igor Sysoev
|
|
|
|
*
|
|
|
|
* Additional changes are licensed under the same terms as NGINX and
|
|
|
|
* copyright Joyent, Inc. and other Node contributors. All rights reserved.
|
|
|
|
*
|
|
|
|
* Permission is hereby granted, free of charge, to any person obtaining a copy
|
|
|
|
* of this software and associated documentation files (the "Software"), to
|
|
|
|
* deal in the Software without restriction, including without limitation the
|
|
|
|
* rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
|
|
|
|
* sell copies of the Software, and to permit persons to whom the Software is
|
|
|
|
* furnished to do so, subject to the following conditions:
|
|
|
|
*
|
|
|
|
* The above copyright notice and this permission notice shall be included in
|
|
|
|
* all copies or substantial portions of the Software.
|
|
|
|
*
|
|
|
|
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
|
|
|
|
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
|
|
|
|
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
|
|
|
|
* AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
|
|
|
|
* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
|
|
|
|
* FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
|
|
|
|
* IN THE SOFTWARE.
|
|
|
|
*/
|
|
|
|
#include "http_parser.h"
|
|
|
|
#include <assert.h>
|
|
|
|
#include <stddef.h>
|
|
|
|
#include <ctype.h>
|
|
|
|
#include <stdlib.h>
|
|
|
|
#include <string.h>
|
|
|
|
#include <limits.h>
|
|
|
|
|
|
|
|
#ifndef ULLONG_MAX
|
|
|
|
# define ULLONG_MAX ((uint64_t) -1) /* 2^64-1 */
|
|
|
|
#endif
|
|
|
|
|
|
|
|
#ifndef MIN
|
|
|
|
# define MIN(a,b) ((a) < (b) ? (a) : (b))
|
|
|
|
#endif
|
|
|
|
|
2012-09-10 17:12:23 +02:00
|
|
|
#ifndef ARRAY_SIZE
|
|
|
|
# define ARRAY_SIZE(a) (sizeof(a) / sizeof((a)[0]))
|
|
|
|
#endif
|
|
|
|
|
|
|
|
#ifndef BIT_AT
|
|
|
|
# define BIT_AT(a, i) \
|
|
|
|
(!!((unsigned int) (a)[(unsigned int) (i) >> 3] & \
|
|
|
|
(1 << ((unsigned int) (i) & 7))))
|
|
|
|
#endif
|
|
|
|
|
|
|
|
#ifndef ELEM_AT
|
|
|
|
# define ELEM_AT(a, i, v) ((unsigned int) (i) < ARRAY_SIZE(a) ? (a)[(i)] : (v))
|
|
|
|
#endif
|
2012-07-11 11:31:32 +02:00
|
|
|
|
|
|
|
#define SET_ERRNO(e) \
|
|
|
|
do { \
|
|
|
|
parser->http_errno = (e); \
|
|
|
|
} while(0)
|
|
|
|
|
2015-01-05 10:28:25 +01:00
|
|
|
#define CURRENT_STATE() p_state
|
2015-02-20 11:51:41 +01:00
|
|
|
#define UPDATE_STATE(V) p_state = (enum state) (V);
|
2015-01-05 10:28:25 +01:00
|
|
|
#define RETURN(V) \
|
|
|
|
do { \
|
|
|
|
parser->state = CURRENT_STATE(); \
|
|
|
|
return (V); \
|
|
|
|
} while (0);
|
|
|
|
#define REEXECUTE() \
|
2015-02-20 11:51:41 +01:00
|
|
|
goto reexecute; \
|
2015-01-05 10:28:25 +01:00
|
|
|
|
|
|
|
|
|
|
|
#ifdef __GNUC__
|
|
|
|
# define LIKELY(X) __builtin_expect(!!(X), 1)
|
|
|
|
# define UNLIKELY(X) __builtin_expect(!!(X), 0)
|
|
|
|
#else
|
|
|
|
# define LIKELY(X) (X)
|
|
|
|
# define UNLIKELY(X) (X)
|
|
|
|
#endif
|
|
|
|
|
2012-07-11 11:31:32 +02:00
|
|
|
|
|
|
|
/* Run the notify callback FOR, returning ER if it fails */
|
|
|
|
#define CALLBACK_NOTIFY_(FOR, ER) \
|
|
|
|
do { \
|
|
|
|
assert(HTTP_PARSER_ERRNO(parser) == HPE_OK); \
|
|
|
|
\
|
2015-01-05 10:28:25 +01:00
|
|
|
if (LIKELY(settings->on_##FOR)) { \
|
|
|
|
parser->state = CURRENT_STATE(); \
|
|
|
|
if (UNLIKELY(0 != settings->on_##FOR(parser))) { \
|
2012-07-11 11:31:32 +02:00
|
|
|
SET_ERRNO(HPE_CB_##FOR); \
|
|
|
|
} \
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(parser->state); \
|
2012-07-11 11:31:32 +02:00
|
|
|
\
|
|
|
|
/* We either errored above or got paused; get out */ \
|
2015-01-05 10:28:25 +01:00
|
|
|
if (UNLIKELY(HTTP_PARSER_ERRNO(parser) != HPE_OK)) { \
|
2012-07-11 11:31:32 +02:00
|
|
|
return (ER); \
|
|
|
|
} \
|
|
|
|
} \
|
|
|
|
} while (0)
|
|
|
|
|
|
|
|
/* Run the notify callback FOR and consume the current byte */
|
|
|
|
#define CALLBACK_NOTIFY(FOR) CALLBACK_NOTIFY_(FOR, p - data + 1)
|
|
|
|
|
|
|
|
/* Run the notify callback FOR and don't consume the current byte */
|
|
|
|
#define CALLBACK_NOTIFY_NOADVANCE(FOR) CALLBACK_NOTIFY_(FOR, p - data)
|
|
|
|
|
|
|
|
/* Run data callback FOR with LEN bytes, returning ER if it fails */
|
|
|
|
#define CALLBACK_DATA_(FOR, LEN, ER) \
|
|
|
|
do { \
|
|
|
|
assert(HTTP_PARSER_ERRNO(parser) == HPE_OK); \
|
|
|
|
\
|
|
|
|
if (FOR##_mark) { \
|
2015-01-05 10:28:25 +01:00
|
|
|
if (LIKELY(settings->on_##FOR)) { \
|
|
|
|
parser->state = CURRENT_STATE(); \
|
|
|
|
if (UNLIKELY(0 != \
|
|
|
|
settings->on_##FOR(parser, FOR##_mark, (LEN)))) { \
|
2012-07-11 11:31:32 +02:00
|
|
|
SET_ERRNO(HPE_CB_##FOR); \
|
|
|
|
} \
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(parser->state); \
|
2012-07-11 11:31:32 +02:00
|
|
|
\
|
|
|
|
/* We either errored above or got paused; get out */ \
|
2015-01-05 10:28:25 +01:00
|
|
|
if (UNLIKELY(HTTP_PARSER_ERRNO(parser) != HPE_OK)) { \
|
2012-07-11 11:31:32 +02:00
|
|
|
return (ER); \
|
|
|
|
} \
|
|
|
|
} \
|
|
|
|
FOR##_mark = NULL; \
|
|
|
|
} \
|
|
|
|
} while (0)
|
2016-04-11 16:14:15 +02:00
|
|
|
|
2012-07-11 11:31:32 +02:00
|
|
|
/* Run the data callback FOR and consume the current byte */
|
|
|
|
#define CALLBACK_DATA(FOR) \
|
|
|
|
CALLBACK_DATA_(FOR, p - FOR##_mark, p - data + 1)
|
|
|
|
|
|
|
|
/* Run the data callback FOR and don't consume the current byte */
|
|
|
|
#define CALLBACK_DATA_NOADVANCE(FOR) \
|
|
|
|
CALLBACK_DATA_(FOR, p - FOR##_mark, p - data)
|
|
|
|
|
|
|
|
/* Set the mark FOR; non-destructive if mark is already set */
|
|
|
|
#define MARK(FOR) \
|
|
|
|
do { \
|
|
|
|
if (!FOR##_mark) { \
|
|
|
|
FOR##_mark = p; \
|
|
|
|
} \
|
|
|
|
} while (0)
|
|
|
|
|
2015-01-05 10:28:25 +01:00
|
|
|
/* Don't allow the total size of the HTTP headers (including the status
|
|
|
|
* line) to exceed HTTP_MAX_HEADER_SIZE. This check is here to protect
|
|
|
|
* embedders against denial-of-service attacks where the attacker feeds
|
|
|
|
* us a never-ending header that the embedder keeps buffering.
|
|
|
|
*
|
|
|
|
* This check is arguably the responsibility of embedders but we're doing
|
|
|
|
* it on the embedder's behalf because most won't bother and this way we
|
|
|
|
* make the web a little safer. HTTP_MAX_HEADER_SIZE is still far bigger
|
|
|
|
* than any reasonable request or response so this should never affect
|
|
|
|
* day-to-day operation.
|
|
|
|
*/
|
|
|
|
#define COUNT_HEADER_SIZE(V) \
|
|
|
|
do { \
|
|
|
|
parser->nread += (V); \
|
|
|
|
if (UNLIKELY(parser->nread > (HTTP_MAX_HEADER_SIZE))) { \
|
|
|
|
SET_ERRNO(HPE_HEADER_OVERFLOW); \
|
|
|
|
goto error; \
|
|
|
|
} \
|
|
|
|
} while (0)
|
|
|
|
|
2012-07-11 11:31:32 +02:00
|
|
|
|
|
|
|
#define PROXY_CONNECTION "proxy-connection"
|
|
|
|
#define CONNECTION "connection"
|
|
|
|
#define CONTENT_LENGTH "content-length"
|
|
|
|
#define TRANSFER_ENCODING "transfer-encoding"
|
|
|
|
#define UPGRADE "upgrade"
|
|
|
|
#define CHUNKED "chunked"
|
|
|
|
#define KEEP_ALIVE "keep-alive"
|
|
|
|
#define CLOSE "close"
|
|
|
|
|
|
|
|
|
|
|
|
static const char *method_strings[] =
|
|
|
|
{
|
|
|
|
#define XX(num, name, string) #string,
|
|
|
|
HTTP_METHOD_MAP(XX)
|
|
|
|
#undef XX
|
|
|
|
};
|
|
|
|
|
|
|
|
|
|
|
|
/* Tokens as defined by rfc 2616. Also lowercases them.
|
|
|
|
* token = 1*<any CHAR except CTLs or separators>
|
|
|
|
* separators = "(" | ")" | "<" | ">" | "@"
|
|
|
|
* | "," | ";" | ":" | "\" | <">
|
|
|
|
* | "/" | "[" | "]" | "?" | "="
|
|
|
|
* | "{" | "}" | SP | HT
|
|
|
|
*/
|
|
|
|
static const char tokens[256] = {
|
|
|
|
/* 0 nul 1 soh 2 stx 3 etx 4 eot 5 enq 6 ack 7 bel */
|
|
|
|
0, 0, 0, 0, 0, 0, 0, 0,
|
|
|
|
/* 8 bs 9 ht 10 nl 11 vt 12 np 13 cr 14 so 15 si */
|
|
|
|
0, 0, 0, 0, 0, 0, 0, 0,
|
|
|
|
/* 16 dle 17 dc1 18 dc2 19 dc3 20 dc4 21 nak 22 syn 23 etb */
|
|
|
|
0, 0, 0, 0, 0, 0, 0, 0,
|
|
|
|
/* 24 can 25 em 26 sub 27 esc 28 fs 29 gs 30 rs 31 us */
|
|
|
|
0, 0, 0, 0, 0, 0, 0, 0,
|
|
|
|
/* 32 sp 33 ! 34 " 35 # 36 $ 37 % 38 & 39 ' */
|
|
|
|
0, '!', 0, '#', '$', '%', '&', '\'',
|
|
|
|
/* 40 ( 41 ) 42 * 43 + 44 , 45 - 46 . 47 / */
|
|
|
|
0, 0, '*', '+', 0, '-', '.', 0,
|
|
|
|
/* 48 0 49 1 50 2 51 3 52 4 53 5 54 6 55 7 */
|
|
|
|
'0', '1', '2', '3', '4', '5', '6', '7',
|
|
|
|
/* 56 8 57 9 58 : 59 ; 60 < 61 = 62 > 63 ? */
|
|
|
|
'8', '9', 0, 0, 0, 0, 0, 0,
|
|
|
|
/* 64 @ 65 A 66 B 67 C 68 D 69 E 70 F 71 G */
|
|
|
|
0, 'a', 'b', 'c', 'd', 'e', 'f', 'g',
|
|
|
|
/* 72 H 73 I 74 J 75 K 76 L 77 M 78 N 79 O */
|
|
|
|
'h', 'i', 'j', 'k', 'l', 'm', 'n', 'o',
|
|
|
|
/* 80 P 81 Q 82 R 83 S 84 T 85 U 86 V 87 W */
|
|
|
|
'p', 'q', 'r', 's', 't', 'u', 'v', 'w',
|
|
|
|
/* 88 X 89 Y 90 Z 91 [ 92 \ 93 ] 94 ^ 95 _ */
|
|
|
|
'x', 'y', 'z', 0, 0, 0, '^', '_',
|
|
|
|
/* 96 ` 97 a 98 b 99 c 100 d 101 e 102 f 103 g */
|
|
|
|
'`', 'a', 'b', 'c', 'd', 'e', 'f', 'g',
|
|
|
|
/* 104 h 105 i 106 j 107 k 108 l 109 m 110 n 111 o */
|
|
|
|
'h', 'i', 'j', 'k', 'l', 'm', 'n', 'o',
|
|
|
|
/* 112 p 113 q 114 r 115 s 116 t 117 u 118 v 119 w */
|
|
|
|
'p', 'q', 'r', 's', 't', 'u', 'v', 'w',
|
|
|
|
/* 120 x 121 y 122 z 123 { 124 | 125 } 126 ~ 127 del */
|
|
|
|
'x', 'y', 'z', 0, '|', 0, '~', 0 };
|
|
|
|
|
|
|
|
|
|
|
|
static const int8_t unhex[256] =
|
|
|
|
{-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1
|
|
|
|
,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1
|
|
|
|
,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1
|
|
|
|
, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9,-1,-1,-1,-1,-1,-1
|
|
|
|
,-1,10,11,12,13,14,15,-1,-1,-1,-1,-1,-1,-1,-1,-1
|
|
|
|
,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1
|
|
|
|
,-1,10,11,12,13,14,15,-1,-1,-1,-1,-1,-1,-1,-1,-1
|
|
|
|
,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1
|
|
|
|
};
|
|
|
|
|
|
|
|
|
|
|
|
#if HTTP_PARSER_STRICT
|
2012-09-10 17:12:23 +02:00
|
|
|
# define T(v) 0
|
2012-07-11 11:31:32 +02:00
|
|
|
#else
|
2012-09-10 17:12:23 +02:00
|
|
|
# define T(v) v
|
2012-07-11 11:31:32 +02:00
|
|
|
#endif
|
|
|
|
|
|
|
|
|
2012-09-10 17:12:23 +02:00
|
|
|
static const uint8_t normal_url_char[32] = {
|
2012-07-11 11:31:32 +02:00
|
|
|
/* 0 nul 1 soh 2 stx 3 etx 4 eot 5 enq 6 ack 7 bel */
|
2012-09-10 17:12:23 +02:00
|
|
|
0 | 0 | 0 | 0 | 0 | 0 | 0 | 0,
|
2012-07-11 11:31:32 +02:00
|
|
|
/* 8 bs 9 ht 10 nl 11 vt 12 np 13 cr 14 so 15 si */
|
2012-09-10 17:12:23 +02:00
|
|
|
0 | T(2) | 0 | 0 | T(16) | 0 | 0 | 0,
|
2012-07-11 11:31:32 +02:00
|
|
|
/* 16 dle 17 dc1 18 dc2 19 dc3 20 dc4 21 nak 22 syn 23 etb */
|
2012-09-10 17:12:23 +02:00
|
|
|
0 | 0 | 0 | 0 | 0 | 0 | 0 | 0,
|
2012-07-11 11:31:32 +02:00
|
|
|
/* 24 can 25 em 26 sub 27 esc 28 fs 29 gs 30 rs 31 us */
|
2012-09-10 17:12:23 +02:00
|
|
|
0 | 0 | 0 | 0 | 0 | 0 | 0 | 0,
|
2012-07-11 11:31:32 +02:00
|
|
|
/* 32 sp 33 ! 34 " 35 # 36 $ 37 % 38 & 39 ' */
|
2012-09-10 17:12:23 +02:00
|
|
|
0 | 2 | 4 | 0 | 16 | 32 | 64 | 128,
|
2012-07-11 11:31:32 +02:00
|
|
|
/* 40 ( 41 ) 42 * 43 + 44 , 45 - 46 . 47 / */
|
2012-09-10 17:12:23 +02:00
|
|
|
1 | 2 | 4 | 8 | 16 | 32 | 64 | 128,
|
2012-07-11 11:31:32 +02:00
|
|
|
/* 48 0 49 1 50 2 51 3 52 4 53 5 54 6 55 7 */
|
2012-09-10 17:12:23 +02:00
|
|
|
1 | 2 | 4 | 8 | 16 | 32 | 64 | 128,
|
2012-07-11 11:31:32 +02:00
|
|
|
/* 56 8 57 9 58 : 59 ; 60 < 61 = 62 > 63 ? */
|
2012-09-10 17:12:23 +02:00
|
|
|
1 | 2 | 4 | 8 | 16 | 32 | 64 | 0,
|
2012-07-11 11:31:32 +02:00
|
|
|
/* 64 @ 65 A 66 B 67 C 68 D 69 E 70 F 71 G */
|
2012-09-10 17:12:23 +02:00
|
|
|
1 | 2 | 4 | 8 | 16 | 32 | 64 | 128,
|
2012-07-11 11:31:32 +02:00
|
|
|
/* 72 H 73 I 74 J 75 K 76 L 77 M 78 N 79 O */
|
2012-09-10 17:12:23 +02:00
|
|
|
1 | 2 | 4 | 8 | 16 | 32 | 64 | 128,
|
2012-07-11 11:31:32 +02:00
|
|
|
/* 80 P 81 Q 82 R 83 S 84 T 85 U 86 V 87 W */
|
2012-09-10 17:12:23 +02:00
|
|
|
1 | 2 | 4 | 8 | 16 | 32 | 64 | 128,
|
2012-07-11 11:31:32 +02:00
|
|
|
/* 88 X 89 Y 90 Z 91 [ 92 \ 93 ] 94 ^ 95 _ */
|
2012-09-10 17:12:23 +02:00
|
|
|
1 | 2 | 4 | 8 | 16 | 32 | 64 | 128,
|
2012-07-11 11:31:32 +02:00
|
|
|
/* 96 ` 97 a 98 b 99 c 100 d 101 e 102 f 103 g */
|
2012-09-10 17:12:23 +02:00
|
|
|
1 | 2 | 4 | 8 | 16 | 32 | 64 | 128,
|
2012-07-11 11:31:32 +02:00
|
|
|
/* 104 h 105 i 106 j 107 k 108 l 109 m 110 n 111 o */
|
2012-09-10 17:12:23 +02:00
|
|
|
1 | 2 | 4 | 8 | 16 | 32 | 64 | 128,
|
2012-07-11 11:31:32 +02:00
|
|
|
/* 112 p 113 q 114 r 115 s 116 t 117 u 118 v 119 w */
|
2012-09-10 17:12:23 +02:00
|
|
|
1 | 2 | 4 | 8 | 16 | 32 | 64 | 128,
|
2012-07-11 11:31:32 +02:00
|
|
|
/* 120 x 121 y 122 z 123 { 124 | 125 } 126 ~ 127 del */
|
2012-09-10 17:12:23 +02:00
|
|
|
1 | 2 | 4 | 8 | 16 | 32 | 64 | 0, };
|
2012-07-11 11:31:32 +02:00
|
|
|
|
|
|
|
#undef T
|
|
|
|
|
|
|
|
enum state
|
|
|
|
{ s_dead = 1 /* important that this is > 0 */
|
|
|
|
|
|
|
|
, s_start_req_or_res
|
2016-05-28 09:08:20 +02:00
|
|
|
, s_res_or_resp_mark_H
|
2012-07-11 11:31:32 +02:00
|
|
|
, s_res_or_resp_H
|
|
|
|
, s_start_res
|
|
|
|
, s_res_H
|
|
|
|
, s_res_HT
|
|
|
|
, s_res_HTT
|
|
|
|
, s_res_HTTP
|
|
|
|
, s_res_first_http_major
|
|
|
|
, s_res_http_major
|
|
|
|
, s_res_first_http_minor
|
|
|
|
, s_res_http_minor
|
|
|
|
, s_res_first_status_code
|
|
|
|
, s_res_status_code
|
2014-05-15 17:23:03 +02:00
|
|
|
, s_res_status_start
|
2012-07-11 11:31:32 +02:00
|
|
|
, s_res_status
|
|
|
|
, s_res_line_almost_done
|
|
|
|
|
|
|
|
, s_start_req
|
|
|
|
|
2016-05-28 09:08:20 +02:00
|
|
|
, s_req_method_start
|
2012-07-11 11:31:32 +02:00
|
|
|
, s_req_method
|
2016-05-28 09:08:20 +02:00
|
|
|
#if HTTP_PARSER_METHOD_CB
|
|
|
|
, s_req_method_unknown
|
|
|
|
#endif
|
2012-07-11 11:31:32 +02:00
|
|
|
, s_req_spaces_before_url
|
|
|
|
, s_req_schema
|
|
|
|
, s_req_schema_slash
|
|
|
|
, s_req_schema_slash_slash
|
2012-09-10 17:12:23 +02:00
|
|
|
, s_req_server_start
|
|
|
|
, s_req_server
|
|
|
|
, s_req_server_with_at
|
2012-07-11 11:31:32 +02:00
|
|
|
, s_req_path
|
|
|
|
, s_req_query_string_start
|
|
|
|
, s_req_query_string
|
|
|
|
, s_req_fragment_start
|
|
|
|
, s_req_fragment
|
|
|
|
, s_req_http_start
|
|
|
|
, s_req_http_H
|
|
|
|
, s_req_http_HT
|
|
|
|
, s_req_http_HTT
|
|
|
|
, s_req_http_HTTP
|
|
|
|
, s_req_first_http_major
|
|
|
|
, s_req_http_major
|
|
|
|
, s_req_first_http_minor
|
|
|
|
, s_req_http_minor
|
|
|
|
, s_req_line_almost_done
|
|
|
|
|
|
|
|
, s_header_field_start
|
|
|
|
, s_header_field
|
2014-05-15 17:23:03 +02:00
|
|
|
, s_header_value_discard_ws
|
|
|
|
, s_header_value_discard_ws_almost_done
|
|
|
|
, s_header_value_discard_lws
|
2012-07-11 11:31:32 +02:00
|
|
|
, s_header_value_start
|
|
|
|
, s_header_value
|
|
|
|
, s_header_value_lws
|
|
|
|
|
|
|
|
, s_header_almost_done
|
|
|
|
|
|
|
|
, s_chunk_size_start
|
|
|
|
, s_chunk_size
|
|
|
|
, s_chunk_parameters
|
|
|
|
, s_chunk_size_almost_done
|
|
|
|
|
|
|
|
, s_headers_almost_done
|
|
|
|
, s_headers_done
|
|
|
|
|
|
|
|
/* Important: 's_headers_done' must be the last 'header' state. All
|
|
|
|
* states beyond this must be 'body' states. It is used for overflow
|
|
|
|
* checking. See the PARSING_HEADER() macro.
|
|
|
|
*/
|
|
|
|
|
|
|
|
, s_chunk_data
|
|
|
|
, s_chunk_data_almost_done
|
|
|
|
, s_chunk_data_done
|
|
|
|
|
|
|
|
, s_body_identity
|
|
|
|
, s_body_identity_eof
|
|
|
|
|
|
|
|
, s_message_done
|
|
|
|
};
|
|
|
|
|
|
|
|
|
|
|
|
#define PARSING_HEADER(state) (state <= s_headers_done)
|
|
|
|
|
|
|
|
|
|
|
|
enum header_states
|
|
|
|
{ h_general = 0
|
|
|
|
, h_C
|
|
|
|
, h_CO
|
|
|
|
, h_CON
|
|
|
|
|
|
|
|
, h_matching_connection
|
|
|
|
, h_matching_proxy_connection
|
|
|
|
, h_matching_content_length
|
|
|
|
, h_matching_transfer_encoding
|
|
|
|
, h_matching_upgrade
|
|
|
|
|
|
|
|
, h_connection
|
|
|
|
, h_content_length
|
|
|
|
, h_transfer_encoding
|
|
|
|
, h_upgrade
|
|
|
|
|
|
|
|
, h_matching_transfer_encoding_chunked
|
2015-01-05 10:28:25 +01:00
|
|
|
, h_matching_connection_token_start
|
2012-07-11 11:31:32 +02:00
|
|
|
, h_matching_connection_keep_alive
|
|
|
|
, h_matching_connection_close
|
2015-01-05 10:28:25 +01:00
|
|
|
, h_matching_connection_upgrade
|
|
|
|
, h_matching_connection_token
|
2012-07-11 11:31:32 +02:00
|
|
|
|
|
|
|
, h_transfer_encoding_chunked
|
|
|
|
, h_connection_keep_alive
|
|
|
|
, h_connection_close
|
2015-01-05 10:28:25 +01:00
|
|
|
, h_connection_upgrade
|
2012-07-11 11:31:32 +02:00
|
|
|
};
|
|
|
|
|
2012-09-10 17:12:23 +02:00
|
|
|
enum http_host_state
|
|
|
|
{
|
|
|
|
s_http_host_dead = 1
|
|
|
|
, s_http_userinfo_start
|
|
|
|
, s_http_userinfo
|
|
|
|
, s_http_host_start
|
|
|
|
, s_http_host_v6_start
|
|
|
|
, s_http_host
|
|
|
|
, s_http_host_v6
|
|
|
|
, s_http_host_v6_end
|
2015-07-25 05:53:17 +02:00
|
|
|
, s_http_host_v6_zone_start
|
|
|
|
, s_http_host_v6_zone
|
2012-09-10 17:12:23 +02:00
|
|
|
, s_http_host_port_start
|
|
|
|
, s_http_host_port
|
|
|
|
};
|
2012-07-11 11:31:32 +02:00
|
|
|
|
|
|
|
/* Macros for character classes; depends on strict-mode */
|
|
|
|
#define CR '\r'
|
|
|
|
#define LF '\n'
|
|
|
|
#define LOWER(c) (unsigned char)(c | 0x20)
|
|
|
|
#define IS_ALPHA(c) (LOWER(c) >= 'a' && LOWER(c) <= 'z')
|
|
|
|
#define IS_NUM(c) ((c) >= '0' && (c) <= '9')
|
|
|
|
#define IS_ALPHANUM(c) (IS_ALPHA(c) || IS_NUM(c))
|
|
|
|
#define IS_HEX(c) (IS_NUM(c) || (LOWER(c) >= 'a' && LOWER(c) <= 'f'))
|
2012-09-10 17:12:23 +02:00
|
|
|
#define IS_MARK(c) ((c) == '-' || (c) == '_' || (c) == '.' || \
|
|
|
|
(c) == '!' || (c) == '~' || (c) == '*' || (c) == '\'' || (c) == '(' || \
|
|
|
|
(c) == ')')
|
|
|
|
#define IS_USERINFO_CHAR(c) (IS_ALPHANUM(c) || IS_MARK(c) || (c) == '%' || \
|
|
|
|
(c) == ';' || (c) == ':' || (c) == '&' || (c) == '=' || (c) == '+' || \
|
|
|
|
(c) == '$' || (c) == ',')
|
2012-07-11 11:31:32 +02:00
|
|
|
|
2015-01-05 10:28:25 +01:00
|
|
|
#define STRICT_TOKEN(c) (tokens[(unsigned char)c])
|
|
|
|
|
2012-07-11 11:31:32 +02:00
|
|
|
#if HTTP_PARSER_STRICT
|
|
|
|
#define TOKEN(c) (tokens[(unsigned char)c])
|
2012-09-10 17:12:23 +02:00
|
|
|
#define IS_URL_CHAR(c) (BIT_AT(normal_url_char, (unsigned char)c))
|
2012-07-11 11:31:32 +02:00
|
|
|
#define IS_HOST_CHAR(c) (IS_ALPHANUM(c) || (c) == '.' || (c) == '-')
|
|
|
|
#else
|
|
|
|
#define TOKEN(c) ((c == ' ') ? ' ' : tokens[(unsigned char)c])
|
|
|
|
#define IS_URL_CHAR(c) \
|
2012-09-10 17:12:23 +02:00
|
|
|
(BIT_AT(normal_url_char, (unsigned char)c) || ((c) & 0x80))
|
2012-07-11 11:31:32 +02:00
|
|
|
#define IS_HOST_CHAR(c) \
|
|
|
|
(IS_ALPHANUM(c) || (c) == '.' || (c) == '-' || (c) == '_')
|
|
|
|
#endif
|
|
|
|
|
2016-04-11 16:14:15 +02:00
|
|
|
/**
|
|
|
|
* Verify that a char is a valid visible (printable) US-ASCII
|
|
|
|
* character or %x80-FF
|
|
|
|
**/
|
|
|
|
#define IS_HEADER_CHAR(ch) \
|
|
|
|
(ch == CR || ch == LF || ch == 9 || ((unsigned char)ch > 31 && ch != 127))
|
2012-07-11 11:31:32 +02:00
|
|
|
|
|
|
|
#define start_state (parser->type == HTTP_REQUEST ? s_start_req : s_start_res)
|
|
|
|
|
|
|
|
|
|
|
|
#if HTTP_PARSER_STRICT
|
|
|
|
# define STRICT_CHECK(cond) \
|
|
|
|
do { \
|
|
|
|
if (cond) { \
|
|
|
|
SET_ERRNO(HPE_STRICT); \
|
|
|
|
goto error; \
|
|
|
|
} \
|
|
|
|
} while (0)
|
|
|
|
# define NEW_MESSAGE() (http_should_keep_alive(parser) ? start_state : s_dead)
|
|
|
|
#else
|
|
|
|
# define STRICT_CHECK(cond)
|
|
|
|
# define NEW_MESSAGE() start_state
|
|
|
|
#endif
|
|
|
|
|
|
|
|
|
|
|
|
/* Map errno values to strings for human-readable output */
|
|
|
|
#define HTTP_STRERROR_GEN(n, s) { "HPE_" #n, s },
|
|
|
|
static struct {
|
|
|
|
const char *name;
|
|
|
|
const char *description;
|
|
|
|
} http_strerror_tab[] = {
|
|
|
|
HTTP_ERRNO_MAP(HTTP_STRERROR_GEN)
|
|
|
|
};
|
|
|
|
#undef HTTP_STRERROR_GEN
|
|
|
|
|
2012-09-10 17:12:23 +02:00
|
|
|
int http_message_needs_eof(const http_parser *parser);
|
2012-07-11 11:31:32 +02:00
|
|
|
|
|
|
|
/* Our URL parser.
|
|
|
|
*
|
|
|
|
* This is designed to be shared by http_parser_execute() for URL validation,
|
|
|
|
* hence it has a state transition + byte-for-byte interface. In addition, it
|
|
|
|
* is meant to be embedded in http_parser_parse_url(), which does the dirty
|
|
|
|
* work of turning state transitions URL components for its API.
|
|
|
|
*
|
|
|
|
* This function should only be invoked with non-space characters. It is
|
|
|
|
* assumed that the caller cares about (and can detect) the transition between
|
|
|
|
* URL and non-URL states by looking for these.
|
|
|
|
*/
|
|
|
|
static enum state
|
|
|
|
parse_url_char(enum state s, const char ch)
|
|
|
|
{
|
|
|
|
if (ch == ' ' || ch == '\r' || ch == '\n') {
|
|
|
|
return s_dead;
|
|
|
|
}
|
|
|
|
|
|
|
|
#if HTTP_PARSER_STRICT
|
|
|
|
if (ch == '\t' || ch == '\f') {
|
|
|
|
return s_dead;
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
|
|
|
|
switch (s) {
|
|
|
|
case s_req_spaces_before_url:
|
|
|
|
/* Proxied requests are followed by scheme of an absolute URI (alpha).
|
|
|
|
* All methods except CONNECT are followed by '/' or '*'.
|
|
|
|
*/
|
|
|
|
|
|
|
|
if (ch == '/' || ch == '*') {
|
|
|
|
return s_req_path;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (IS_ALPHA(ch)) {
|
|
|
|
return s_req_schema;
|
|
|
|
}
|
|
|
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
case s_req_schema:
|
|
|
|
if (IS_ALPHA(ch)) {
|
|
|
|
return s;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (ch == ':') {
|
|
|
|
return s_req_schema_slash;
|
|
|
|
}
|
|
|
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
case s_req_schema_slash:
|
|
|
|
if (ch == '/') {
|
|
|
|
return s_req_schema_slash_slash;
|
|
|
|
}
|
|
|
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
case s_req_schema_slash_slash:
|
|
|
|
if (ch == '/') {
|
2012-09-10 17:12:23 +02:00
|
|
|
return s_req_server_start;
|
2012-07-11 11:31:32 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
break;
|
|
|
|
|
2012-09-10 17:12:23 +02:00
|
|
|
case s_req_server_with_at:
|
|
|
|
if (ch == '@') {
|
|
|
|
return s_dead;
|
2012-07-11 11:31:32 +02:00
|
|
|
}
|
|
|
|
|
2012-09-10 17:12:23 +02:00
|
|
|
/* FALLTHROUGH */
|
|
|
|
case s_req_server_start:
|
|
|
|
case s_req_server:
|
|
|
|
if (ch == '/') {
|
|
|
|
return s_req_path;
|
2012-07-11 11:31:32 +02:00
|
|
|
}
|
|
|
|
|
2012-09-10 17:12:23 +02:00
|
|
|
if (ch == '?') {
|
|
|
|
return s_req_query_string_start;
|
2012-07-11 11:31:32 +02:00
|
|
|
}
|
|
|
|
|
2012-09-10 17:12:23 +02:00
|
|
|
if (ch == '@') {
|
|
|
|
return s_req_server_with_at;
|
2012-07-11 11:31:32 +02:00
|
|
|
}
|
|
|
|
|
2012-09-10 17:12:23 +02:00
|
|
|
if (IS_USERINFO_CHAR(ch) || ch == '[' || ch == ']') {
|
|
|
|
return s_req_server;
|
2012-07-11 11:31:32 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
case s_req_path:
|
|
|
|
if (IS_URL_CHAR(ch)) {
|
|
|
|
return s;
|
|
|
|
}
|
|
|
|
|
|
|
|
switch (ch) {
|
|
|
|
case '?':
|
|
|
|
return s_req_query_string_start;
|
|
|
|
|
|
|
|
case '#':
|
|
|
|
return s_req_fragment_start;
|
|
|
|
}
|
|
|
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
case s_req_query_string_start:
|
|
|
|
case s_req_query_string:
|
|
|
|
if (IS_URL_CHAR(ch)) {
|
|
|
|
return s_req_query_string;
|
|
|
|
}
|
|
|
|
|
|
|
|
switch (ch) {
|
|
|
|
case '?':
|
|
|
|
/* allow extra '?' in query string */
|
|
|
|
return s_req_query_string;
|
|
|
|
|
|
|
|
case '#':
|
|
|
|
return s_req_fragment_start;
|
|
|
|
}
|
|
|
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
case s_req_fragment_start:
|
|
|
|
if (IS_URL_CHAR(ch)) {
|
|
|
|
return s_req_fragment;
|
|
|
|
}
|
|
|
|
|
|
|
|
switch (ch) {
|
|
|
|
case '?':
|
|
|
|
return s_req_fragment;
|
|
|
|
|
|
|
|
case '#':
|
|
|
|
return s;
|
|
|
|
}
|
|
|
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
case s_req_fragment:
|
|
|
|
if (IS_URL_CHAR(ch)) {
|
|
|
|
return s;
|
|
|
|
}
|
|
|
|
|
|
|
|
switch (ch) {
|
|
|
|
case '?':
|
|
|
|
case '#':
|
|
|
|
return s;
|
|
|
|
}
|
|
|
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
default:
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* We should never fall out of the switch above unless there's an error */
|
|
|
|
return s_dead;
|
|
|
|
}
|
|
|
|
|
|
|
|
size_t http_parser_execute (http_parser *parser,
|
|
|
|
const http_parser_settings *settings,
|
|
|
|
const char *data,
|
|
|
|
size_t len)
|
|
|
|
{
|
|
|
|
char c, ch;
|
|
|
|
int8_t unhex_val;
|
|
|
|
const char *p = data;
|
|
|
|
const char *header_field_mark = 0;
|
|
|
|
const char *header_value_mark = 0;
|
|
|
|
const char *url_mark = 0;
|
|
|
|
const char *body_mark = 0;
|
2014-05-15 17:23:03 +02:00
|
|
|
const char *status_mark = 0;
|
2016-05-28 09:08:20 +02:00
|
|
|
#if HTTP_PARSER_METHOD_CB
|
|
|
|
const char *method_mark = 0;
|
|
|
|
#endif
|
2015-02-20 11:51:41 +01:00
|
|
|
enum state p_state = (enum state) parser->state;
|
2016-04-11 16:14:15 +02:00
|
|
|
const unsigned int lenient = parser->lenient_http_headers;
|
2012-07-11 11:31:32 +02:00
|
|
|
|
|
|
|
/* We're in an error state. Don't bother doing anything. */
|
|
|
|
if (HTTP_PARSER_ERRNO(parser) != HPE_OK) {
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (len == 0) {
|
2015-01-05 10:28:25 +01:00
|
|
|
switch (CURRENT_STATE()) {
|
2012-07-11 11:31:32 +02:00
|
|
|
case s_body_identity_eof:
|
|
|
|
/* Use of CALLBACK_NOTIFY() here would erroneously return 1 byte read if
|
|
|
|
* we got paused.
|
|
|
|
*/
|
|
|
|
CALLBACK_NOTIFY_NOADVANCE(message_complete);
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
case s_dead:
|
|
|
|
case s_start_req_or_res:
|
|
|
|
case s_start_res:
|
|
|
|
case s_start_req:
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
default:
|
|
|
|
SET_ERRNO(HPE_INVALID_EOF_STATE);
|
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
|
2015-01-05 10:28:25 +01:00
|
|
|
if (CURRENT_STATE() == s_header_field)
|
2012-07-11 11:31:32 +02:00
|
|
|
header_field_mark = data;
|
2015-01-05 10:28:25 +01:00
|
|
|
if (CURRENT_STATE() == s_header_value)
|
2012-07-11 11:31:32 +02:00
|
|
|
header_value_mark = data;
|
2015-01-05 10:28:25 +01:00
|
|
|
switch (CURRENT_STATE()) {
|
2012-07-11 11:31:32 +02:00
|
|
|
case s_req_path:
|
|
|
|
case s_req_schema:
|
|
|
|
case s_req_schema_slash:
|
|
|
|
case s_req_schema_slash_slash:
|
2012-09-10 17:12:23 +02:00
|
|
|
case s_req_server_start:
|
|
|
|
case s_req_server:
|
|
|
|
case s_req_server_with_at:
|
2012-07-11 11:31:32 +02:00
|
|
|
case s_req_query_string_start:
|
|
|
|
case s_req_query_string:
|
|
|
|
case s_req_fragment_start:
|
|
|
|
case s_req_fragment:
|
|
|
|
url_mark = data;
|
|
|
|
break;
|
2016-05-28 09:08:20 +02:00
|
|
|
#if HTTP_PARSER_METHOD_CB
|
|
|
|
case s_res_or_resp_H:
|
|
|
|
case s_req_method_start:
|
|
|
|
case s_req_method:
|
|
|
|
case s_req_method_unknown:
|
|
|
|
method_mark = data;
|
|
|
|
break;
|
|
|
|
#endif
|
2014-05-15 17:23:03 +02:00
|
|
|
case s_res_status:
|
|
|
|
status_mark = data;
|
|
|
|
break;
|
2015-01-05 10:28:25 +01:00
|
|
|
default:
|
|
|
|
break;
|
2012-07-11 11:31:32 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
for (p=data; p != data + len; p++) {
|
|
|
|
ch = *p;
|
|
|
|
|
2015-01-05 10:28:25 +01:00
|
|
|
if (PARSING_HEADER(CURRENT_STATE()))
|
|
|
|
COUNT_HEADER_SIZE(1);
|
2012-07-11 11:31:32 +02:00
|
|
|
|
2015-02-20 11:51:41 +01:00
|
|
|
reexecute:
|
2015-01-05 10:28:25 +01:00
|
|
|
switch (CURRENT_STATE()) {
|
2012-07-11 11:31:32 +02:00
|
|
|
|
|
|
|
case s_dead:
|
|
|
|
/* this state is used after a 'Connection: close' message
|
|
|
|
* the parser will error out if it reads another message
|
|
|
|
*/
|
2015-01-05 10:28:25 +01:00
|
|
|
if (LIKELY(ch == CR || ch == LF))
|
2012-07-11 11:31:32 +02:00
|
|
|
break;
|
|
|
|
|
|
|
|
SET_ERRNO(HPE_CLOSED_CONNECTION);
|
|
|
|
goto error;
|
|
|
|
|
|
|
|
case s_start_req_or_res:
|
|
|
|
{
|
|
|
|
if (ch == CR || ch == LF)
|
|
|
|
break;
|
|
|
|
parser->flags = 0;
|
|
|
|
parser->content_length = ULLONG_MAX;
|
|
|
|
|
|
|
|
if (ch == 'H') {
|
2016-05-28 09:08:20 +02:00
|
|
|
UPDATE_STATE(s_res_or_resp_mark_H);
|
2012-07-11 11:31:32 +02:00
|
|
|
|
2016-05-28 09:08:20 +02:00
|
|
|
CALLBACK_NOTIFY_NOADVANCE(message_begin);
|
|
|
|
REEXECUTE();
|
2012-07-11 11:31:32 +02:00
|
|
|
} else {
|
|
|
|
parser->type = HTTP_REQUEST;
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_start_req);
|
|
|
|
REEXECUTE();
|
2012-07-11 11:31:32 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
2016-05-28 09:08:20 +02:00
|
|
|
case s_res_or_resp_mark_H:
|
|
|
|
assert(ch == 'H');
|
|
|
|
UPDATE_STATE(s_res_or_resp_H);
|
|
|
|
#if HTTP_PARSER_METHOD_CB
|
|
|
|
/* TODO This will call on_method even if type == HTTP_BOTH is
|
|
|
|
* used and it later turns out that this is response.
|
|
|
|
* Automatic handling bites us hard.
|
|
|
|
*/
|
|
|
|
MARK(method);
|
|
|
|
#endif
|
|
|
|
break;
|
|
|
|
|
2012-07-11 11:31:32 +02:00
|
|
|
case s_res_or_resp_H:
|
|
|
|
if (ch == 'T') {
|
|
|
|
parser->type = HTTP_RESPONSE;
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_res_HT);
|
2012-07-11 11:31:32 +02:00
|
|
|
} else {
|
2016-05-28 09:08:20 +02:00
|
|
|
#if HTTP_PARSER_METHOD_CB
|
|
|
|
if (UNLIKELY(!TOKEN(ch))) {
|
|
|
|
SET_ERRNO(HPE_INVALID_CONSTANT);
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (UNLIKELY(ch != 'E')) {
|
|
|
|
parser->method = HTTP_METHOD_UNKNOWN;
|
|
|
|
} else {
|
|
|
|
parser->method = HTTP_HEAD;
|
|
|
|
parser->index = 2;
|
|
|
|
}
|
|
|
|
#else
|
2015-01-05 10:28:25 +01:00
|
|
|
if (UNLIKELY(ch != 'E')) {
|
2012-07-11 11:31:32 +02:00
|
|
|
SET_ERRNO(HPE_INVALID_CONSTANT);
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
|
|
|
|
parser->method = HTTP_HEAD;
|
|
|
|
parser->index = 2;
|
2016-05-28 09:08:20 +02:00
|
|
|
#endif
|
|
|
|
|
|
|
|
parser->type = HTTP_REQUEST;
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_req_method);
|
2012-07-11 11:31:32 +02:00
|
|
|
}
|
|
|
|
break;
|
|
|
|
|
|
|
|
case s_start_res:
|
|
|
|
{
|
|
|
|
parser->flags = 0;
|
|
|
|
parser->content_length = ULLONG_MAX;
|
|
|
|
|
|
|
|
switch (ch) {
|
|
|
|
case 'H':
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_res_H);
|
2012-07-11 11:31:32 +02:00
|
|
|
break;
|
|
|
|
|
|
|
|
case CR:
|
|
|
|
case LF:
|
|
|
|
break;
|
|
|
|
|
|
|
|
default:
|
|
|
|
SET_ERRNO(HPE_INVALID_CONSTANT);
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
|
|
|
|
CALLBACK_NOTIFY(message_begin);
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
case s_res_H:
|
|
|
|
STRICT_CHECK(ch != 'T');
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_res_HT);
|
2012-07-11 11:31:32 +02:00
|
|
|
break;
|
|
|
|
|
|
|
|
case s_res_HT:
|
|
|
|
STRICT_CHECK(ch != 'T');
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_res_HTT);
|
2012-07-11 11:31:32 +02:00
|
|
|
break;
|
|
|
|
|
|
|
|
case s_res_HTT:
|
|
|
|
STRICT_CHECK(ch != 'P');
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_res_HTTP);
|
2012-07-11 11:31:32 +02:00
|
|
|
break;
|
|
|
|
|
|
|
|
case s_res_HTTP:
|
|
|
|
STRICT_CHECK(ch != '/');
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_res_first_http_major);
|
2012-07-11 11:31:32 +02:00
|
|
|
break;
|
|
|
|
|
|
|
|
case s_res_first_http_major:
|
2015-01-05 10:28:25 +01:00
|
|
|
if (UNLIKELY(ch < '0' || ch > '9')) {
|
2012-07-11 11:31:32 +02:00
|
|
|
SET_ERRNO(HPE_INVALID_VERSION);
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
|
|
|
|
parser->http_major = ch - '0';
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_res_http_major);
|
2012-07-11 11:31:32 +02:00
|
|
|
break;
|
|
|
|
|
|
|
|
/* major HTTP version or dot */
|
|
|
|
case s_res_http_major:
|
|
|
|
{
|
|
|
|
if (ch == '.') {
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_res_first_http_minor);
|
2012-07-11 11:31:32 +02:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (!IS_NUM(ch)) {
|
|
|
|
SET_ERRNO(HPE_INVALID_VERSION);
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
|
|
|
|
parser->http_major *= 10;
|
|
|
|
parser->http_major += ch - '0';
|
|
|
|
|
2015-01-05 10:28:25 +01:00
|
|
|
if (UNLIKELY(parser->http_major > 999)) {
|
2012-07-11 11:31:32 +02:00
|
|
|
SET_ERRNO(HPE_INVALID_VERSION);
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* first digit of minor HTTP version */
|
|
|
|
case s_res_first_http_minor:
|
2015-01-05 10:28:25 +01:00
|
|
|
if (UNLIKELY(!IS_NUM(ch))) {
|
2012-07-11 11:31:32 +02:00
|
|
|
SET_ERRNO(HPE_INVALID_VERSION);
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
|
|
|
|
parser->http_minor = ch - '0';
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_res_http_minor);
|
2012-07-11 11:31:32 +02:00
|
|
|
break;
|
|
|
|
|
|
|
|
/* minor HTTP version or end of request line */
|
|
|
|
case s_res_http_minor:
|
|
|
|
{
|
|
|
|
if (ch == ' ') {
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_res_first_status_code);
|
2012-07-11 11:31:32 +02:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
2015-01-05 10:28:25 +01:00
|
|
|
if (UNLIKELY(!IS_NUM(ch))) {
|
2012-07-11 11:31:32 +02:00
|
|
|
SET_ERRNO(HPE_INVALID_VERSION);
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
|
|
|
|
parser->http_minor *= 10;
|
|
|
|
parser->http_minor += ch - '0';
|
|
|
|
|
2015-01-05 10:28:25 +01:00
|
|
|
if (UNLIKELY(parser->http_minor > 999)) {
|
2012-07-11 11:31:32 +02:00
|
|
|
SET_ERRNO(HPE_INVALID_VERSION);
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
case s_res_first_status_code:
|
|
|
|
{
|
|
|
|
if (!IS_NUM(ch)) {
|
|
|
|
if (ch == ' ') {
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
SET_ERRNO(HPE_INVALID_STATUS);
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
parser->status_code = ch - '0';
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_res_status_code);
|
2012-07-11 11:31:32 +02:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
case s_res_status_code:
|
|
|
|
{
|
|
|
|
if (!IS_NUM(ch)) {
|
|
|
|
switch (ch) {
|
|
|
|
case ' ':
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_res_status_start);
|
2012-07-11 11:31:32 +02:00
|
|
|
break;
|
|
|
|
case CR:
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_res_line_almost_done);
|
2012-07-11 11:31:32 +02:00
|
|
|
break;
|
|
|
|
case LF:
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_header_field_start);
|
2012-07-11 11:31:32 +02:00
|
|
|
break;
|
|
|
|
default:
|
|
|
|
SET_ERRNO(HPE_INVALID_STATUS);
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
parser->status_code *= 10;
|
|
|
|
parser->status_code += ch - '0';
|
|
|
|
|
2015-01-05 10:28:25 +01:00
|
|
|
if (UNLIKELY(parser->status_code > 999)) {
|
2012-07-11 11:31:32 +02:00
|
|
|
SET_ERRNO(HPE_INVALID_STATUS);
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
2014-05-15 17:23:03 +02:00
|
|
|
case s_res_status_start:
|
|
|
|
{
|
|
|
|
if (ch == CR) {
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_res_line_almost_done);
|
2014-05-15 17:23:03 +02:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (ch == LF) {
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_header_field_start);
|
2014-05-15 17:23:03 +02:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
MARK(status);
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_res_status);
|
2014-05-15 17:23:03 +02:00
|
|
|
parser->index = 0;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
2012-07-11 11:31:32 +02:00
|
|
|
case s_res_status:
|
|
|
|
if (ch == CR) {
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_res_line_almost_done);
|
2014-05-15 17:23:03 +02:00
|
|
|
CALLBACK_DATA(status);
|
2012-07-11 11:31:32 +02:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (ch == LF) {
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_header_field_start);
|
2014-05-15 17:23:03 +02:00
|
|
|
CALLBACK_DATA(status);
|
2012-07-11 11:31:32 +02:00
|
|
|
break;
|
|
|
|
}
|
2014-05-15 17:23:03 +02:00
|
|
|
|
2012-07-11 11:31:32 +02:00
|
|
|
break;
|
|
|
|
|
|
|
|
case s_res_line_almost_done:
|
|
|
|
STRICT_CHECK(ch != LF);
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_header_field_start);
|
2012-07-11 11:31:32 +02:00
|
|
|
break;
|
|
|
|
|
|
|
|
case s_start_req:
|
|
|
|
if (ch == CR || ch == LF)
|
|
|
|
break;
|
|
|
|
parser->flags = 0;
|
|
|
|
parser->content_length = ULLONG_MAX;
|
|
|
|
|
2015-01-05 10:28:25 +01:00
|
|
|
if (UNLIKELY(!IS_ALPHA(ch))) {
|
2012-07-11 11:31:32 +02:00
|
|
|
SET_ERRNO(HPE_INVALID_METHOD);
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
|
2016-05-28 09:08:20 +02:00
|
|
|
UPDATE_STATE(s_req_method_start);
|
|
|
|
#if HTTP_PARSER_METHOD_CB
|
|
|
|
MARK(method);
|
|
|
|
#endif
|
|
|
|
|
|
|
|
CALLBACK_NOTIFY_NOADVANCE(message_begin);
|
|
|
|
REEXECUTE();
|
|
|
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
case s_req_method_start:
|
|
|
|
{
|
|
|
|
enum state next_state = s_req_method;
|
|
|
|
|
|
|
|
if (UNLIKELY(!IS_ALPHA(ch))) {
|
|
|
|
SET_ERRNO(HPE_INVALID_METHOD);
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
|
2012-07-11 11:31:32 +02:00
|
|
|
parser->method = (enum http_method) 0;
|
|
|
|
parser->index = 1;
|
|
|
|
switch (ch) {
|
2015-07-25 05:53:17 +02:00
|
|
|
case 'A': parser->method = HTTP_ACL; break;
|
|
|
|
case 'B': parser->method = HTTP_BIND; break;
|
2012-07-11 11:31:32 +02:00
|
|
|
case 'C': parser->method = HTTP_CONNECT; /* or COPY, CHECKOUT */ break;
|
|
|
|
case 'D': parser->method = HTTP_DELETE; break;
|
|
|
|
case 'G': parser->method = HTTP_GET; break;
|
|
|
|
case 'H': parser->method = HTTP_HEAD; break;
|
2016-01-27 12:46:59 +01:00
|
|
|
case 'L': parser->method = HTTP_LOCK; /* or LINK */ break;
|
2014-08-08 13:12:34 +02:00
|
|
|
case 'M': parser->method = HTTP_MKCOL; /* or MOVE, MKACTIVITY, MERGE, M-SEARCH, MKCALENDAR */ break;
|
2012-07-11 11:31:32 +02:00
|
|
|
case 'N': parser->method = HTTP_NOTIFY; break;
|
|
|
|
case 'O': parser->method = HTTP_OPTIONS; break;
|
|
|
|
case 'P': parser->method = HTTP_POST;
|
|
|
|
/* or PROPFIND|PROPPATCH|PUT|PATCH|PURGE */
|
|
|
|
break;
|
2015-07-25 05:53:17 +02:00
|
|
|
case 'R': parser->method = HTTP_REPORT; /* or REBIND */ break;
|
2012-07-11 11:31:32 +02:00
|
|
|
case 'S': parser->method = HTTP_SUBSCRIBE; /* or SEARCH */ break;
|
|
|
|
case 'T': parser->method = HTTP_TRACE; break;
|
2016-01-27 12:46:59 +01:00
|
|
|
case 'U': parser->method = HTTP_UNLOCK; /* or UNSUBSCRIBE, UNBIND, UNLINK */ break;
|
2012-07-11 11:31:32 +02:00
|
|
|
default:
|
2016-05-28 09:08:20 +02:00
|
|
|
#if HTTP_PARSER_METHOD_CB
|
|
|
|
if (UNLIKELY(!TOKEN(ch))) {
|
|
|
|
SET_ERRNO(HPE_INVALID_METHOD);
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
|
|
|
|
parser->method = (unsigned int) HTTP_METHOD_UNKNOWN;
|
|
|
|
|
|
|
|
next_state = s_req_method_unknown;
|
|
|
|
#else
|
2012-07-11 11:31:32 +02:00
|
|
|
SET_ERRNO(HPE_INVALID_METHOD);
|
|
|
|
goto error;
|
2016-05-28 09:08:20 +02:00
|
|
|
#endif
|
2012-07-11 11:31:32 +02:00
|
|
|
}
|
|
|
|
|
2016-05-28 09:08:20 +02:00
|
|
|
UPDATE_STATE(next_state);
|
|
|
|
|
|
|
|
#if HTTP_PARSER_METHOD_CB
|
|
|
|
MARK(method);
|
|
|
|
#endif
|
2012-07-11 11:31:32 +02:00
|
|
|
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
case s_req_method:
|
|
|
|
{
|
|
|
|
const char *matcher;
|
2015-01-05 10:28:25 +01:00
|
|
|
if (UNLIKELY(ch == '\0')) {
|
2012-07-11 11:31:32 +02:00
|
|
|
SET_ERRNO(HPE_INVALID_METHOD);
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
|
|
|
|
matcher = method_strings[parser->method];
|
2016-05-28 09:08:20 +02:00
|
|
|
|
|
|
|
if (ch == ' ') {
|
|
|
|
if (matcher[parser->index] != '\0') {
|
|
|
|
#if HTTP_PARSER_METHOD_CB
|
|
|
|
parser->method = (unsigned int) HTTP_METHOD_UNKNOWN;
|
|
|
|
#else
|
|
|
|
SET_ERRNO(HPE_INVALID_METHOD);
|
|
|
|
goto error;
|
|
|
|
#endif
|
|
|
|
}
|
|
|
|
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_req_spaces_before_url);
|
2016-05-28 09:08:20 +02:00
|
|
|
#if HTTP_PARSER_METHOD_CB
|
|
|
|
CALLBACK_DATA(method);
|
|
|
|
#endif
|
2012-07-11 11:31:32 +02:00
|
|
|
} else if (ch == matcher[parser->index]) {
|
|
|
|
; /* nada */
|
2016-04-11 16:14:15 +02:00
|
|
|
} else if (IS_ALPHA(ch)) {
|
|
|
|
|
|
|
|
switch (parser->method << 16 | parser->index << 8 | ch) {
|
|
|
|
#define XX(meth, pos, ch, new_meth) \
|
|
|
|
case (HTTP_##meth << 16 | pos << 8 | ch): \
|
|
|
|
parser->method = HTTP_##new_meth; break;
|
|
|
|
|
|
|
|
XX(POST, 1, 'U', PUT)
|
|
|
|
XX(POST, 1, 'A', PATCH)
|
|
|
|
XX(CONNECT, 1, 'H', CHECKOUT)
|
|
|
|
XX(CONNECT, 2, 'P', COPY)
|
|
|
|
XX(MKCOL, 1, 'O', MOVE)
|
|
|
|
XX(MKCOL, 1, 'E', MERGE)
|
|
|
|
XX(MKCOL, 2, 'A', MKACTIVITY)
|
|
|
|
XX(MKCOL, 3, 'A', MKCALENDAR)
|
|
|
|
XX(SUBSCRIBE, 1, 'E', SEARCH)
|
|
|
|
XX(REPORT, 2, 'B', REBIND)
|
|
|
|
XX(POST, 1, 'R', PROPFIND)
|
|
|
|
XX(PROPFIND, 4, 'P', PROPPATCH)
|
|
|
|
XX(PUT, 2, 'R', PURGE)
|
|
|
|
XX(LOCK, 1, 'I', LINK)
|
|
|
|
XX(UNLOCK, 2, 'S', UNSUBSCRIBE)
|
|
|
|
XX(UNLOCK, 2, 'B', UNBIND)
|
|
|
|
XX(UNLOCK, 3, 'I', UNLINK)
|
|
|
|
#undef XX
|
|
|
|
|
|
|
|
default:
|
2016-05-28 09:08:20 +02:00
|
|
|
#if HTTP_PARSER_METHOD_CB
|
|
|
|
if (UNLIKELY(!TOKEN(ch))) {
|
|
|
|
SET_ERRNO(HPE_INVALID_METHOD);
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
|
|
|
|
parser->method = (unsigned int) HTTP_METHOD_UNKNOWN;
|
|
|
|
|
|
|
|
UPDATE_STATE(s_req_method_unknown);
|
|
|
|
#else
|
2013-11-14 15:53:48 +01:00
|
|
|
SET_ERRNO(HPE_INVALID_METHOD);
|
|
|
|
goto error;
|
2016-05-28 09:08:20 +02:00
|
|
|
#endif
|
2012-07-11 11:31:32 +02:00
|
|
|
}
|
2016-04-11 16:14:15 +02:00
|
|
|
} else if (ch == '-' &&
|
|
|
|
parser->index == 1 &&
|
|
|
|
parser->method == HTTP_MKCOL) {
|
|
|
|
parser->method = HTTP_MSEARCH;
|
2012-07-11 11:31:32 +02:00
|
|
|
} else {
|
2016-05-28 09:08:20 +02:00
|
|
|
#if HTTP_PARSER_METHOD_CB
|
|
|
|
if (UNLIKELY(!TOKEN(ch))) {
|
|
|
|
SET_ERRNO(HPE_INVALID_METHOD);
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
|
|
|
|
parser->method = (unsigned int) HTTP_METHOD_UNKNOWN;
|
|
|
|
|
|
|
|
UPDATE_STATE(s_req_method_unknown);
|
|
|
|
#else
|
2012-07-11 11:31:32 +02:00
|
|
|
SET_ERRNO(HPE_INVALID_METHOD);
|
|
|
|
goto error;
|
2016-05-28 09:08:20 +02:00
|
|
|
#endif
|
2012-07-11 11:31:32 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
++parser->index;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
2016-05-28 09:08:20 +02:00
|
|
|
#if HTTP_PARSER_METHOD_CB
|
|
|
|
case s_req_method_unknown:
|
|
|
|
{
|
|
|
|
if (UNLIKELY(ch == '\0')) {
|
|
|
|
SET_ERRNO(HPE_INVALID_METHOD);
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (ch == ' ') {
|
|
|
|
UPDATE_STATE(s_req_spaces_before_url);
|
|
|
|
CALLBACK_DATA(method);
|
|
|
|
} else if (UNLIKELY(!TOKEN(ch))) {
|
|
|
|
SET_ERRNO(HPE_INVALID_METHOD);
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
|
|
|
|
++parser->index;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
|
2012-07-11 11:31:32 +02:00
|
|
|
case s_req_spaces_before_url:
|
|
|
|
{
|
|
|
|
if (ch == ' ') break;
|
|
|
|
|
|
|
|
MARK(url);
|
|
|
|
if (parser->method == HTTP_CONNECT) {
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_req_server_start);
|
2012-07-11 11:31:32 +02:00
|
|
|
}
|
|
|
|
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(parse_url_char(CURRENT_STATE(), ch));
|
|
|
|
if (UNLIKELY(CURRENT_STATE() == s_dead)) {
|
2012-07-11 11:31:32 +02:00
|
|
|
SET_ERRNO(HPE_INVALID_URL);
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
case s_req_schema:
|
|
|
|
case s_req_schema_slash:
|
|
|
|
case s_req_schema_slash_slash:
|
2012-09-10 17:12:23 +02:00
|
|
|
case s_req_server_start:
|
2012-07-11 11:31:32 +02:00
|
|
|
{
|
|
|
|
switch (ch) {
|
|
|
|
/* No whitespace allowed here */
|
|
|
|
case ' ':
|
|
|
|
case CR:
|
|
|
|
case LF:
|
|
|
|
SET_ERRNO(HPE_INVALID_URL);
|
|
|
|
goto error;
|
|
|
|
default:
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(parse_url_char(CURRENT_STATE(), ch));
|
|
|
|
if (UNLIKELY(CURRENT_STATE() == s_dead)) {
|
2012-07-11 11:31:32 +02:00
|
|
|
SET_ERRNO(HPE_INVALID_URL);
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
2012-09-10 17:12:23 +02:00
|
|
|
case s_req_server:
|
|
|
|
case s_req_server_with_at:
|
2012-07-11 11:31:32 +02:00
|
|
|
case s_req_path:
|
|
|
|
case s_req_query_string_start:
|
|
|
|
case s_req_query_string:
|
|
|
|
case s_req_fragment_start:
|
|
|
|
case s_req_fragment:
|
|
|
|
{
|
|
|
|
switch (ch) {
|
|
|
|
case ' ':
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_req_http_start);
|
2012-07-11 11:31:32 +02:00
|
|
|
CALLBACK_DATA(url);
|
|
|
|
break;
|
|
|
|
case CR:
|
|
|
|
case LF:
|
|
|
|
parser->http_major = 0;
|
|
|
|
parser->http_minor = 9;
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE((ch == CR) ?
|
2012-07-11 11:31:32 +02:00
|
|
|
s_req_line_almost_done :
|
2015-01-05 10:28:25 +01:00
|
|
|
s_header_field_start);
|
2012-07-11 11:31:32 +02:00
|
|
|
CALLBACK_DATA(url);
|
|
|
|
break;
|
|
|
|
default:
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(parse_url_char(CURRENT_STATE(), ch));
|
|
|
|
if (UNLIKELY(CURRENT_STATE() == s_dead)) {
|
2012-07-11 11:31:32 +02:00
|
|
|
SET_ERRNO(HPE_INVALID_URL);
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
case s_req_http_start:
|
|
|
|
switch (ch) {
|
|
|
|
case 'H':
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_req_http_H);
|
2012-07-11 11:31:32 +02:00
|
|
|
break;
|
|
|
|
case ' ':
|
|
|
|
break;
|
|
|
|
default:
|
|
|
|
SET_ERRNO(HPE_INVALID_CONSTANT);
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
|
|
|
|
case s_req_http_H:
|
|
|
|
STRICT_CHECK(ch != 'T');
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_req_http_HT);
|
2012-07-11 11:31:32 +02:00
|
|
|
break;
|
|
|
|
|
|
|
|
case s_req_http_HT:
|
|
|
|
STRICT_CHECK(ch != 'T');
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_req_http_HTT);
|
2012-07-11 11:31:32 +02:00
|
|
|
break;
|
|
|
|
|
|
|
|
case s_req_http_HTT:
|
|
|
|
STRICT_CHECK(ch != 'P');
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_req_http_HTTP);
|
2012-07-11 11:31:32 +02:00
|
|
|
break;
|
|
|
|
|
|
|
|
case s_req_http_HTTP:
|
|
|
|
STRICT_CHECK(ch != '/');
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_req_first_http_major);
|
2012-07-11 11:31:32 +02:00
|
|
|
break;
|
|
|
|
|
|
|
|
/* first digit of major HTTP version */
|
|
|
|
case s_req_first_http_major:
|
2015-01-05 10:28:25 +01:00
|
|
|
if (UNLIKELY(ch < '1' || ch > '9')) {
|
2012-07-11 11:31:32 +02:00
|
|
|
SET_ERRNO(HPE_INVALID_VERSION);
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
|
|
|
|
parser->http_major = ch - '0';
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_req_http_major);
|
2012-07-11 11:31:32 +02:00
|
|
|
break;
|
|
|
|
|
|
|
|
/* major HTTP version or dot */
|
|
|
|
case s_req_http_major:
|
|
|
|
{
|
|
|
|
if (ch == '.') {
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_req_first_http_minor);
|
2012-07-11 11:31:32 +02:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
2015-01-05 10:28:25 +01:00
|
|
|
if (UNLIKELY(!IS_NUM(ch))) {
|
2012-07-11 11:31:32 +02:00
|
|
|
SET_ERRNO(HPE_INVALID_VERSION);
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
|
|
|
|
parser->http_major *= 10;
|
|
|
|
parser->http_major += ch - '0';
|
|
|
|
|
2015-01-05 10:28:25 +01:00
|
|
|
if (UNLIKELY(parser->http_major > 999)) {
|
2012-07-11 11:31:32 +02:00
|
|
|
SET_ERRNO(HPE_INVALID_VERSION);
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* first digit of minor HTTP version */
|
|
|
|
case s_req_first_http_minor:
|
2015-01-05 10:28:25 +01:00
|
|
|
if (UNLIKELY(!IS_NUM(ch))) {
|
2012-07-11 11:31:32 +02:00
|
|
|
SET_ERRNO(HPE_INVALID_VERSION);
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
|
|
|
|
parser->http_minor = ch - '0';
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_req_http_minor);
|
2012-07-11 11:31:32 +02:00
|
|
|
break;
|
|
|
|
|
|
|
|
/* minor HTTP version or end of request line */
|
|
|
|
case s_req_http_minor:
|
|
|
|
{
|
|
|
|
if (ch == CR) {
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_req_line_almost_done);
|
2012-07-11 11:31:32 +02:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (ch == LF) {
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_header_field_start);
|
2012-07-11 11:31:32 +02:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* XXX allow spaces after digit? */
|
|
|
|
|
2015-01-05 10:28:25 +01:00
|
|
|
if (UNLIKELY(!IS_NUM(ch))) {
|
2012-07-11 11:31:32 +02:00
|
|
|
SET_ERRNO(HPE_INVALID_VERSION);
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
|
|
|
|
parser->http_minor *= 10;
|
|
|
|
parser->http_minor += ch - '0';
|
|
|
|
|
2015-01-05 10:28:25 +01:00
|
|
|
if (UNLIKELY(parser->http_minor > 999)) {
|
2012-07-11 11:31:32 +02:00
|
|
|
SET_ERRNO(HPE_INVALID_VERSION);
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* end of request line */
|
|
|
|
case s_req_line_almost_done:
|
|
|
|
{
|
2015-01-05 10:28:25 +01:00
|
|
|
if (UNLIKELY(ch != LF)) {
|
2012-07-11 11:31:32 +02:00
|
|
|
SET_ERRNO(HPE_LF_EXPECTED);
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_header_field_start);
|
2012-07-11 11:31:32 +02:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
case s_header_field_start:
|
|
|
|
{
|
|
|
|
if (ch == CR) {
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_headers_almost_done);
|
2012-07-11 11:31:32 +02:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (ch == LF) {
|
|
|
|
/* they might be just sending \n instead of \r\n so this would be
|
|
|
|
* the second \n to denote the end of headers*/
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_headers_almost_done);
|
|
|
|
REEXECUTE();
|
2012-07-11 11:31:32 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
c = TOKEN(ch);
|
|
|
|
|
2015-01-05 10:28:25 +01:00
|
|
|
if (UNLIKELY(!c)) {
|
2012-07-11 11:31:32 +02:00
|
|
|
SET_ERRNO(HPE_INVALID_HEADER_TOKEN);
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
|
|
|
|
MARK(header_field);
|
|
|
|
|
|
|
|
parser->index = 0;
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_header_field);
|
2012-07-11 11:31:32 +02:00
|
|
|
|
|
|
|
switch (c) {
|
|
|
|
case 'c':
|
|
|
|
parser->header_state = h_C;
|
|
|
|
break;
|
|
|
|
|
|
|
|
case 'p':
|
|
|
|
parser->header_state = h_matching_proxy_connection;
|
|
|
|
break;
|
|
|
|
|
|
|
|
case 't':
|
|
|
|
parser->header_state = h_matching_transfer_encoding;
|
|
|
|
break;
|
|
|
|
|
|
|
|
case 'u':
|
|
|
|
parser->header_state = h_matching_upgrade;
|
|
|
|
break;
|
|
|
|
|
|
|
|
default:
|
|
|
|
parser->header_state = h_general;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
case s_header_field:
|
|
|
|
{
|
2015-01-05 10:28:25 +01:00
|
|
|
const char* start = p;
|
|
|
|
for (; p != data + len; p++) {
|
|
|
|
ch = *p;
|
|
|
|
c = TOKEN(ch);
|
|
|
|
|
|
|
|
if (!c)
|
|
|
|
break;
|
2012-07-11 11:31:32 +02:00
|
|
|
|
|
|
|
switch (parser->header_state) {
|
|
|
|
case h_general:
|
|
|
|
break;
|
|
|
|
|
|
|
|
case h_C:
|
|
|
|
parser->index++;
|
|
|
|
parser->header_state = (c == 'o' ? h_CO : h_general);
|
|
|
|
break;
|
|
|
|
|
|
|
|
case h_CO:
|
|
|
|
parser->index++;
|
|
|
|
parser->header_state = (c == 'n' ? h_CON : h_general);
|
|
|
|
break;
|
|
|
|
|
|
|
|
case h_CON:
|
|
|
|
parser->index++;
|
|
|
|
switch (c) {
|
|
|
|
case 'n':
|
|
|
|
parser->header_state = h_matching_connection;
|
|
|
|
break;
|
|
|
|
case 't':
|
|
|
|
parser->header_state = h_matching_content_length;
|
|
|
|
break;
|
|
|
|
default:
|
|
|
|
parser->header_state = h_general;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
|
|
|
|
/* connection */
|
|
|
|
|
|
|
|
case h_matching_connection:
|
|
|
|
parser->index++;
|
|
|
|
if (parser->index > sizeof(CONNECTION)-1
|
|
|
|
|| c != CONNECTION[parser->index]) {
|
|
|
|
parser->header_state = h_general;
|
|
|
|
} else if (parser->index == sizeof(CONNECTION)-2) {
|
|
|
|
parser->header_state = h_connection;
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
|
|
|
|
/* proxy-connection */
|
|
|
|
|
|
|
|
case h_matching_proxy_connection:
|
|
|
|
parser->index++;
|
|
|
|
if (parser->index > sizeof(PROXY_CONNECTION)-1
|
|
|
|
|| c != PROXY_CONNECTION[parser->index]) {
|
|
|
|
parser->header_state = h_general;
|
|
|
|
} else if (parser->index == sizeof(PROXY_CONNECTION)-2) {
|
|
|
|
parser->header_state = h_connection;
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
|
|
|
|
/* content-length */
|
|
|
|
|
|
|
|
case h_matching_content_length:
|
|
|
|
parser->index++;
|
|
|
|
if (parser->index > sizeof(CONTENT_LENGTH)-1
|
|
|
|
|| c != CONTENT_LENGTH[parser->index]) {
|
|
|
|
parser->header_state = h_general;
|
|
|
|
} else if (parser->index == sizeof(CONTENT_LENGTH)-2) {
|
2016-04-11 16:14:15 +02:00
|
|
|
if (parser->flags & F_CONTENTLENGTH) {
|
|
|
|
SET_ERRNO(HPE_UNEXPECTED_CONTENT_LENGTH);
|
|
|
|
goto error;
|
|
|
|
}
|
2012-07-11 11:31:32 +02:00
|
|
|
parser->header_state = h_content_length;
|
2016-04-11 16:14:15 +02:00
|
|
|
parser->flags |= F_CONTENTLENGTH;
|
2012-07-11 11:31:32 +02:00
|
|
|
}
|
|
|
|
break;
|
|
|
|
|
|
|
|
/* transfer-encoding */
|
|
|
|
|
|
|
|
case h_matching_transfer_encoding:
|
|
|
|
parser->index++;
|
|
|
|
if (parser->index > sizeof(TRANSFER_ENCODING)-1
|
|
|
|
|| c != TRANSFER_ENCODING[parser->index]) {
|
|
|
|
parser->header_state = h_general;
|
|
|
|
} else if (parser->index == sizeof(TRANSFER_ENCODING)-2) {
|
|
|
|
parser->header_state = h_transfer_encoding;
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
|
|
|
|
/* upgrade */
|
|
|
|
|
|
|
|
case h_matching_upgrade:
|
|
|
|
parser->index++;
|
|
|
|
if (parser->index > sizeof(UPGRADE)-1
|
|
|
|
|| c != UPGRADE[parser->index]) {
|
|
|
|
parser->header_state = h_general;
|
|
|
|
} else if (parser->index == sizeof(UPGRADE)-2) {
|
|
|
|
parser->header_state = h_upgrade;
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
|
|
|
|
case h_connection:
|
|
|
|
case h_content_length:
|
|
|
|
case h_transfer_encoding:
|
|
|
|
case h_upgrade:
|
|
|
|
if (ch != ' ') parser->header_state = h_general;
|
|
|
|
break;
|
|
|
|
|
|
|
|
default:
|
|
|
|
assert(0 && "Unknown header_state");
|
|
|
|
break;
|
|
|
|
}
|
2015-01-05 10:28:25 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
COUNT_HEADER_SIZE(p - start);
|
|
|
|
|
|
|
|
if (p == data + len) {
|
|
|
|
--p;
|
2012-07-11 11:31:32 +02:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (ch == ':') {
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_header_value_discard_ws);
|
2012-07-11 11:31:32 +02:00
|
|
|
CALLBACK_DATA(header_field);
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
SET_ERRNO(HPE_INVALID_HEADER_TOKEN);
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
|
2014-05-15 17:23:03 +02:00
|
|
|
case s_header_value_discard_ws:
|
2012-07-11 11:31:32 +02:00
|
|
|
if (ch == ' ' || ch == '\t') break;
|
|
|
|
|
|
|
|
if (ch == CR) {
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_header_value_discard_ws_almost_done);
|
2012-07-11 11:31:32 +02:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (ch == LF) {
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_header_value_discard_lws);
|
2012-07-11 11:31:32 +02:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
2014-05-15 17:23:03 +02:00
|
|
|
/* FALLTHROUGH */
|
|
|
|
|
|
|
|
case s_header_value_start:
|
|
|
|
{
|
|
|
|
MARK(header_value);
|
|
|
|
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_header_value);
|
2014-05-15 17:23:03 +02:00
|
|
|
parser->index = 0;
|
|
|
|
|
2012-07-11 11:31:32 +02:00
|
|
|
c = LOWER(ch);
|
|
|
|
|
|
|
|
switch (parser->header_state) {
|
|
|
|
case h_upgrade:
|
|
|
|
parser->flags |= F_UPGRADE;
|
|
|
|
parser->header_state = h_general;
|
|
|
|
break;
|
|
|
|
|
|
|
|
case h_transfer_encoding:
|
|
|
|
/* looking for 'Transfer-Encoding: chunked' */
|
|
|
|
if ('c' == c) {
|
|
|
|
parser->header_state = h_matching_transfer_encoding_chunked;
|
|
|
|
} else {
|
|
|
|
parser->header_state = h_general;
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
|
|
|
|
case h_content_length:
|
2015-01-05 10:28:25 +01:00
|
|
|
if (UNLIKELY(!IS_NUM(ch))) {
|
2012-07-11 11:31:32 +02:00
|
|
|
SET_ERRNO(HPE_INVALID_CONTENT_LENGTH);
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
|
|
|
|
parser->content_length = ch - '0';
|
|
|
|
break;
|
|
|
|
|
|
|
|
case h_connection:
|
|
|
|
/* looking for 'Connection: keep-alive' */
|
|
|
|
if (c == 'k') {
|
|
|
|
parser->header_state = h_matching_connection_keep_alive;
|
|
|
|
/* looking for 'Connection: close' */
|
|
|
|
} else if (c == 'c') {
|
|
|
|
parser->header_state = h_matching_connection_close;
|
2015-01-05 10:28:25 +01:00
|
|
|
} else if (c == 'u') {
|
|
|
|
parser->header_state = h_matching_connection_upgrade;
|
2012-07-11 11:31:32 +02:00
|
|
|
} else {
|
2015-01-05 10:28:25 +01:00
|
|
|
parser->header_state = h_matching_connection_token;
|
2012-07-11 11:31:32 +02:00
|
|
|
}
|
|
|
|
break;
|
|
|
|
|
2015-01-05 10:28:25 +01:00
|
|
|
/* Multi-value `Connection` header */
|
|
|
|
case h_matching_connection_token_start:
|
|
|
|
break;
|
|
|
|
|
2012-07-11 11:31:32 +02:00
|
|
|
default:
|
|
|
|
parser->header_state = h_general;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
case s_header_value:
|
|
|
|
{
|
2015-01-05 10:28:25 +01:00
|
|
|
const char* start = p;
|
2015-02-20 11:51:41 +01:00
|
|
|
enum header_states h_state = (enum header_states) parser->header_state;
|
2015-01-05 10:28:25 +01:00
|
|
|
for (; p != data + len; p++) {
|
|
|
|
ch = *p;
|
|
|
|
if (ch == CR) {
|
|
|
|
UPDATE_STATE(s_header_almost_done);
|
|
|
|
parser->header_state = h_state;
|
|
|
|
CALLBACK_DATA(header_value);
|
|
|
|
break;
|
|
|
|
}
|
2012-07-11 11:31:32 +02:00
|
|
|
|
2015-01-05 10:28:25 +01:00
|
|
|
if (ch == LF) {
|
|
|
|
UPDATE_STATE(s_header_almost_done);
|
|
|
|
COUNT_HEADER_SIZE(p - start);
|
|
|
|
parser->header_state = h_state;
|
|
|
|
CALLBACK_DATA_NOADVANCE(header_value);
|
|
|
|
REEXECUTE();
|
|
|
|
}
|
2012-07-11 11:31:32 +02:00
|
|
|
|
2016-04-11 16:14:15 +02:00
|
|
|
if (!lenient && !IS_HEADER_CHAR(ch)) {
|
|
|
|
SET_ERRNO(HPE_INVALID_HEADER_TOKEN);
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
|
2015-01-05 10:28:25 +01:00
|
|
|
c = LOWER(ch);
|
2012-07-11 11:31:32 +02:00
|
|
|
|
2015-01-05 10:28:25 +01:00
|
|
|
switch (h_state) {
|
|
|
|
case h_general:
|
|
|
|
{
|
|
|
|
const char* p_cr;
|
|
|
|
const char* p_lf;
|
|
|
|
size_t limit = data + len - p;
|
|
|
|
|
|
|
|
limit = MIN(limit, HTTP_MAX_HEADER_SIZE);
|
|
|
|
|
2015-02-20 11:51:41 +01:00
|
|
|
p_cr = (const char*) memchr(p, CR, limit);
|
|
|
|
p_lf = (const char*) memchr(p, LF, limit);
|
2015-01-05 10:28:25 +01:00
|
|
|
if (p_cr != NULL) {
|
|
|
|
if (p_lf != NULL && p_cr >= p_lf)
|
|
|
|
p = p_lf;
|
|
|
|
else
|
|
|
|
p = p_cr;
|
|
|
|
} else if (UNLIKELY(p_lf != NULL)) {
|
|
|
|
p = p_lf;
|
|
|
|
} else {
|
|
|
|
p = data + len;
|
|
|
|
}
|
|
|
|
--p;
|
2012-07-11 11:31:32 +02:00
|
|
|
|
2015-01-05 10:28:25 +01:00
|
|
|
break;
|
|
|
|
}
|
2012-07-11 11:31:32 +02:00
|
|
|
|
2015-01-05 10:28:25 +01:00
|
|
|
case h_connection:
|
|
|
|
case h_transfer_encoding:
|
|
|
|
assert(0 && "Shouldn't get here.");
|
|
|
|
break;
|
2012-07-11 11:31:32 +02:00
|
|
|
|
2015-01-05 10:28:25 +01:00
|
|
|
case h_content_length:
|
|
|
|
{
|
|
|
|
uint64_t t;
|
2012-07-11 11:31:32 +02:00
|
|
|
|
2015-01-05 10:28:25 +01:00
|
|
|
if (ch == ' ') break;
|
2012-07-11 11:31:32 +02:00
|
|
|
|
2015-01-05 10:28:25 +01:00
|
|
|
if (UNLIKELY(!IS_NUM(ch))) {
|
|
|
|
SET_ERRNO(HPE_INVALID_CONTENT_LENGTH);
|
|
|
|
parser->header_state = h_state;
|
|
|
|
goto error;
|
|
|
|
}
|
2012-07-11 11:31:32 +02:00
|
|
|
|
2015-01-05 10:28:25 +01:00
|
|
|
t = parser->content_length;
|
|
|
|
t *= 10;
|
|
|
|
t += ch - '0';
|
2012-07-11 11:31:32 +02:00
|
|
|
|
2015-01-05 10:28:25 +01:00
|
|
|
/* Overflow? Test against a conservative limit for simplicity. */
|
|
|
|
if (UNLIKELY((ULLONG_MAX - 10) / 10 < parser->content_length)) {
|
|
|
|
SET_ERRNO(HPE_INVALID_CONTENT_LENGTH);
|
|
|
|
parser->header_state = h_state;
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
|
|
|
|
parser->content_length = t;
|
|
|
|
break;
|
2012-07-11 11:31:32 +02:00
|
|
|
}
|
|
|
|
|
2015-01-05 10:28:25 +01:00
|
|
|
/* Transfer-Encoding: chunked */
|
|
|
|
case h_matching_transfer_encoding_chunked:
|
|
|
|
parser->index++;
|
|
|
|
if (parser->index > sizeof(CHUNKED)-1
|
|
|
|
|| c != CHUNKED[parser->index]) {
|
|
|
|
h_state = h_general;
|
|
|
|
} else if (parser->index == sizeof(CHUNKED)-2) {
|
|
|
|
h_state = h_transfer_encoding_chunked;
|
|
|
|
}
|
|
|
|
break;
|
2012-07-11 11:31:32 +02:00
|
|
|
|
2015-01-05 10:28:25 +01:00
|
|
|
case h_matching_connection_token_start:
|
|
|
|
/* looking for 'Connection: keep-alive' */
|
|
|
|
if (c == 'k') {
|
|
|
|
h_state = h_matching_connection_keep_alive;
|
|
|
|
/* looking for 'Connection: close' */
|
|
|
|
} else if (c == 'c') {
|
|
|
|
h_state = h_matching_connection_close;
|
|
|
|
} else if (c == 'u') {
|
|
|
|
h_state = h_matching_connection_upgrade;
|
|
|
|
} else if (STRICT_TOKEN(c)) {
|
|
|
|
h_state = h_matching_connection_token;
|
2015-02-20 11:51:41 +01:00
|
|
|
} else if (c == ' ' || c == '\t') {
|
|
|
|
/* Skip lws */
|
2015-01-05 10:28:25 +01:00
|
|
|
} else {
|
|
|
|
h_state = h_general;
|
|
|
|
}
|
|
|
|
break;
|
2012-07-11 11:31:32 +02:00
|
|
|
|
2015-01-05 10:28:25 +01:00
|
|
|
/* looking for 'Connection: keep-alive' */
|
|
|
|
case h_matching_connection_keep_alive:
|
|
|
|
parser->index++;
|
|
|
|
if (parser->index > sizeof(KEEP_ALIVE)-1
|
|
|
|
|| c != KEEP_ALIVE[parser->index]) {
|
|
|
|
h_state = h_matching_connection_token;
|
|
|
|
} else if (parser->index == sizeof(KEEP_ALIVE)-2) {
|
|
|
|
h_state = h_connection_keep_alive;
|
|
|
|
}
|
|
|
|
break;
|
2012-07-11 11:31:32 +02:00
|
|
|
|
2015-01-05 10:28:25 +01:00
|
|
|
/* looking for 'Connection: close' */
|
|
|
|
case h_matching_connection_close:
|
|
|
|
parser->index++;
|
|
|
|
if (parser->index > sizeof(CLOSE)-1 || c != CLOSE[parser->index]) {
|
|
|
|
h_state = h_matching_connection_token;
|
|
|
|
} else if (parser->index == sizeof(CLOSE)-2) {
|
|
|
|
h_state = h_connection_close;
|
|
|
|
}
|
|
|
|
break;
|
2012-07-11 11:31:32 +02:00
|
|
|
|
2015-01-05 10:28:25 +01:00
|
|
|
/* looking for 'Connection: upgrade' */
|
|
|
|
case h_matching_connection_upgrade:
|
|
|
|
parser->index++;
|
|
|
|
if (parser->index > sizeof(UPGRADE) - 1 ||
|
|
|
|
c != UPGRADE[parser->index]) {
|
|
|
|
h_state = h_matching_connection_token;
|
|
|
|
} else if (parser->index == sizeof(UPGRADE)-2) {
|
|
|
|
h_state = h_connection_upgrade;
|
|
|
|
}
|
|
|
|
break;
|
2012-07-11 11:31:32 +02:00
|
|
|
|
2015-01-05 10:28:25 +01:00
|
|
|
case h_matching_connection_token:
|
|
|
|
if (ch == ',') {
|
|
|
|
h_state = h_matching_connection_token_start;
|
|
|
|
parser->index = 0;
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
|
|
|
|
case h_transfer_encoding_chunked:
|
|
|
|
if (ch != ' ') h_state = h_general;
|
|
|
|
break;
|
|
|
|
|
|
|
|
case h_connection_keep_alive:
|
|
|
|
case h_connection_close:
|
|
|
|
case h_connection_upgrade:
|
|
|
|
if (ch == ',') {
|
|
|
|
if (h_state == h_connection_keep_alive) {
|
|
|
|
parser->flags |= F_CONNECTION_KEEP_ALIVE;
|
|
|
|
} else if (h_state == h_connection_close) {
|
|
|
|
parser->flags |= F_CONNECTION_CLOSE;
|
|
|
|
} else if (h_state == h_connection_upgrade) {
|
|
|
|
parser->flags |= F_CONNECTION_UPGRADE;
|
|
|
|
}
|
|
|
|
h_state = h_matching_connection_token_start;
|
|
|
|
parser->index = 0;
|
|
|
|
} else if (ch != ' ') {
|
|
|
|
h_state = h_matching_connection_token;
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
|
|
|
|
default:
|
|
|
|
UPDATE_STATE(s_header_value);
|
|
|
|
h_state = h_general;
|
|
|
|
break;
|
|
|
|
}
|
2012-07-11 11:31:32 +02:00
|
|
|
}
|
2015-01-05 10:28:25 +01:00
|
|
|
parser->header_state = h_state;
|
|
|
|
|
|
|
|
COUNT_HEADER_SIZE(p - start);
|
|
|
|
|
|
|
|
if (p == data + len)
|
|
|
|
--p;
|
2012-07-11 11:31:32 +02:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
case s_header_almost_done:
|
|
|
|
{
|
2016-04-11 16:14:15 +02:00
|
|
|
if (UNLIKELY(ch != LF)) {
|
|
|
|
SET_ERRNO(HPE_LF_EXPECTED);
|
|
|
|
goto error;
|
|
|
|
}
|
2012-07-11 11:31:32 +02:00
|
|
|
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_header_value_lws);
|
2014-05-15 17:23:03 +02:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
case s_header_value_lws:
|
|
|
|
{
|
|
|
|
if (ch == ' ' || ch == '\t') {
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_header_value_start);
|
|
|
|
REEXECUTE();
|
2014-05-15 17:23:03 +02:00
|
|
|
}
|
2012-07-11 11:31:32 +02:00
|
|
|
|
2014-05-15 17:23:03 +02:00
|
|
|
/* finished the header */
|
2012-07-11 11:31:32 +02:00
|
|
|
switch (parser->header_state) {
|
|
|
|
case h_connection_keep_alive:
|
|
|
|
parser->flags |= F_CONNECTION_KEEP_ALIVE;
|
|
|
|
break;
|
|
|
|
case h_connection_close:
|
|
|
|
parser->flags |= F_CONNECTION_CLOSE;
|
|
|
|
break;
|
|
|
|
case h_transfer_encoding_chunked:
|
|
|
|
parser->flags |= F_CHUNKED;
|
|
|
|
break;
|
2015-01-05 10:28:25 +01:00
|
|
|
case h_connection_upgrade:
|
|
|
|
parser->flags |= F_CONNECTION_UPGRADE;
|
|
|
|
break;
|
2012-07-11 11:31:32 +02:00
|
|
|
default:
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_header_field_start);
|
|
|
|
REEXECUTE();
|
2014-05-15 17:23:03 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
case s_header_value_discard_ws_almost_done:
|
|
|
|
{
|
|
|
|
STRICT_CHECK(ch != LF);
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_header_value_discard_lws);
|
2012-07-11 11:31:32 +02:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
2014-05-15 17:23:03 +02:00
|
|
|
case s_header_value_discard_lws:
|
2012-07-11 11:31:32 +02:00
|
|
|
{
|
2014-05-15 17:23:03 +02:00
|
|
|
if (ch == ' ' || ch == '\t') {
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_header_value_discard_ws);
|
2014-05-15 17:23:03 +02:00
|
|
|
break;
|
|
|
|
} else {
|
2015-01-05 10:28:25 +01:00
|
|
|
switch (parser->header_state) {
|
|
|
|
case h_connection_keep_alive:
|
|
|
|
parser->flags |= F_CONNECTION_KEEP_ALIVE;
|
|
|
|
break;
|
|
|
|
case h_connection_close:
|
|
|
|
parser->flags |= F_CONNECTION_CLOSE;
|
|
|
|
break;
|
|
|
|
case h_connection_upgrade:
|
|
|
|
parser->flags |= F_CONNECTION_UPGRADE;
|
|
|
|
break;
|
|
|
|
case h_transfer_encoding_chunked:
|
|
|
|
parser->flags |= F_CHUNKED;
|
|
|
|
break;
|
|
|
|
default:
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
2014-05-15 17:23:03 +02:00
|
|
|
/* header value was empty */
|
|
|
|
MARK(header_value);
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_header_field_start);
|
2014-05-15 17:23:03 +02:00
|
|
|
CALLBACK_DATA_NOADVANCE(header_value);
|
2015-01-05 10:28:25 +01:00
|
|
|
REEXECUTE();
|
2012-07-11 11:31:32 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
case s_headers_almost_done:
|
|
|
|
{
|
|
|
|
STRICT_CHECK(ch != LF);
|
|
|
|
|
|
|
|
if (parser->flags & F_TRAILING) {
|
|
|
|
/* End of a chunked request */
|
2015-05-21 18:55:47 +02:00
|
|
|
UPDATE_STATE(s_message_done);
|
|
|
|
CALLBACK_NOTIFY_NOADVANCE(chunk_complete);
|
|
|
|
REEXECUTE();
|
2012-07-11 11:31:32 +02:00
|
|
|
}
|
|
|
|
|
2016-04-11 16:14:15 +02:00
|
|
|
/* Cannot use chunked encoding and a content-length header together
|
|
|
|
per the HTTP specification. */
|
|
|
|
if ((parser->flags & F_CHUNKED) &&
|
|
|
|
(parser->flags & F_CONTENTLENGTH)) {
|
|
|
|
SET_ERRNO(HPE_UNEXPECTED_CONTENT_LENGTH);
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_headers_done);
|
2012-07-11 11:31:32 +02:00
|
|
|
|
|
|
|
/* Set this here so that on_headers_complete() callbacks can see it */
|
|
|
|
parser->upgrade =
|
2015-01-05 10:28:25 +01:00
|
|
|
((parser->flags & (F_UPGRADE | F_CONNECTION_UPGRADE)) ==
|
|
|
|
(F_UPGRADE | F_CONNECTION_UPGRADE) ||
|
|
|
|
parser->method == HTTP_CONNECT);
|
2012-07-11 11:31:32 +02:00
|
|
|
|
|
|
|
/* Here we call the headers_complete callback. This is somewhat
|
|
|
|
* different than other callbacks because if the user returns 1, we
|
|
|
|
* will interpret that as saying that this message has no body. This
|
|
|
|
* is needed for the annoying case of recieving a response to a HEAD
|
|
|
|
* request.
|
|
|
|
*
|
|
|
|
* We'd like to use CALLBACK_NOTIFY_NOADVANCE() here but we cannot, so
|
|
|
|
* we have to simulate it by handling a change in errno below.
|
|
|
|
*/
|
|
|
|
if (settings->on_headers_complete) {
|
|
|
|
switch (settings->on_headers_complete(parser)) {
|
|
|
|
case 0:
|
|
|
|
break;
|
|
|
|
|
2016-05-28 05:17:17 +02:00
|
|
|
case 2:
|
|
|
|
parser->upgrade = 1;
|
|
|
|
|
2012-07-11 11:31:32 +02:00
|
|
|
case 1:
|
|
|
|
parser->flags |= F_SKIPBODY;
|
|
|
|
break;
|
|
|
|
|
|
|
|
default:
|
|
|
|
SET_ERRNO(HPE_CB_headers_complete);
|
2015-01-05 10:28:25 +01:00
|
|
|
RETURN(p - data); /* Error */
|
2012-07-11 11:31:32 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
if (HTTP_PARSER_ERRNO(parser) != HPE_OK) {
|
2015-01-05 10:28:25 +01:00
|
|
|
RETURN(p - data);
|
2012-07-11 11:31:32 +02:00
|
|
|
}
|
|
|
|
|
2015-01-05 10:28:25 +01:00
|
|
|
REEXECUTE();
|
2012-07-11 11:31:32 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
case s_headers_done:
|
|
|
|
{
|
2015-07-25 05:53:17 +02:00
|
|
|
int hasBody;
|
2012-07-11 11:31:32 +02:00
|
|
|
STRICT_CHECK(ch != LF);
|
|
|
|
|
|
|
|
parser->nread = 0;
|
|
|
|
|
2015-07-25 05:53:17 +02:00
|
|
|
hasBody = parser->flags & F_CHUNKED ||
|
2015-05-21 18:55:47 +02:00
|
|
|
(parser->content_length > 0 && parser->content_length != ULLONG_MAX);
|
|
|
|
if (parser->upgrade && (parser->method == HTTP_CONNECT ||
|
|
|
|
(parser->flags & F_SKIPBODY) || !hasBody)) {
|
|
|
|
/* Exit, the rest of the message is in a different protocol. */
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(NEW_MESSAGE());
|
2013-07-31 14:48:37 +02:00
|
|
|
CALLBACK_NOTIFY(message_complete);
|
2015-01-05 10:28:25 +01:00
|
|
|
RETURN((p - data) + 1);
|
2012-07-11 11:31:32 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
if (parser->flags & F_SKIPBODY) {
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(NEW_MESSAGE());
|
2012-07-11 11:31:32 +02:00
|
|
|
CALLBACK_NOTIFY(message_complete);
|
|
|
|
} else if (parser->flags & F_CHUNKED) {
|
|
|
|
/* chunked encoding - ignore Content-Length header */
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_chunk_size_start);
|
2012-07-11 11:31:32 +02:00
|
|
|
} else {
|
|
|
|
if (parser->content_length == 0) {
|
|
|
|
/* Content-Length header given but zero: Content-Length: 0\r\n */
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(NEW_MESSAGE());
|
2012-07-11 11:31:32 +02:00
|
|
|
CALLBACK_NOTIFY(message_complete);
|
|
|
|
} else if (parser->content_length != ULLONG_MAX) {
|
|
|
|
/* Content-Length header given and non-zero */
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_body_identity);
|
2012-07-11 11:31:32 +02:00
|
|
|
} else {
|
2015-05-21 18:55:47 +02:00
|
|
|
if (!http_message_needs_eof(parser)) {
|
2012-07-11 11:31:32 +02:00
|
|
|
/* Assume content-length 0 - read the next */
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(NEW_MESSAGE());
|
2012-07-11 11:31:32 +02:00
|
|
|
CALLBACK_NOTIFY(message_complete);
|
|
|
|
} else {
|
|
|
|
/* Read body until EOF */
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_body_identity_eof);
|
2012-07-11 11:31:32 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
case s_body_identity:
|
|
|
|
{
|
|
|
|
uint64_t to_read = MIN(parser->content_length,
|
|
|
|
(uint64_t) ((data + len) - p));
|
|
|
|
|
|
|
|
assert(parser->content_length != 0
|
|
|
|
&& parser->content_length != ULLONG_MAX);
|
|
|
|
|
|
|
|
/* The difference between advancing content_length and p is because
|
|
|
|
* the latter will automaticaly advance on the next loop iteration.
|
|
|
|
* Further, if content_length ends up at 0, we want to see the last
|
|
|
|
* byte again for our message complete callback.
|
|
|
|
*/
|
|
|
|
MARK(body);
|
|
|
|
parser->content_length -= to_read;
|
|
|
|
p += to_read - 1;
|
|
|
|
|
|
|
|
if (parser->content_length == 0) {
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_message_done);
|
2012-07-11 11:31:32 +02:00
|
|
|
|
|
|
|
/* Mimic CALLBACK_DATA_NOADVANCE() but with one extra byte.
|
|
|
|
*
|
|
|
|
* The alternative to doing this is to wait for the next byte to
|
|
|
|
* trigger the data callback, just as in every other case. The
|
|
|
|
* problem with this is that this makes it difficult for the test
|
|
|
|
* harness to distinguish between complete-on-EOF and
|
|
|
|
* complete-on-length. It's not clear that this distinction is
|
|
|
|
* important for applications, but let's keep it for now.
|
|
|
|
*/
|
|
|
|
CALLBACK_DATA_(body, p - body_mark + 1, p - data);
|
2015-01-05 10:28:25 +01:00
|
|
|
REEXECUTE();
|
2012-07-11 11:31:32 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* read until EOF */
|
|
|
|
case s_body_identity_eof:
|
|
|
|
MARK(body);
|
|
|
|
p = data + len - 1;
|
|
|
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
case s_message_done:
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(NEW_MESSAGE());
|
2012-07-11 11:31:32 +02:00
|
|
|
CALLBACK_NOTIFY(message_complete);
|
2015-05-21 18:55:47 +02:00
|
|
|
if (parser->upgrade) {
|
|
|
|
/* Exit, the rest of the message is in a different protocol. */
|
|
|
|
RETURN((p - data) + 1);
|
|
|
|
}
|
2012-07-11 11:31:32 +02:00
|
|
|
break;
|
|
|
|
|
|
|
|
case s_chunk_size_start:
|
|
|
|
{
|
|
|
|
assert(parser->nread == 1);
|
|
|
|
assert(parser->flags & F_CHUNKED);
|
|
|
|
|
|
|
|
unhex_val = unhex[(unsigned char)ch];
|
2015-01-05 10:28:25 +01:00
|
|
|
if (UNLIKELY(unhex_val == -1)) {
|
2012-07-11 11:31:32 +02:00
|
|
|
SET_ERRNO(HPE_INVALID_CHUNK_SIZE);
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
|
|
|
|
parser->content_length = unhex_val;
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_chunk_size);
|
2012-07-11 11:31:32 +02:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
case s_chunk_size:
|
|
|
|
{
|
|
|
|
uint64_t t;
|
|
|
|
|
|
|
|
assert(parser->flags & F_CHUNKED);
|
|
|
|
|
|
|
|
if (ch == CR) {
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_chunk_size_almost_done);
|
2012-07-11 11:31:32 +02:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
unhex_val = unhex[(unsigned char)ch];
|
|
|
|
|
|
|
|
if (unhex_val == -1) {
|
|
|
|
if (ch == ';' || ch == ' ') {
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_chunk_parameters);
|
2012-07-11 11:31:32 +02:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
SET_ERRNO(HPE_INVALID_CHUNK_SIZE);
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
|
|
|
|
t = parser->content_length;
|
|
|
|
t *= 16;
|
|
|
|
t += unhex_val;
|
|
|
|
|
2014-05-15 17:23:03 +02:00
|
|
|
/* Overflow? Test against a conservative limit for simplicity. */
|
2015-01-05 10:28:25 +01:00
|
|
|
if (UNLIKELY((ULLONG_MAX - 16) / 16 < parser->content_length)) {
|
2012-07-11 11:31:32 +02:00
|
|
|
SET_ERRNO(HPE_INVALID_CONTENT_LENGTH);
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
|
|
|
|
parser->content_length = t;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
case s_chunk_parameters:
|
|
|
|
{
|
|
|
|
assert(parser->flags & F_CHUNKED);
|
|
|
|
/* just ignore this shit. TODO check for overflow */
|
|
|
|
if (ch == CR) {
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_chunk_size_almost_done);
|
2012-07-11 11:31:32 +02:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
case s_chunk_size_almost_done:
|
|
|
|
{
|
|
|
|
assert(parser->flags & F_CHUNKED);
|
|
|
|
STRICT_CHECK(ch != LF);
|
|
|
|
|
|
|
|
parser->nread = 0;
|
|
|
|
|
|
|
|
if (parser->content_length == 0) {
|
|
|
|
parser->flags |= F_TRAILING;
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_header_field_start);
|
2012-07-11 11:31:32 +02:00
|
|
|
} else {
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_chunk_data);
|
2012-07-11 11:31:32 +02:00
|
|
|
}
|
2015-05-21 18:55:47 +02:00
|
|
|
CALLBACK_NOTIFY(chunk_header);
|
2012-07-11 11:31:32 +02:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
case s_chunk_data:
|
|
|
|
{
|
|
|
|
uint64_t to_read = MIN(parser->content_length,
|
|
|
|
(uint64_t) ((data + len) - p));
|
|
|
|
|
|
|
|
assert(parser->flags & F_CHUNKED);
|
|
|
|
assert(parser->content_length != 0
|
|
|
|
&& parser->content_length != ULLONG_MAX);
|
|
|
|
|
|
|
|
/* See the explanation in s_body_identity for why the content
|
|
|
|
* length and data pointers are managed this way.
|
|
|
|
*/
|
|
|
|
MARK(body);
|
|
|
|
parser->content_length -= to_read;
|
|
|
|
p += to_read - 1;
|
|
|
|
|
|
|
|
if (parser->content_length == 0) {
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_chunk_data_almost_done);
|
2012-07-11 11:31:32 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
case s_chunk_data_almost_done:
|
|
|
|
assert(parser->flags & F_CHUNKED);
|
|
|
|
assert(parser->content_length == 0);
|
|
|
|
STRICT_CHECK(ch != CR);
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_chunk_data_done);
|
2012-07-11 11:31:32 +02:00
|
|
|
CALLBACK_DATA(body);
|
|
|
|
break;
|
|
|
|
|
|
|
|
case s_chunk_data_done:
|
|
|
|
assert(parser->flags & F_CHUNKED);
|
|
|
|
STRICT_CHECK(ch != LF);
|
|
|
|
parser->nread = 0;
|
2015-01-05 10:28:25 +01:00
|
|
|
UPDATE_STATE(s_chunk_size_start);
|
2015-05-21 18:55:47 +02:00
|
|
|
CALLBACK_NOTIFY(chunk_complete);
|
2012-07-11 11:31:32 +02:00
|
|
|
break;
|
|
|
|
|
|
|
|
default:
|
|
|
|
assert(0 && "unhandled state");
|
|
|
|
SET_ERRNO(HPE_INVALID_INTERNAL_STATE);
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Run callbacks for any marks that we have leftover after we ran our of
|
|
|
|
* bytes. There should be at most one of these set, so it's OK to invoke
|
|
|
|
* them in series (unset marks will not result in callbacks).
|
|
|
|
*
|
|
|
|
* We use the NOADVANCE() variety of callbacks here because 'p' has already
|
|
|
|
* overflowed 'data' and this allows us to correct for the off-by-one that
|
|
|
|
* we'd otherwise have (since CALLBACK_DATA() is meant to be run with a 'p'
|
|
|
|
* value that's in-bounds).
|
|
|
|
*/
|
|
|
|
|
|
|
|
assert(((header_field_mark ? 1 : 0) +
|
|
|
|
(header_value_mark ? 1 : 0) +
|
|
|
|
(url_mark ? 1 : 0) +
|
2014-05-15 17:23:03 +02:00
|
|
|
(body_mark ? 1 : 0) +
|
|
|
|
(status_mark ? 1 : 0)) <= 1);
|
2012-07-11 11:31:32 +02:00
|
|
|
|
|
|
|
CALLBACK_DATA_NOADVANCE(header_field);
|
|
|
|
CALLBACK_DATA_NOADVANCE(header_value);
|
|
|
|
CALLBACK_DATA_NOADVANCE(url);
|
|
|
|
CALLBACK_DATA_NOADVANCE(body);
|
2014-05-15 17:23:03 +02:00
|
|
|
CALLBACK_DATA_NOADVANCE(status);
|
2016-05-28 09:08:20 +02:00
|
|
|
#if HTTP_PARSER_METHOD_CB
|
|
|
|
CALLBACK_DATA_NOADVANCE(method);
|
|
|
|
#endif
|
2012-07-11 11:31:32 +02:00
|
|
|
|
2015-01-05 10:28:25 +01:00
|
|
|
RETURN(len);
|
2012-07-11 11:31:32 +02:00
|
|
|
|
|
|
|
error:
|
|
|
|
if (HTTP_PARSER_ERRNO(parser) == HPE_OK) {
|
|
|
|
SET_ERRNO(HPE_UNKNOWN);
|
|
|
|
}
|
|
|
|
|
2015-01-05 10:28:25 +01:00
|
|
|
RETURN(p - data);
|
2012-07-11 11:31:32 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
/* Does the parser need to see an EOF to find the end of the message? */
|
|
|
|
int
|
2012-09-10 17:12:23 +02:00
|
|
|
http_message_needs_eof (const http_parser *parser)
|
2012-07-11 11:31:32 +02:00
|
|
|
{
|
|
|
|
if (parser->type == HTTP_REQUEST) {
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* See RFC 2616 section 4.4 */
|
|
|
|
if (parser->status_code / 100 == 1 || /* 1xx e.g. Continue */
|
|
|
|
parser->status_code == 204 || /* No Content */
|
|
|
|
parser->status_code == 304 || /* Not Modified */
|
|
|
|
parser->flags & F_SKIPBODY) { /* response to a HEAD request */
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
if ((parser->flags & F_CHUNKED) || parser->content_length != ULLONG_MAX) {
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
int
|
2012-09-10 17:12:23 +02:00
|
|
|
http_should_keep_alive (const http_parser *parser)
|
2012-07-11 11:31:32 +02:00
|
|
|
{
|
|
|
|
if (parser->http_major > 0 && parser->http_minor > 0) {
|
|
|
|
/* HTTP/1.1 */
|
|
|
|
if (parser->flags & F_CONNECTION_CLOSE) {
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
} else {
|
|
|
|
/* HTTP/1.0 or earlier */
|
|
|
|
if (!(parser->flags & F_CONNECTION_KEEP_ALIVE)) {
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
return !http_message_needs_eof(parser);
|
|
|
|
}
|
|
|
|
|
|
|
|
|
2012-09-10 17:12:23 +02:00
|
|
|
const char *
|
|
|
|
http_method_str (enum http_method m)
|
2012-07-11 11:31:32 +02:00
|
|
|
{
|
2012-09-10 17:12:23 +02:00
|
|
|
return ELEM_AT(method_strings, m, "<unknown>");
|
2012-07-11 11:31:32 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
void
|
|
|
|
http_parser_init (http_parser *parser, enum http_parser_type t)
|
|
|
|
{
|
|
|
|
void *data = parser->data; /* preserve application data */
|
|
|
|
memset(parser, 0, sizeof(*parser));
|
|
|
|
parser->data = data;
|
|
|
|
parser->type = t;
|
|
|
|
parser->state = (t == HTTP_REQUEST ? s_start_req : (t == HTTP_RESPONSE ? s_start_res : s_start_req_or_res));
|
|
|
|
parser->http_errno = HPE_OK;
|
|
|
|
}
|
|
|
|
|
2015-02-20 11:51:41 +01:00
|
|
|
void
|
|
|
|
http_parser_settings_init(http_parser_settings *settings)
|
|
|
|
{
|
|
|
|
memset(settings, 0, sizeof(*settings));
|
|
|
|
}
|
|
|
|
|
2012-07-11 11:31:32 +02:00
|
|
|
const char *
|
|
|
|
http_errno_name(enum http_errno err) {
|
2015-07-25 05:53:17 +02:00
|
|
|
assert(((size_t) err) < ARRAY_SIZE(http_strerror_tab));
|
2012-07-11 11:31:32 +02:00
|
|
|
return http_strerror_tab[err].name;
|
|
|
|
}
|
|
|
|
|
|
|
|
const char *
|
|
|
|
http_errno_description(enum http_errno err) {
|
2015-07-25 05:53:17 +02:00
|
|
|
assert(((size_t) err) < ARRAY_SIZE(http_strerror_tab));
|
2012-07-11 11:31:32 +02:00
|
|
|
return http_strerror_tab[err].description;
|
|
|
|
}
|
|
|
|
|
2012-09-10 17:12:23 +02:00
|
|
|
static enum http_host_state
|
|
|
|
http_parse_host_char(enum http_host_state s, const char ch) {
|
|
|
|
switch(s) {
|
|
|
|
case s_http_userinfo:
|
|
|
|
case s_http_userinfo_start:
|
|
|
|
if (ch == '@') {
|
|
|
|
return s_http_host_start;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (IS_USERINFO_CHAR(ch)) {
|
|
|
|
return s_http_userinfo;
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
|
|
|
|
case s_http_host_start:
|
|
|
|
if (ch == '[') {
|
|
|
|
return s_http_host_v6_start;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (IS_HOST_CHAR(ch)) {
|
|
|
|
return s_http_host;
|
|
|
|
}
|
|
|
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
case s_http_host:
|
|
|
|
if (IS_HOST_CHAR(ch)) {
|
|
|
|
return s_http_host;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* FALLTHROUGH */
|
|
|
|
case s_http_host_v6_end:
|
|
|
|
if (ch == ':') {
|
|
|
|
return s_http_host_port_start;
|
|
|
|
}
|
|
|
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
case s_http_host_v6:
|
|
|
|
if (ch == ']') {
|
|
|
|
return s_http_host_v6_end;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* FALLTHROUGH */
|
|
|
|
case s_http_host_v6_start:
|
2012-12-22 17:13:02 +01:00
|
|
|
if (IS_HEX(ch) || ch == ':' || ch == '.') {
|
2012-09-10 17:12:23 +02:00
|
|
|
return s_http_host_v6;
|
|
|
|
}
|
|
|
|
|
2015-07-25 05:53:17 +02:00
|
|
|
if (s == s_http_host_v6 && ch == '%') {
|
|
|
|
return s_http_host_v6_zone_start;
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
|
|
|
|
case s_http_host_v6_zone:
|
|
|
|
if (ch == ']') {
|
|
|
|
return s_http_host_v6_end;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* FALLTHROUGH */
|
|
|
|
case s_http_host_v6_zone_start:
|
|
|
|
/* RFC 6874 Zone ID consists of 1*( unreserved / pct-encoded) */
|
|
|
|
if (IS_ALPHANUM(ch) || ch == '%' || ch == '.' || ch == '-' || ch == '_' ||
|
|
|
|
ch == '~') {
|
|
|
|
return s_http_host_v6_zone;
|
|
|
|
}
|
2012-09-10 17:12:23 +02:00
|
|
|
break;
|
|
|
|
|
|
|
|
case s_http_host_port:
|
|
|
|
case s_http_host_port_start:
|
|
|
|
if (IS_NUM(ch)) {
|
|
|
|
return s_http_host_port;
|
|
|
|
}
|
|
|
|
|
|
|
|
break;
|
|
|
|
|
|
|
|
default:
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
return s_http_host_dead;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int
|
|
|
|
http_parse_host(const char * buf, struct http_parser_url *u, int found_at) {
|
|
|
|
enum http_host_state s;
|
|
|
|
|
|
|
|
const char *p;
|
|
|
|
size_t buflen = u->field_data[UF_HOST].off + u->field_data[UF_HOST].len;
|
|
|
|
|
2016-05-28 05:17:17 +02:00
|
|
|
assert(u->field_set & (1 << UF_HOST));
|
|
|
|
|
2012-09-10 17:12:23 +02:00
|
|
|
u->field_data[UF_HOST].len = 0;
|
|
|
|
|
|
|
|
s = found_at ? s_http_userinfo_start : s_http_host_start;
|
|
|
|
|
|
|
|
for (p = buf + u->field_data[UF_HOST].off; p < buf + buflen; p++) {
|
|
|
|
enum http_host_state new_s = http_parse_host_char(s, *p);
|
|
|
|
|
|
|
|
if (new_s == s_http_host_dead) {
|
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
|
|
|
|
switch(new_s) {
|
|
|
|
case s_http_host:
|
|
|
|
if (s != s_http_host) {
|
|
|
|
u->field_data[UF_HOST].off = p - buf;
|
|
|
|
}
|
|
|
|
u->field_data[UF_HOST].len++;
|
|
|
|
break;
|
|
|
|
|
|
|
|
case s_http_host_v6:
|
|
|
|
if (s != s_http_host_v6) {
|
|
|
|
u->field_data[UF_HOST].off = p - buf;
|
|
|
|
}
|
|
|
|
u->field_data[UF_HOST].len++;
|
|
|
|
break;
|
|
|
|
|
2015-07-25 05:53:17 +02:00
|
|
|
case s_http_host_v6_zone_start:
|
|
|
|
case s_http_host_v6_zone:
|
|
|
|
u->field_data[UF_HOST].len++;
|
|
|
|
break;
|
|
|
|
|
2012-09-10 17:12:23 +02:00
|
|
|
case s_http_host_port:
|
|
|
|
if (s != s_http_host_port) {
|
|
|
|
u->field_data[UF_PORT].off = p - buf;
|
|
|
|
u->field_data[UF_PORT].len = 0;
|
|
|
|
u->field_set |= (1 << UF_PORT);
|
|
|
|
}
|
|
|
|
u->field_data[UF_PORT].len++;
|
|
|
|
break;
|
|
|
|
|
|
|
|
case s_http_userinfo:
|
|
|
|
if (s != s_http_userinfo) {
|
|
|
|
u->field_data[UF_USERINFO].off = p - buf ;
|
|
|
|
u->field_data[UF_USERINFO].len = 0;
|
|
|
|
u->field_set |= (1 << UF_USERINFO);
|
|
|
|
}
|
|
|
|
u->field_data[UF_USERINFO].len++;
|
|
|
|
break;
|
|
|
|
|
|
|
|
default:
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
s = new_s;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Make sure we don't end somewhere unexpected */
|
|
|
|
switch (s) {
|
|
|
|
case s_http_host_start:
|
|
|
|
case s_http_host_v6_start:
|
|
|
|
case s_http_host_v6:
|
2015-07-25 05:53:17 +02:00
|
|
|
case s_http_host_v6_zone_start:
|
|
|
|
case s_http_host_v6_zone:
|
2012-09-10 17:12:23 +02:00
|
|
|
case s_http_host_port_start:
|
|
|
|
case s_http_userinfo:
|
|
|
|
case s_http_userinfo_start:
|
|
|
|
return 1;
|
|
|
|
default:
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2016-01-27 12:46:59 +01:00
|
|
|
void
|
|
|
|
http_parser_url_init(struct http_parser_url *u) {
|
|
|
|
memset(u, 0, sizeof(*u));
|
|
|
|
}
|
|
|
|
|
2012-07-11 11:31:32 +02:00
|
|
|
int
|
|
|
|
http_parser_parse_url(const char *buf, size_t buflen, int is_connect,
|
|
|
|
struct http_parser_url *u)
|
|
|
|
{
|
|
|
|
enum state s;
|
|
|
|
const char *p;
|
|
|
|
enum http_parser_url_fields uf, old_uf;
|
2012-09-10 17:12:23 +02:00
|
|
|
int found_at = 0;
|
2012-07-11 11:31:32 +02:00
|
|
|
|
|
|
|
u->port = u->field_set = 0;
|
2012-09-10 17:12:23 +02:00
|
|
|
s = is_connect ? s_req_server_start : s_req_spaces_before_url;
|
2014-08-08 13:12:34 +02:00
|
|
|
old_uf = UF_MAX;
|
2012-07-11 11:31:32 +02:00
|
|
|
|
|
|
|
for (p = buf; p < buf + buflen; p++) {
|
|
|
|
s = parse_url_char(s, *p);
|
|
|
|
|
|
|
|
/* Figure out the next field that we're operating on */
|
|
|
|
switch (s) {
|
|
|
|
case s_dead:
|
|
|
|
return 1;
|
|
|
|
|
|
|
|
/* Skip delimeters */
|
|
|
|
case s_req_schema_slash:
|
|
|
|
case s_req_schema_slash_slash:
|
2012-09-10 17:12:23 +02:00
|
|
|
case s_req_server_start:
|
2012-07-11 11:31:32 +02:00
|
|
|
case s_req_query_string_start:
|
|
|
|
case s_req_fragment_start:
|
|
|
|
continue;
|
|
|
|
|
|
|
|
case s_req_schema:
|
|
|
|
uf = UF_SCHEMA;
|
|
|
|
break;
|
|
|
|
|
2012-09-10 17:12:23 +02:00
|
|
|
case s_req_server_with_at:
|
|
|
|
found_at = 1;
|
2012-07-11 11:31:32 +02:00
|
|
|
|
2012-09-10 17:12:23 +02:00
|
|
|
/* FALLTROUGH */
|
|
|
|
case s_req_server:
|
|
|
|
uf = UF_HOST;
|
2012-07-11 11:31:32 +02:00
|
|
|
break;
|
|
|
|
|
|
|
|
case s_req_path:
|
|
|
|
uf = UF_PATH;
|
|
|
|
break;
|
|
|
|
|
|
|
|
case s_req_query_string:
|
|
|
|
uf = UF_QUERY;
|
|
|
|
break;
|
|
|
|
|
|
|
|
case s_req_fragment:
|
|
|
|
uf = UF_FRAGMENT;
|
|
|
|
break;
|
|
|
|
|
|
|
|
default:
|
|
|
|
assert(!"Unexpected state");
|
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Nothing's changed; soldier on */
|
|
|
|
if (uf == old_uf) {
|
|
|
|
u->field_data[uf].len++;
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
|
|
|
|
u->field_data[uf].off = p - buf;
|
|
|
|
u->field_data[uf].len = 1;
|
|
|
|
|
|
|
|
u->field_set |= (1 << uf);
|
|
|
|
old_uf = uf;
|
|
|
|
}
|
|
|
|
|
2012-09-10 17:12:23 +02:00
|
|
|
/* host must be present if there is a schema */
|
|
|
|
/* parsing http:///toto will fail */
|
2015-05-21 18:55:47 +02:00
|
|
|
if ((u->field_set & (1 << UF_SCHEMA)) &&
|
|
|
|
(u->field_set & (1 << UF_HOST)) == 0) {
|
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (u->field_set & (1 << UF_HOST)) {
|
2012-09-10 17:12:23 +02:00
|
|
|
if (http_parse_host(buf, u, found_at) != 0) {
|
|
|
|
return 1;
|
|
|
|
}
|
2012-07-11 11:31:32 +02:00
|
|
|
}
|
|
|
|
|
2012-09-10 17:12:23 +02:00
|
|
|
/* CONNECT requests can only contain "hostname:port" */
|
|
|
|
if (is_connect && u->field_set != ((1 << UF_HOST)|(1 << UF_PORT))) {
|
2012-07-11 11:31:32 +02:00
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (u->field_set & (1 << UF_PORT)) {
|
|
|
|
/* Don't bother with endp; we've already validated the string */
|
|
|
|
unsigned long v = strtoul(buf + u->field_data[UF_PORT].off, NULL, 10);
|
|
|
|
|
|
|
|
/* Ports have a max value of 2^16 */
|
|
|
|
if (v > 0xffff) {
|
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
|
|
|
|
u->port = (uint16_t) v;
|
|
|
|
}
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
void
|
|
|
|
http_parser_pause(http_parser *parser, int paused) {
|
|
|
|
/* Users should only be pausing/unpausing a parser that is not in an error
|
|
|
|
* state. In non-debug builds, there's not much that we can do about this
|
|
|
|
* other than ignore it.
|
|
|
|
*/
|
|
|
|
if (HTTP_PARSER_ERRNO(parser) == HPE_OK ||
|
|
|
|
HTTP_PARSER_ERRNO(parser) == HPE_PAUSED) {
|
|
|
|
SET_ERRNO((paused) ? HPE_PAUSED : HPE_OK);
|
|
|
|
} else {
|
|
|
|
assert(0 && "Attempting to pause parser in error state");
|
|
|
|
}
|
|
|
|
}
|
2012-09-10 17:12:23 +02:00
|
|
|
|
|
|
|
int
|
|
|
|
http_body_is_final(const struct http_parser *parser) {
|
|
|
|
return parser->state == s_message_done;
|
|
|
|
}
|
2013-11-14 15:53:48 +01:00
|
|
|
|
|
|
|
unsigned long
|
|
|
|
http_parser_version(void) {
|
|
|
|
return HTTP_PARSER_VERSION_MAJOR * 0x10000 |
|
|
|
|
HTTP_PARSER_VERSION_MINOR * 0x00100 |
|
|
|
|
HTTP_PARSER_VERSION_PATCH * 0x00001;
|
|
|
|
}
|