ruby_http_parser 0.0.1
Sign up to get free protection for your applications and to get access to all the features.
- data/README.md +27 -0
- data/Rakefile +11 -0
- data/bench/thin.rb +57 -0
- data/ext/http-parser/LICENSE-MIT +19 -0
- data/ext/http-parser/Makefile +33 -0
- data/ext/http-parser/README.md +136 -0
- data/ext/http-parser/http_parser.c +1468 -0
- data/ext/http-parser/http_parser.h +141 -0
- data/ext/http-parser/test.c +1176 -0
- data/ext/ruby_http_parser/ext_help.h +18 -0
- data/ext/ruby_http_parser/extconf.rb +7 -0
- data/ext/ruby_http_parser/ruby_http_parser.c +275 -0
- data/lib/net/http/parser.rb +1 -0
- data/lib/ruby_http_parser.bundle +0 -0
- data/ruby_http_parser.gemspec +13 -0
- data/spec/request_parser_spec.rb +27 -0
- data/spec/spec_helper.rb +3 -0
- data/tmp/universal-darwin10.0/ruby_http_parser/1.8.7/Makefile +157 -0
- data/tmp/universal-darwin10.0/ruby_http_parser/1.8.7/ruby_http_parser.bundle +0 -0
- data/tmp/universal-darwin10.0/ruby_http_parser/1.8.7/ruby_http_parser.o +0 -0
- metadata +75 -0
@@ -0,0 +1,141 @@
|
|
1
|
+
/* Copyright 2009 Ryan Dahl <ry@tinyclouds.org>
|
2
|
+
*
|
3
|
+
* Permission is hereby granted, free of charge, to any person obtaining a copy
|
4
|
+
* of this software and associated documentation files (the "Software"), to
|
5
|
+
* deal in the Software without restriction, including without limitation the
|
6
|
+
* rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
|
7
|
+
* sell copies of the Software, and to permit persons to whom the Software is
|
8
|
+
* furnished to do so, subject to the following conditions:
|
9
|
+
*
|
10
|
+
* The above copyright notice and this permission notice shall be included in
|
11
|
+
* all copies or substantial portions of the Software.
|
12
|
+
*
|
13
|
+
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
|
14
|
+
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
|
15
|
+
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
|
16
|
+
* AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
|
17
|
+
* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
|
18
|
+
* FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
|
19
|
+
* IN THE SOFTWARE.
|
20
|
+
*/
|
21
|
+
#ifndef http_parser_h
|
22
|
+
#define http_parser_h
|
23
|
+
#ifdef __cplusplus
|
24
|
+
extern "C" {
|
25
|
+
#endif
|
26
|
+
|
27
|
+
#ifdef _MSC_VER
|
28
|
+
# include <stddef.h>
|
29
|
+
#endif
|
30
|
+
#include <sys/types.h>
|
31
|
+
|
32
|
+
/* Compile with -DHTTP_PARSER_STRICT=0 to make less checks, but run
|
33
|
+
* faster
|
34
|
+
*/
|
35
|
+
#ifndef HTTP_PARSER_STRICT
|
36
|
+
# define HTTP_PARSER_STRICT 1
|
37
|
+
#else
|
38
|
+
# define HTTP_PARSER_STRICT 0
|
39
|
+
#endif
|
40
|
+
|
41
|
+
typedef struct http_parser http_parser;
|
42
|
+
|
43
|
+
/* Callbacks should return non-zero to indicate an error. The parse will
|
44
|
+
* then halt execution.
|
45
|
+
*
|
46
|
+
* http_data_cb does not return data chunks. It will be call arbitrarally
|
47
|
+
* many times for each string. E.G. you might get 10 callbacks for "on_path"
|
48
|
+
* each providing just a few characters more data.
|
49
|
+
*/
|
50
|
+
typedef int (*http_data_cb) (http_parser*, const char *at, size_t length);
|
51
|
+
typedef int (*http_cb) (http_parser*);
|
52
|
+
|
53
|
+
/* Should be at least one longer than the longest request method */
|
54
|
+
#define HTTP_PARSER_MAX_METHOD_LEN 10
|
55
|
+
|
56
|
+
/* Request Methods */
|
57
|
+
enum http_method
|
58
|
+
{ HTTP_DELETE = 0x0001
|
59
|
+
, HTTP_GET = 0x0002
|
60
|
+
, HTTP_HEAD = 0x0004
|
61
|
+
, HTTP_POST = 0x0008
|
62
|
+
, HTTP_PUT = 0x0010
|
63
|
+
/* pathological */
|
64
|
+
, HTTP_CONNECT = 0x0020
|
65
|
+
, HTTP_OPTIONS = 0x0040
|
66
|
+
, HTTP_TRACE = 0x0080
|
67
|
+
/* webdav */
|
68
|
+
, HTTP_COPY = 0x0100
|
69
|
+
, HTTP_LOCK = 0x0200
|
70
|
+
, HTTP_MKCOL = 0x0400
|
71
|
+
, HTTP_MOVE = 0x0800
|
72
|
+
, HTTP_PROPFIND = 0x1000
|
73
|
+
, HTTP_PROPPATCH = 0x2000
|
74
|
+
, HTTP_UNLOCK = 0x4000
|
75
|
+
};
|
76
|
+
|
77
|
+
struct http_parser {
|
78
|
+
/** PRIVATE **/
|
79
|
+
unsigned short state;
|
80
|
+
unsigned short header_state;
|
81
|
+
size_t index;
|
82
|
+
|
83
|
+
char flags;
|
84
|
+
|
85
|
+
ssize_t body_read;
|
86
|
+
ssize_t content_length;
|
87
|
+
|
88
|
+
const char *header_field_mark;
|
89
|
+
size_t header_field_size;
|
90
|
+
const char *header_value_mark;
|
91
|
+
size_t header_value_size;
|
92
|
+
const char *query_string_mark;
|
93
|
+
size_t query_string_size;
|
94
|
+
const char *path_mark;
|
95
|
+
size_t path_size;
|
96
|
+
const char *url_mark;
|
97
|
+
size_t url_size;
|
98
|
+
const char *fragment_mark;
|
99
|
+
size_t fragment_size;
|
100
|
+
|
101
|
+
/** READ-ONLY **/
|
102
|
+
unsigned short status_code; /* responses only */
|
103
|
+
enum http_method method; /* requests only */
|
104
|
+
unsigned short http_major;
|
105
|
+
unsigned short http_minor;
|
106
|
+
char buffer[HTTP_PARSER_MAX_METHOD_LEN];
|
107
|
+
|
108
|
+
/** PUBLIC **/
|
109
|
+
void *data; /* A pointer to get hook to the "connection" or "socket" object */
|
110
|
+
|
111
|
+
/* an ordered list of callbacks */
|
112
|
+
|
113
|
+
http_cb on_message_begin;
|
114
|
+
|
115
|
+
/* requests only */
|
116
|
+
http_data_cb on_path;
|
117
|
+
http_data_cb on_query_string;
|
118
|
+
http_data_cb on_url;
|
119
|
+
http_data_cb on_fragment;
|
120
|
+
|
121
|
+
http_data_cb on_header_field;
|
122
|
+
http_data_cb on_header_value;
|
123
|
+
http_cb on_headers_complete;
|
124
|
+
http_data_cb on_body;
|
125
|
+
http_cb on_message_complete;
|
126
|
+
};
|
127
|
+
|
128
|
+
void http_parser_init(http_parser *parser);
|
129
|
+
size_t http_parse_requests(http_parser *parser, const char *data, size_t len);
|
130
|
+
size_t http_parse_responses(http_parser *parser, const char *data, size_t len);
|
131
|
+
/* Call this in the on_headers_complete or on_message_complete callback to
|
132
|
+
* determine if this will be the last message on the connection.
|
133
|
+
* If you are the server, respond with the "Connection: close" header
|
134
|
+
* if you are the client, close the connection.
|
135
|
+
*/
|
136
|
+
int http_should_keep_alive(http_parser *parser);
|
137
|
+
|
138
|
+
#ifdef __cplusplus
|
139
|
+
}
|
140
|
+
#endif
|
141
|
+
#endif
|
@@ -0,0 +1,1176 @@
|
|
1
|
+
/* Copyright 2009 Ryan Dahl <ry@tinyclouds.org>
|
2
|
+
*
|
3
|
+
* Permission is hereby granted, free of charge, to any person obtaining a copy
|
4
|
+
* of this software and associated documentation files (the "Software"), to
|
5
|
+
* deal in the Software without restriction, including without limitation the
|
6
|
+
* rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
|
7
|
+
* sell copies of the Software, and to permit persons to whom the Software is
|
8
|
+
* furnished to do so, subject to the following conditions:
|
9
|
+
*
|
10
|
+
* The above copyright notice and this permission notice shall be included in
|
11
|
+
* all copies or substantial portions of the Software.
|
12
|
+
*
|
13
|
+
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
|
14
|
+
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
|
15
|
+
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
|
16
|
+
* AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
|
17
|
+
* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
|
18
|
+
* FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
|
19
|
+
* IN THE SOFTWARE.
|
20
|
+
*/
|
21
|
+
#include "http_parser.h"
|
22
|
+
#include <stdlib.h>
|
23
|
+
#include <assert.h>
|
24
|
+
#include <stdio.h>
|
25
|
+
#include <stdlib.h> /* rand */
|
26
|
+
#include <string.h>
|
27
|
+
#include <stdarg.h>
|
28
|
+
|
29
|
+
#undef TRUE
|
30
|
+
#define TRUE 1
|
31
|
+
#undef FALSE
|
32
|
+
#define FALSE 0
|
33
|
+
|
34
|
+
#define MAX_HEADERS 10
|
35
|
+
#define MAX_ELEMENT_SIZE 500
|
36
|
+
|
37
|
+
enum message_type { REQUEST, RESPONSE };
|
38
|
+
|
39
|
+
static http_parser *parser;
|
40
|
+
|
41
|
+
struct message {
|
42
|
+
const char *name; // for debugging purposes
|
43
|
+
const char *raw;
|
44
|
+
enum message_type type;
|
45
|
+
enum http_method method;
|
46
|
+
int status_code;
|
47
|
+
char request_path[MAX_ELEMENT_SIZE];
|
48
|
+
char request_url[MAX_ELEMENT_SIZE];
|
49
|
+
char fragment[MAX_ELEMENT_SIZE];
|
50
|
+
char query_string[MAX_ELEMENT_SIZE];
|
51
|
+
char body[MAX_ELEMENT_SIZE];
|
52
|
+
int num_headers;
|
53
|
+
enum { NONE=0, FIELD, VALUE } last_header_element;
|
54
|
+
char headers [MAX_HEADERS][2][MAX_ELEMENT_SIZE];
|
55
|
+
int should_keep_alive;
|
56
|
+
|
57
|
+
unsigned short http_major;
|
58
|
+
unsigned short http_minor;
|
59
|
+
|
60
|
+
int message_begin_cb_called;
|
61
|
+
int headers_complete_cb_called;
|
62
|
+
int message_complete_cb_called;
|
63
|
+
int message_complete_on_eof;
|
64
|
+
};
|
65
|
+
|
66
|
+
static int currently_parsing_eof;
|
67
|
+
|
68
|
+
inline size_t parse (enum message_type t, const char *buf, size_t len)
|
69
|
+
{
|
70
|
+
size_t nparsed;
|
71
|
+
currently_parsing_eof = (len == 0);
|
72
|
+
nparsed = (t == REQUEST ? http_parse_requests(parser, buf, len)
|
73
|
+
: http_parse_responses(parser, buf, len));
|
74
|
+
return nparsed;
|
75
|
+
}
|
76
|
+
|
77
|
+
static struct message messages[5];
|
78
|
+
static int num_messages;
|
79
|
+
|
80
|
+
/* * R E Q U E S T S * */
|
81
|
+
const struct message requests[] =
|
82
|
+
#define CURL_GET 0
|
83
|
+
{ {.name= "curl get"
|
84
|
+
,.type= REQUEST
|
85
|
+
,.raw= "GET /test HTTP/1.1\r\n"
|
86
|
+
"User-Agent: curl/7.18.0 (i486-pc-linux-gnu) libcurl/7.18.0 OpenSSL/0.9.8g zlib/1.2.3.3 libidn/1.1\r\n"
|
87
|
+
"Host: 0.0.0.0=5000\r\n"
|
88
|
+
"Accept: */*\r\n"
|
89
|
+
"\r\n"
|
90
|
+
,.should_keep_alive= TRUE
|
91
|
+
,.message_complete_on_eof= FALSE
|
92
|
+
,.http_major= 1
|
93
|
+
,.http_minor= 1
|
94
|
+
,.method= HTTP_GET
|
95
|
+
,.query_string= ""
|
96
|
+
,.fragment= ""
|
97
|
+
,.request_path= "/test"
|
98
|
+
,.request_url= "/test"
|
99
|
+
,.num_headers= 3
|
100
|
+
,.headers=
|
101
|
+
{ { "User-Agent", "curl/7.18.0 (i486-pc-linux-gnu) libcurl/7.18.0 OpenSSL/0.9.8g zlib/1.2.3.3 libidn/1.1" }
|
102
|
+
, { "Host", "0.0.0.0=5000" }
|
103
|
+
, { "Accept", "*/*" }
|
104
|
+
}
|
105
|
+
,.body= ""
|
106
|
+
}
|
107
|
+
|
108
|
+
#define FIREFOX_GET 1
|
109
|
+
, {.name= "firefox get"
|
110
|
+
,.type= REQUEST
|
111
|
+
,.raw= "GET /favicon.ico HTTP/1.1\r\n"
|
112
|
+
"Host: 0.0.0.0=5000\r\n"
|
113
|
+
"User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.9) Gecko/2008061015 Firefox/3.0\r\n"
|
114
|
+
"Accept: text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8\r\n"
|
115
|
+
"Accept-Language: en-us,en;q=0.5\r\n"
|
116
|
+
"Accept-Encoding: gzip,deflate\r\n"
|
117
|
+
"Accept-Charset: ISO-8859-1,utf-8;q=0.7,*;q=0.7\r\n"
|
118
|
+
"Keep-Alive: 300\r\n"
|
119
|
+
"Connection: keep-alive\r\n"
|
120
|
+
"\r\n"
|
121
|
+
,.should_keep_alive= TRUE
|
122
|
+
,.message_complete_on_eof= FALSE
|
123
|
+
,.http_major= 1
|
124
|
+
,.http_minor= 1
|
125
|
+
,.method= HTTP_GET
|
126
|
+
,.query_string= ""
|
127
|
+
,.fragment= ""
|
128
|
+
,.request_path= "/favicon.ico"
|
129
|
+
,.request_url= "/favicon.ico"
|
130
|
+
,.num_headers= 8
|
131
|
+
,.headers=
|
132
|
+
{ { "Host", "0.0.0.0=5000" }
|
133
|
+
, { "User-Agent", "Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.9) Gecko/2008061015 Firefox/3.0" }
|
134
|
+
, { "Accept", "text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8" }
|
135
|
+
, { "Accept-Language", "en-us,en;q=0.5" }
|
136
|
+
, { "Accept-Encoding", "gzip,deflate" }
|
137
|
+
, { "Accept-Charset", "ISO-8859-1,utf-8;q=0.7,*;q=0.7" }
|
138
|
+
, { "Keep-Alive", "300" }
|
139
|
+
, { "Connection", "keep-alive" }
|
140
|
+
}
|
141
|
+
,.body= ""
|
142
|
+
}
|
143
|
+
|
144
|
+
#define DUMBFUCK 2
|
145
|
+
, {.name= "dumbfuck"
|
146
|
+
,.type= REQUEST
|
147
|
+
,.raw= "GET /dumbfuck HTTP/1.1\r\n"
|
148
|
+
"aaaaaaaaaaaaa:++++++++++\r\n"
|
149
|
+
"\r\n"
|
150
|
+
,.should_keep_alive= TRUE
|
151
|
+
,.message_complete_on_eof= FALSE
|
152
|
+
,.http_major= 1
|
153
|
+
,.http_minor= 1
|
154
|
+
,.method= HTTP_GET
|
155
|
+
,.query_string= ""
|
156
|
+
,.fragment= ""
|
157
|
+
,.request_path= "/dumbfuck"
|
158
|
+
,.request_url= "/dumbfuck"
|
159
|
+
,.num_headers= 1
|
160
|
+
,.headers=
|
161
|
+
{ { "aaaaaaaaaaaaa", "++++++++++" }
|
162
|
+
}
|
163
|
+
,.body= ""
|
164
|
+
}
|
165
|
+
|
166
|
+
#define FRAGMENT_IN_URI 3
|
167
|
+
, {.name= "fragment in url"
|
168
|
+
,.type= REQUEST
|
169
|
+
,.raw= "GET /forums/1/topics/2375?page=1#posts-17408 HTTP/1.1\r\n"
|
170
|
+
"\r\n"
|
171
|
+
,.should_keep_alive= TRUE
|
172
|
+
,.message_complete_on_eof= FALSE
|
173
|
+
,.http_major= 1
|
174
|
+
,.http_minor= 1
|
175
|
+
,.method= HTTP_GET
|
176
|
+
,.query_string= "page=1"
|
177
|
+
,.fragment= "posts-17408"
|
178
|
+
,.request_path= "/forums/1/topics/2375"
|
179
|
+
/* XXX request url does include fragment? */
|
180
|
+
,.request_url= "/forums/1/topics/2375?page=1#posts-17408"
|
181
|
+
,.num_headers= 0
|
182
|
+
,.body= ""
|
183
|
+
}
|
184
|
+
|
185
|
+
#define GET_NO_HEADERS_NO_BODY 4
|
186
|
+
, {.name= "get no headers no body"
|
187
|
+
,.type= REQUEST
|
188
|
+
,.raw= "GET /get_no_headers_no_body/world HTTP/1.1\r\n"
|
189
|
+
"\r\n"
|
190
|
+
,.should_keep_alive= TRUE
|
191
|
+
,.message_complete_on_eof= FALSE /* would need Connection: close */
|
192
|
+
,.http_major= 1
|
193
|
+
,.http_minor= 1
|
194
|
+
,.method= HTTP_GET
|
195
|
+
,.query_string= ""
|
196
|
+
,.fragment= ""
|
197
|
+
,.request_path= "/get_no_headers_no_body/world"
|
198
|
+
,.request_url= "/get_no_headers_no_body/world"
|
199
|
+
,.num_headers= 0
|
200
|
+
,.body= ""
|
201
|
+
}
|
202
|
+
|
203
|
+
#define GET_ONE_HEADER_NO_BODY 5
|
204
|
+
, {.name= "get one header no body"
|
205
|
+
,.type= REQUEST
|
206
|
+
,.raw= "GET /get_one_header_no_body HTTP/1.1\r\n"
|
207
|
+
"Accept: */*\r\n"
|
208
|
+
"\r\n"
|
209
|
+
,.should_keep_alive= TRUE
|
210
|
+
,.message_complete_on_eof= FALSE /* would need Connection: close */
|
211
|
+
,.http_major= 1
|
212
|
+
,.http_minor= 1
|
213
|
+
,.method= HTTP_GET
|
214
|
+
,.query_string= ""
|
215
|
+
,.fragment= ""
|
216
|
+
,.request_path= "/get_one_header_no_body"
|
217
|
+
,.request_url= "/get_one_header_no_body"
|
218
|
+
,.num_headers= 1
|
219
|
+
,.headers=
|
220
|
+
{ { "Accept" , "*/*" }
|
221
|
+
}
|
222
|
+
,.body= ""
|
223
|
+
}
|
224
|
+
|
225
|
+
#define GET_FUNKY_CONTENT_LENGTH 6
|
226
|
+
, {.name= "get funky content length body hello"
|
227
|
+
,.type= REQUEST
|
228
|
+
,.raw= "GET /get_funky_content_length_body_hello HTTP/1.0\r\n"
|
229
|
+
"conTENT-Length: 5\r\n"
|
230
|
+
"\r\n"
|
231
|
+
"HELLO"
|
232
|
+
,.should_keep_alive= FALSE
|
233
|
+
,.message_complete_on_eof= FALSE
|
234
|
+
,.http_major= 1
|
235
|
+
,.http_minor= 0
|
236
|
+
,.method= HTTP_GET
|
237
|
+
,.query_string= ""
|
238
|
+
,.fragment= ""
|
239
|
+
,.request_path= "/get_funky_content_length_body_hello"
|
240
|
+
,.request_url= "/get_funky_content_length_body_hello"
|
241
|
+
,.num_headers= 1
|
242
|
+
,.headers=
|
243
|
+
{ { "conTENT-Length" , "5" }
|
244
|
+
}
|
245
|
+
,.body= "HELLO"
|
246
|
+
}
|
247
|
+
|
248
|
+
#define POST_IDENTITY_BODY_WORLD 7
|
249
|
+
, {.name= "post identity body world"
|
250
|
+
,.type= REQUEST
|
251
|
+
,.raw= "POST /post_identity_body_world?q=search#hey HTTP/1.1\r\n"
|
252
|
+
"Accept: */*\r\n"
|
253
|
+
"Transfer-Encoding: identity\r\n"
|
254
|
+
"Content-Length: 5\r\n"
|
255
|
+
"\r\n"
|
256
|
+
"World"
|
257
|
+
,.should_keep_alive= TRUE
|
258
|
+
,.message_complete_on_eof= FALSE
|
259
|
+
,.http_major= 1
|
260
|
+
,.http_minor= 1
|
261
|
+
,.method= HTTP_POST
|
262
|
+
,.query_string= "q=search"
|
263
|
+
,.fragment= "hey"
|
264
|
+
,.request_path= "/post_identity_body_world"
|
265
|
+
,.request_url= "/post_identity_body_world?q=search#hey"
|
266
|
+
,.num_headers= 3
|
267
|
+
,.headers=
|
268
|
+
{ { "Accept", "*/*" }
|
269
|
+
, { "Transfer-Encoding", "identity" }
|
270
|
+
, { "Content-Length", "5" }
|
271
|
+
}
|
272
|
+
,.body= "World"
|
273
|
+
}
|
274
|
+
|
275
|
+
#define POST_CHUNKED_ALL_YOUR_BASE 8
|
276
|
+
, {.name= "post - chunked body: all your base are belong to us"
|
277
|
+
,.type= REQUEST
|
278
|
+
,.raw= "POST /post_chunked_all_your_base HTTP/1.1\r\n"
|
279
|
+
"Transfer-Encoding: chunked\r\n"
|
280
|
+
"\r\n"
|
281
|
+
"1e\r\nall your base are belong to us\r\n"
|
282
|
+
"0\r\n"
|
283
|
+
"\r\n"
|
284
|
+
,.should_keep_alive= TRUE
|
285
|
+
,.message_complete_on_eof= FALSE
|
286
|
+
,.http_major= 1
|
287
|
+
,.http_minor= 1
|
288
|
+
,.method= HTTP_POST
|
289
|
+
,.query_string= ""
|
290
|
+
,.fragment= ""
|
291
|
+
,.request_path= "/post_chunked_all_your_base"
|
292
|
+
,.request_url= "/post_chunked_all_your_base"
|
293
|
+
,.num_headers= 1
|
294
|
+
,.headers=
|
295
|
+
{ { "Transfer-Encoding" , "chunked" }
|
296
|
+
}
|
297
|
+
,.body= "all your base are belong to us"
|
298
|
+
}
|
299
|
+
|
300
|
+
#define TWO_CHUNKS_MULT_ZERO_END 9
|
301
|
+
, {.name= "two chunks ; triple zero ending"
|
302
|
+
,.type= REQUEST
|
303
|
+
,.raw= "POST /two_chunks_mult_zero_end HTTP/1.1\r\n"
|
304
|
+
"Transfer-Encoding: chunked\r\n"
|
305
|
+
"\r\n"
|
306
|
+
"5\r\nhello\r\n"
|
307
|
+
"6\r\n world\r\n"
|
308
|
+
"000\r\n"
|
309
|
+
"\r\n"
|
310
|
+
,.should_keep_alive= TRUE
|
311
|
+
,.message_complete_on_eof= FALSE
|
312
|
+
,.http_major= 1
|
313
|
+
,.http_minor= 1
|
314
|
+
,.method= HTTP_POST
|
315
|
+
,.query_string= ""
|
316
|
+
,.fragment= ""
|
317
|
+
,.request_path= "/two_chunks_mult_zero_end"
|
318
|
+
,.request_url= "/two_chunks_mult_zero_end"
|
319
|
+
,.num_headers= 1
|
320
|
+
,.headers=
|
321
|
+
{ { "Transfer-Encoding", "chunked" }
|
322
|
+
}
|
323
|
+
,.body= "hello world"
|
324
|
+
}
|
325
|
+
|
326
|
+
#define CHUNKED_W_TRAILING_HEADERS 10
|
327
|
+
, {.name= "chunked with trailing headers. blech."
|
328
|
+
,.type= REQUEST
|
329
|
+
,.raw= "POST /chunked_w_trailing_headers HTTP/1.1\r\n"
|
330
|
+
"Transfer-Encoding: chunked\r\n"
|
331
|
+
"\r\n"
|
332
|
+
"5\r\nhello\r\n"
|
333
|
+
"6\r\n world\r\n"
|
334
|
+
"0\r\n"
|
335
|
+
"Vary: *\r\n"
|
336
|
+
"Content-Type: text/plain\r\n"
|
337
|
+
"\r\n"
|
338
|
+
,.should_keep_alive= TRUE
|
339
|
+
,.message_complete_on_eof= FALSE
|
340
|
+
,.http_major= 1
|
341
|
+
,.http_minor= 1
|
342
|
+
,.method= HTTP_POST
|
343
|
+
,.query_string= ""
|
344
|
+
,.fragment= ""
|
345
|
+
,.request_path= "/chunked_w_trailing_headers"
|
346
|
+
,.request_url= "/chunked_w_trailing_headers"
|
347
|
+
,.num_headers= 3
|
348
|
+
,.headers=
|
349
|
+
{ { "Transfer-Encoding", "chunked" }
|
350
|
+
, { "Vary", "*" }
|
351
|
+
, { "Content-Type", "text/plain" }
|
352
|
+
}
|
353
|
+
,.body= "hello world"
|
354
|
+
}
|
355
|
+
|
356
|
+
#define CHUNKED_W_BULLSHIT_AFTER_LENGTH 11
|
357
|
+
, {.name= "with bullshit after the length"
|
358
|
+
,.type= REQUEST
|
359
|
+
,.raw= "POST /chunked_w_bullshit_after_length HTTP/1.1\r\n"
|
360
|
+
"Transfer-Encoding: chunked\r\n"
|
361
|
+
"\r\n"
|
362
|
+
"5; ihatew3;whatthefuck=aretheseparametersfor\r\nhello\r\n"
|
363
|
+
"6; blahblah; blah\r\n world\r\n"
|
364
|
+
"0\r\n"
|
365
|
+
"\r\n"
|
366
|
+
,.should_keep_alive= TRUE
|
367
|
+
,.message_complete_on_eof= FALSE
|
368
|
+
,.http_major= 1
|
369
|
+
,.http_minor= 1
|
370
|
+
,.method= HTTP_POST
|
371
|
+
,.query_string= ""
|
372
|
+
,.fragment= ""
|
373
|
+
,.request_path= "/chunked_w_bullshit_after_length"
|
374
|
+
,.request_url= "/chunked_w_bullshit_after_length"
|
375
|
+
,.num_headers= 1
|
376
|
+
,.headers=
|
377
|
+
{ { "Transfer-Encoding", "chunked" }
|
378
|
+
}
|
379
|
+
,.body= "hello world"
|
380
|
+
}
|
381
|
+
|
382
|
+
#define WITH_QUOTES 12
|
383
|
+
, {.name= "with quotes"
|
384
|
+
,.type= REQUEST
|
385
|
+
,.raw= "GET /with_\"stupid\"_quotes?foo=\"bar\" HTTP/1.1\r\n\r\n"
|
386
|
+
,.should_keep_alive= TRUE
|
387
|
+
,.message_complete_on_eof= FALSE
|
388
|
+
,.http_major= 1
|
389
|
+
,.http_minor= 1
|
390
|
+
,.method= HTTP_GET
|
391
|
+
,.query_string= "foo=\"bar\""
|
392
|
+
,.fragment= ""
|
393
|
+
,.request_path= "/with_\"stupid\"_quotes"
|
394
|
+
,.request_url= "/with_\"stupid\"_quotes?foo=\"bar\""
|
395
|
+
,.num_headers= 0
|
396
|
+
,.headers= { }
|
397
|
+
,.body= ""
|
398
|
+
}
|
399
|
+
|
400
|
+
#define APACHEBENCH_GET 13
|
401
|
+
/* The server receiving this request SHOULD NOT wait for EOF
|
402
|
+
* to know that content-length == 0.
|
403
|
+
* How to represent this in a unit test? message_complete_on_eof
|
404
|
+
* Compare with NO_CONTENT_LENGTH_RESPONSE.
|
405
|
+
*/
|
406
|
+
, {.name = "apachebench get"
|
407
|
+
,.type= REQUEST
|
408
|
+
,.raw= "GET /test HTTP/1.0\r\n"
|
409
|
+
"Host: 0.0.0.0:5000\r\n"
|
410
|
+
"User-Agent: ApacheBench/2.3\r\n"
|
411
|
+
"Accept: */*\r\n\r\n"
|
412
|
+
,.should_keep_alive= FALSE
|
413
|
+
,.message_complete_on_eof= FALSE
|
414
|
+
,.http_major= 1
|
415
|
+
,.http_minor= 0
|
416
|
+
,.method= HTTP_GET
|
417
|
+
,.query_string= ""
|
418
|
+
,.fragment= ""
|
419
|
+
,.request_path= "/test"
|
420
|
+
,.request_url= "/test"
|
421
|
+
,.num_headers= 3
|
422
|
+
,.headers= { { "Host", "0.0.0.0:5000" }
|
423
|
+
, { "User-Agent", "ApacheBench/2.3" }
|
424
|
+
, { "Accept", "*/*" }
|
425
|
+
}
|
426
|
+
,.body= ""
|
427
|
+
}
|
428
|
+
|
429
|
+
, {.name= NULL } /* sentinel */
|
430
|
+
};
|
431
|
+
|
432
|
+
/* * R E S P O N S E S * */
|
433
|
+
const struct message responses[] =
|
434
|
+
#define GOOGLE_301 0
|
435
|
+
{ {.name= "google 301"
|
436
|
+
,.type= RESPONSE
|
437
|
+
,.raw= "HTTP/1.1 301 Moved Permanently\r\n"
|
438
|
+
"Location: http://www.google.com/\r\n"
|
439
|
+
"Content-Type: text/html; charset=UTF-8\r\n"
|
440
|
+
"Date: Sun, 26 Apr 2009 11:11:49 GMT\r\n"
|
441
|
+
"Expires: Tue, 26 May 2009 11:11:49 GMT\r\n"
|
442
|
+
"Cache-Control: public, max-age=2592000\r\n"
|
443
|
+
"Server: gws\r\n"
|
444
|
+
"Content-Length: 219\r\n"
|
445
|
+
"\r\n"
|
446
|
+
"<HTML><HEAD><meta http-equiv=\"content-type\" content=\"text/html;charset=utf-8\">\n"
|
447
|
+
"<TITLE>301 Moved</TITLE></HEAD><BODY>\n"
|
448
|
+
"<H1>301 Moved</H1>\n"
|
449
|
+
"The document has moved\n"
|
450
|
+
"<A HREF=\"http://www.google.com/\">here</A>.\r\n"
|
451
|
+
"</BODY></HTML>\r\n"
|
452
|
+
,.should_keep_alive= TRUE
|
453
|
+
,.message_complete_on_eof= FALSE
|
454
|
+
,.http_major= 1
|
455
|
+
,.http_minor= 1
|
456
|
+
,.status_code= 301
|
457
|
+
,.num_headers= 7
|
458
|
+
,.headers=
|
459
|
+
{ { "Location", "http://www.google.com/" }
|
460
|
+
, { "Content-Type", "text/html; charset=UTF-8" }
|
461
|
+
, { "Date", "Sun, 26 Apr 2009 11:11:49 GMT" }
|
462
|
+
, { "Expires", "Tue, 26 May 2009 11:11:49 GMT" }
|
463
|
+
, { "Cache-Control", "public, max-age=2592000" }
|
464
|
+
, { "Server", "gws" }
|
465
|
+
, { "Content-Length", "219" }
|
466
|
+
}
|
467
|
+
,.body= "<HTML><HEAD><meta http-equiv=\"content-type\" content=\"text/html;charset=utf-8\">\n"
|
468
|
+
"<TITLE>301 Moved</TITLE></HEAD><BODY>\n"
|
469
|
+
"<H1>301 Moved</H1>\n"
|
470
|
+
"The document has moved\n"
|
471
|
+
"<A HREF=\"http://www.google.com/\">here</A>.\r\n"
|
472
|
+
"</BODY></HTML>\r\n"
|
473
|
+
}
|
474
|
+
|
475
|
+
#define NO_CONTENT_LENGTH_RESPONSE 1
|
476
|
+
/* The client should wait for the server's EOF. That is, when content-length
|
477
|
+
* is not specified, and "Connection: close", the end of body is specified
|
478
|
+
* by the EOF.
|
479
|
+
* Compare with APACHEBENCH_GET
|
480
|
+
*/
|
481
|
+
, {.name= "no content-length response"
|
482
|
+
,.type= RESPONSE
|
483
|
+
,.raw= "HTTP/1.1 200 OK\r\n"
|
484
|
+
"Date: Tue, 04 Aug 2009 07:59:32 GMT\r\n"
|
485
|
+
"Server: Apache\r\n"
|
486
|
+
"X-Powered-By: Servlet/2.5 JSP/2.1\r\n"
|
487
|
+
"Content-Type: text/xml; charset=utf-8\r\n"
|
488
|
+
"Connection: close\r\n"
|
489
|
+
"\r\n"
|
490
|
+
"<?xml version=\"1.0\" encoding=\"UTF-8\"?>\n"
|
491
|
+
"<SOAP-ENV:Envelope xmlns:SOAP-ENV=\"http://schemas.xmlsoap.org/soap/envelope/\">\n"
|
492
|
+
" <SOAP-ENV:Body>\n"
|
493
|
+
" <SOAP-ENV:Fault>\n"
|
494
|
+
" <faultcode>SOAP-ENV:Client</faultcode>\n"
|
495
|
+
" <faultstring>Client Error</faultstring>\n"
|
496
|
+
" </SOAP-ENV:Fault>\n"
|
497
|
+
" </SOAP-ENV:Body>\n"
|
498
|
+
"</SOAP-ENV:Envelope>"
|
499
|
+
,.should_keep_alive= FALSE
|
500
|
+
,.message_complete_on_eof= TRUE
|
501
|
+
,.http_major= 1
|
502
|
+
,.http_minor= 1
|
503
|
+
,.status_code= 200
|
504
|
+
,.num_headers= 5
|
505
|
+
,.headers=
|
506
|
+
{ { "Date", "Tue, 04 Aug 2009 07:59:32 GMT" }
|
507
|
+
, { "Server", "Apache" }
|
508
|
+
, { "X-Powered-By", "Servlet/2.5 JSP/2.1" }
|
509
|
+
, { "Content-Type", "text/xml; charset=utf-8" }
|
510
|
+
, { "Connection", "close" }
|
511
|
+
}
|
512
|
+
,.body= "<?xml version=\"1.0\" encoding=\"UTF-8\"?>\n"
|
513
|
+
"<SOAP-ENV:Envelope xmlns:SOAP-ENV=\"http://schemas.xmlsoap.org/soap/envelope/\">\n"
|
514
|
+
" <SOAP-ENV:Body>\n"
|
515
|
+
" <SOAP-ENV:Fault>\n"
|
516
|
+
" <faultcode>SOAP-ENV:Client</faultcode>\n"
|
517
|
+
" <faultstring>Client Error</faultstring>\n"
|
518
|
+
" </SOAP-ENV:Fault>\n"
|
519
|
+
" </SOAP-ENV:Body>\n"
|
520
|
+
"</SOAP-ENV:Envelope>"
|
521
|
+
}
|
522
|
+
|
523
|
+
#define NO_HEADERS_NO_BODY_404 2
|
524
|
+
, {.name= "404 no headers no body"
|
525
|
+
,.type= RESPONSE
|
526
|
+
,.raw= "HTTP/1.1 404 Not Found\r\n\r\n"
|
527
|
+
,.should_keep_alive= TRUE
|
528
|
+
,.message_complete_on_eof= FALSE
|
529
|
+
,.http_major= 1
|
530
|
+
,.http_minor= 1
|
531
|
+
,.status_code= 404
|
532
|
+
,.num_headers= 0
|
533
|
+
,.headers= {}
|
534
|
+
,.body= ""
|
535
|
+
}
|
536
|
+
|
537
|
+
#define NO_REASON_PHRASE 3
|
538
|
+
, {.name= "301 no response phrase"
|
539
|
+
,.type= RESPONSE
|
540
|
+
,.raw= "HTTP/1.1 301\r\n\r\n"
|
541
|
+
,.should_keep_alive = TRUE
|
542
|
+
,.message_complete_on_eof= FALSE
|
543
|
+
,.http_major= 1
|
544
|
+
,.http_minor= 1
|
545
|
+
,.status_code= 301
|
546
|
+
,.num_headers= 0
|
547
|
+
,.headers= {}
|
548
|
+
,.body= ""
|
549
|
+
}
|
550
|
+
|
551
|
+
#define TRAILING_SPACE_ON_CHUNKED_BODY 4
|
552
|
+
, {.name="200 trailing space on chunked body"
|
553
|
+
,.type= RESPONSE
|
554
|
+
,.raw= "HTTP/1.1 200 OK\r\n"
|
555
|
+
"Content-Type: text/plain\r\n"
|
556
|
+
"Transfer-Encoding: chunked\r\n"
|
557
|
+
"\r\n"
|
558
|
+
"25 \r\n"
|
559
|
+
"This is the data in the first chunk\r\n"
|
560
|
+
"\r\n"
|
561
|
+
"1C\r\n"
|
562
|
+
"and this is the second one\r\n"
|
563
|
+
"\r\n"
|
564
|
+
"0 \r\n"
|
565
|
+
"\r\n"
|
566
|
+
,.should_keep_alive= TRUE
|
567
|
+
,.message_complete_on_eof= FALSE
|
568
|
+
,.http_major= 1
|
569
|
+
,.http_minor= 1
|
570
|
+
,.status_code= 200
|
571
|
+
,.num_headers= 2
|
572
|
+
,.headers=
|
573
|
+
{ {"Content-Type", "text/plain" }
|
574
|
+
, {"Transfer-Encoding", "chunked" }
|
575
|
+
}
|
576
|
+
,.body =
|
577
|
+
"This is the data in the first chunk\r\n"
|
578
|
+
"and this is the second one\r\n"
|
579
|
+
|
580
|
+
}
|
581
|
+
|
582
|
+
, {.name= NULL } /* sentinel */
|
583
|
+
};
|
584
|
+
|
585
|
+
int
|
586
|
+
request_path_cb (http_parser *p, const char *buf, size_t len)
|
587
|
+
{
|
588
|
+
assert(p == parser);
|
589
|
+
strncat(messages[num_messages].request_path, buf, len);
|
590
|
+
return 0;
|
591
|
+
}
|
592
|
+
|
593
|
+
int
|
594
|
+
request_url_cb (http_parser *p, const char *buf, size_t len)
|
595
|
+
{
|
596
|
+
assert(p == parser);
|
597
|
+
strncat(messages[num_messages].request_url, buf, len);
|
598
|
+
return 0;
|
599
|
+
}
|
600
|
+
|
601
|
+
int
|
602
|
+
query_string_cb (http_parser *p, const char *buf, size_t len)
|
603
|
+
{
|
604
|
+
assert(p == parser);
|
605
|
+
strncat(messages[num_messages].query_string, buf, len);
|
606
|
+
return 0;
|
607
|
+
}
|
608
|
+
|
609
|
+
int
|
610
|
+
fragment_cb (http_parser *p, const char *buf, size_t len)
|
611
|
+
{
|
612
|
+
assert(p == parser);
|
613
|
+
strncat(messages[num_messages].fragment, buf, len);
|
614
|
+
return 0;
|
615
|
+
}
|
616
|
+
|
617
|
+
int
|
618
|
+
header_field_cb (http_parser *p, const char *buf, size_t len)
|
619
|
+
{
|
620
|
+
assert(p == parser);
|
621
|
+
struct message *m = &messages[num_messages];
|
622
|
+
|
623
|
+
if (m->last_header_element != FIELD)
|
624
|
+
m->num_headers++;
|
625
|
+
|
626
|
+
strncat(m->headers[m->num_headers-1][0], buf, len);
|
627
|
+
|
628
|
+
m->last_header_element = FIELD;
|
629
|
+
|
630
|
+
return 0;
|
631
|
+
}
|
632
|
+
|
633
|
+
int
|
634
|
+
header_value_cb (http_parser *p, const char *buf, size_t len)
|
635
|
+
{
|
636
|
+
assert(p == parser);
|
637
|
+
struct message *m = &messages[num_messages];
|
638
|
+
|
639
|
+
strncat(m->headers[m->num_headers-1][1], buf, len);
|
640
|
+
|
641
|
+
m->last_header_element = VALUE;
|
642
|
+
|
643
|
+
return 0;
|
644
|
+
}
|
645
|
+
|
646
|
+
int
|
647
|
+
body_cb (http_parser *p, const char *buf, size_t len)
|
648
|
+
{
|
649
|
+
assert(p == parser);
|
650
|
+
strncat(messages[num_messages].body, buf, len);
|
651
|
+
// printf("body_cb: '%s'\n", requests[num_messages].body);
|
652
|
+
return 0;
|
653
|
+
}
|
654
|
+
|
655
|
+
int
|
656
|
+
message_begin_cb (http_parser *p)
|
657
|
+
{
|
658
|
+
assert(p == parser);
|
659
|
+
messages[num_messages].message_begin_cb_called = TRUE;
|
660
|
+
return 0;
|
661
|
+
}
|
662
|
+
|
663
|
+
int
|
664
|
+
headers_complete_cb (http_parser *p)
|
665
|
+
{
|
666
|
+
assert(p == parser);
|
667
|
+
messages[num_messages].method = parser->method;
|
668
|
+
messages[num_messages].status_code = parser->status_code;
|
669
|
+
messages[num_messages].http_major = parser->http_major;
|
670
|
+
messages[num_messages].http_minor = parser->http_minor;
|
671
|
+
messages[num_messages].headers_complete_cb_called = TRUE;
|
672
|
+
messages[num_messages].should_keep_alive = http_should_keep_alive(parser);
|
673
|
+
return 0;
|
674
|
+
}
|
675
|
+
|
676
|
+
int
|
677
|
+
message_complete_cb (http_parser *p)
|
678
|
+
{
|
679
|
+
assert(p == parser);
|
680
|
+
if (messages[num_messages].should_keep_alive != http_should_keep_alive(parser))
|
681
|
+
{
|
682
|
+
fprintf(stderr, "\n\n *** Error http_should_keep_alive() should have same "
|
683
|
+
"value in both on_message_complete and on_headers_complete "
|
684
|
+
"but it doesn't! ***\n\n");
|
685
|
+
assert(0);
|
686
|
+
exit(1);
|
687
|
+
}
|
688
|
+
messages[num_messages].message_complete_cb_called = TRUE;
|
689
|
+
|
690
|
+
messages[num_messages].message_complete_on_eof = currently_parsing_eof;
|
691
|
+
|
692
|
+
num_messages++;
|
693
|
+
return 0;
|
694
|
+
}
|
695
|
+
|
696
|
+
void
|
697
|
+
parser_init ()
|
698
|
+
{
|
699
|
+
num_messages = 0;
|
700
|
+
|
701
|
+
assert(parser == NULL);
|
702
|
+
|
703
|
+
parser = malloc(sizeof(http_parser));
|
704
|
+
|
705
|
+
http_parser_init(parser);
|
706
|
+
|
707
|
+
memset(&messages, 0, sizeof messages);
|
708
|
+
|
709
|
+
parser->on_message_begin = message_begin_cb;
|
710
|
+
parser->on_header_field = header_field_cb;
|
711
|
+
parser->on_header_value = header_value_cb;
|
712
|
+
parser->on_path = request_path_cb;
|
713
|
+
parser->on_url = request_url_cb;
|
714
|
+
parser->on_fragment = fragment_cb;
|
715
|
+
parser->on_query_string = query_string_cb;
|
716
|
+
parser->on_body = body_cb;
|
717
|
+
parser->on_headers_complete = headers_complete_cb;
|
718
|
+
parser->on_message_complete = message_complete_cb;
|
719
|
+
}
|
720
|
+
|
721
|
+
void
|
722
|
+
parser_free ()
|
723
|
+
{
|
724
|
+
assert(parser);
|
725
|
+
free(parser);
|
726
|
+
parser = NULL;
|
727
|
+
}
|
728
|
+
|
729
|
+
static inline int
|
730
|
+
check_str_eq (const struct message *m,
|
731
|
+
const char *prop,
|
732
|
+
const char *expected,
|
733
|
+
const char *found) {
|
734
|
+
if (0 != strcmp(expected, found)) {
|
735
|
+
printf("\n*** Error: %s in '%s' ***\n\n", prop, m->name);
|
736
|
+
printf("expected '%s'\n", expected);
|
737
|
+
printf(" found '%s'\n", found);
|
738
|
+
return 0;
|
739
|
+
}
|
740
|
+
return 1;
|
741
|
+
}
|
742
|
+
|
743
|
+
static inline int
|
744
|
+
check_num_eq (const struct message *m,
|
745
|
+
const char *prop,
|
746
|
+
int expected,
|
747
|
+
int found) {
|
748
|
+
if (expected != found) {
|
749
|
+
printf("\n*** Error: %s in '%s' ***\n\n", prop, m->name);
|
750
|
+
printf("expected %d\n", expected);
|
751
|
+
printf(" found %d\n", found);
|
752
|
+
return 0;
|
753
|
+
}
|
754
|
+
return 1;
|
755
|
+
}
|
756
|
+
|
757
|
+
#define MESSAGE_CHECK_STR_EQ(expected, found, prop) \
|
758
|
+
if (!check_str_eq(expected, #prop, expected->prop, found->prop)) return 0
|
759
|
+
|
760
|
+
#define MESSAGE_CHECK_NUM_EQ(expected, found, prop) \
|
761
|
+
if (!check_num_eq(expected, #prop, expected->prop, found->prop)) return 0
|
762
|
+
|
763
|
+
|
764
|
+
int
|
765
|
+
message_eq (int index, const struct message *expected)
|
766
|
+
{
|
767
|
+
int i;
|
768
|
+
struct message *m = &messages[index];
|
769
|
+
|
770
|
+
MESSAGE_CHECK_NUM_EQ(expected, m, http_major);
|
771
|
+
MESSAGE_CHECK_NUM_EQ(expected, m, http_minor);
|
772
|
+
|
773
|
+
if (expected->type == REQUEST) {
|
774
|
+
MESSAGE_CHECK_NUM_EQ(expected, m, method);
|
775
|
+
} else {
|
776
|
+
MESSAGE_CHECK_NUM_EQ(expected, m, status_code);
|
777
|
+
}
|
778
|
+
|
779
|
+
MESSAGE_CHECK_NUM_EQ(expected, m, should_keep_alive);
|
780
|
+
MESSAGE_CHECK_NUM_EQ(expected, m, message_complete_on_eof);
|
781
|
+
|
782
|
+
assert(m->message_begin_cb_called);
|
783
|
+
assert(m->headers_complete_cb_called);
|
784
|
+
assert(m->message_complete_cb_called);
|
785
|
+
|
786
|
+
|
787
|
+
MESSAGE_CHECK_STR_EQ(expected, m, request_path);
|
788
|
+
MESSAGE_CHECK_STR_EQ(expected, m, query_string);
|
789
|
+
MESSAGE_CHECK_STR_EQ(expected, m, fragment);
|
790
|
+
MESSAGE_CHECK_STR_EQ(expected, m, request_url);
|
791
|
+
MESSAGE_CHECK_STR_EQ(expected, m, body);
|
792
|
+
|
793
|
+
MESSAGE_CHECK_NUM_EQ(expected, m, num_headers);
|
794
|
+
|
795
|
+
int r;
|
796
|
+
for (i = 0; i < m->num_headers; i++) {
|
797
|
+
r = check_str_eq(expected, "header field", expected->headers[i][0], m->headers[i][0]);
|
798
|
+
if (!r) return 0;
|
799
|
+
r = check_str_eq(expected, "header value", expected->headers[i][1], m->headers[i][1]);
|
800
|
+
if (!r) return 0;
|
801
|
+
}
|
802
|
+
|
803
|
+
return 1;
|
804
|
+
}
|
805
|
+
|
806
|
+
static void
|
807
|
+
print_error (const char *raw, size_t error_location)
|
808
|
+
{
|
809
|
+
fprintf(stderr, "\n*** parse error ***\n\n");
|
810
|
+
|
811
|
+
int this_line = 0, char_len = 0;
|
812
|
+
size_t i, j, len = strlen(raw), error_location_line = 0;
|
813
|
+
for (i = 0; i < len; i++) {
|
814
|
+
if (i == error_location) this_line = 1;
|
815
|
+
switch (raw[i]) {
|
816
|
+
case '\r':
|
817
|
+
char_len = 2;
|
818
|
+
fprintf(stderr, "\\r");
|
819
|
+
break;
|
820
|
+
|
821
|
+
case '\n':
|
822
|
+
char_len = 2;
|
823
|
+
fprintf(stderr, "\\n\n");
|
824
|
+
|
825
|
+
if (this_line) goto print;
|
826
|
+
|
827
|
+
error_location_line = 0;
|
828
|
+
continue;
|
829
|
+
|
830
|
+
default:
|
831
|
+
char_len = 1;
|
832
|
+
fputc(raw[i], stderr);
|
833
|
+
break;
|
834
|
+
}
|
835
|
+
if (!this_line) error_location_line += char_len;
|
836
|
+
}
|
837
|
+
|
838
|
+
fprintf(stderr, "[eof]\n");
|
839
|
+
|
840
|
+
print:
|
841
|
+
for (j = 0; j < error_location_line; j++) {
|
842
|
+
fputc(' ', stderr);
|
843
|
+
}
|
844
|
+
fprintf(stderr, "^\n\nerror location: %d\n", error_location);
|
845
|
+
}
|
846
|
+
|
847
|
+
|
848
|
+
void
|
849
|
+
test_message (const struct message *message)
|
850
|
+
{
|
851
|
+
parser_init();
|
852
|
+
|
853
|
+
size_t read;
|
854
|
+
|
855
|
+
read = parse(message->type, message->raw, strlen(message->raw));
|
856
|
+
if (read != strlen(message->raw)) {
|
857
|
+
print_error(message->raw, read);
|
858
|
+
exit(1);
|
859
|
+
}
|
860
|
+
|
861
|
+
read = parse(message->type, NULL, 0);
|
862
|
+
if (read != 0) {
|
863
|
+
print_error(message->raw, read);
|
864
|
+
exit(1);
|
865
|
+
}
|
866
|
+
|
867
|
+
if (num_messages != 1) {
|
868
|
+
printf("\n*** num_messages != 1 after testing '%s' ***\n\n", message->name);
|
869
|
+
exit(1);
|
870
|
+
}
|
871
|
+
|
872
|
+
if(!message_eq(0, message)) exit(1);
|
873
|
+
|
874
|
+
parser_free();
|
875
|
+
}
|
876
|
+
|
877
|
+
void
|
878
|
+
test_error (const char *buf)
|
879
|
+
{
|
880
|
+
parser_init();
|
881
|
+
|
882
|
+
size_t parsed;
|
883
|
+
|
884
|
+
parsed = parse(REQUEST, buf, strlen(buf));
|
885
|
+
if (parsed != strlen(buf)) goto out;
|
886
|
+
parsed = parse(REQUEST, NULL, 0);
|
887
|
+
if (parsed != 0) goto out;
|
888
|
+
|
889
|
+
fprintf(stderr, "\n*** Error expected but none found ***\n\n%s", buf);
|
890
|
+
exit(1);
|
891
|
+
return;
|
892
|
+
|
893
|
+
out:
|
894
|
+
parser_free();
|
895
|
+
}
|
896
|
+
|
897
|
+
void
|
898
|
+
test_multiple3 (const struct message *r1, const struct message *r2, const struct message *r3)
|
899
|
+
{
|
900
|
+
char total[ strlen(r1->raw)
|
901
|
+
+ strlen(r2->raw)
|
902
|
+
+ strlen(r3->raw)
|
903
|
+
+ 1
|
904
|
+
];
|
905
|
+
total[0] = '\0';
|
906
|
+
|
907
|
+
strcat(total, r1->raw);
|
908
|
+
strcat(total, r2->raw);
|
909
|
+
strcat(total, r3->raw);
|
910
|
+
|
911
|
+
parser_init();
|
912
|
+
|
913
|
+
size_t read;
|
914
|
+
|
915
|
+
read = parse(r1->type, total, strlen(total));
|
916
|
+
if (read != strlen(total)) {
|
917
|
+
print_error(total, read);
|
918
|
+
exit(1);
|
919
|
+
}
|
920
|
+
|
921
|
+
read = parse(REQUEST, NULL, 0);
|
922
|
+
if (read != 0) {
|
923
|
+
print_error(total, read);
|
924
|
+
exit(1);
|
925
|
+
}
|
926
|
+
|
927
|
+
if (3 != num_messages) {
|
928
|
+
fprintf(stderr, "\n\n*** Parser didn't see 3 messages only %d *** \n", num_messages);
|
929
|
+
exit(1);
|
930
|
+
}
|
931
|
+
|
932
|
+
if (!message_eq(0, r1)) exit(1);
|
933
|
+
if (!message_eq(1, r2)) exit(1);
|
934
|
+
if (!message_eq(2, r3)) exit(1);
|
935
|
+
|
936
|
+
parser_free();
|
937
|
+
}
|
938
|
+
|
939
|
+
/* SCAN through every possible breaking to make sure the
|
940
|
+
* parser can handle getting the content in any chunks that
|
941
|
+
* might come from the socket
|
942
|
+
*/
|
943
|
+
void
|
944
|
+
test_scan (const struct message *r1, const struct message *r2, const struct message *r3)
|
945
|
+
{
|
946
|
+
char total[80*1024] = "\0";
|
947
|
+
char buf1[80*1024] = "\0";
|
948
|
+
char buf2[80*1024] = "\0";
|
949
|
+
char buf3[80*1024] = "\0";
|
950
|
+
|
951
|
+
strcat(total, r1->raw);
|
952
|
+
strcat(total, r2->raw);
|
953
|
+
strcat(total, r3->raw);
|
954
|
+
|
955
|
+
size_t read;
|
956
|
+
|
957
|
+
int total_len = strlen(total);
|
958
|
+
|
959
|
+
int total_ops = (total_len - 1) * (total_len - 2) / 2;
|
960
|
+
int ops = 0 ;
|
961
|
+
|
962
|
+
size_t buf1_len, buf2_len, buf3_len;
|
963
|
+
|
964
|
+
int i,j;
|
965
|
+
for (j = 2; j < total_len; j ++ ) {
|
966
|
+
for (i = 1; i < j; i ++ ) {
|
967
|
+
|
968
|
+
if (ops % 1000 == 0) {
|
969
|
+
printf("\b\b\b\b%3.0f%%", 100 * (float)ops /(float)total_ops);
|
970
|
+
fflush(stdout);
|
971
|
+
}
|
972
|
+
ops += 1;
|
973
|
+
|
974
|
+
parser_init();
|
975
|
+
|
976
|
+
buf1_len = i;
|
977
|
+
strncpy(buf1, total, buf1_len);
|
978
|
+
buf1[buf1_len] = 0;
|
979
|
+
|
980
|
+
buf2_len = j - i;
|
981
|
+
strncpy(buf2, total+i, buf2_len);
|
982
|
+
buf2[buf2_len] = 0;
|
983
|
+
|
984
|
+
buf3_len = total_len - j;
|
985
|
+
strncpy(buf3, total+j, buf3_len);
|
986
|
+
buf3[buf3_len] = 0;
|
987
|
+
|
988
|
+
read = parse(r1->type, buf1, buf1_len);
|
989
|
+
if (read != buf1_len) {
|
990
|
+
print_error(buf1, read);
|
991
|
+
goto error;
|
992
|
+
}
|
993
|
+
|
994
|
+
read = parse(r1->type, buf2, buf2_len);
|
995
|
+
if (read != buf2_len) {
|
996
|
+
print_error(buf2, read);
|
997
|
+
goto error;
|
998
|
+
}
|
999
|
+
|
1000
|
+
read = parse(r1->type, buf3, buf3_len);
|
1001
|
+
if (read != buf3_len) {
|
1002
|
+
print_error(buf3, read);
|
1003
|
+
goto error;
|
1004
|
+
}
|
1005
|
+
|
1006
|
+
parse(r1->type, NULL, 0);
|
1007
|
+
|
1008
|
+
if (3 != num_messages) {
|
1009
|
+
fprintf(stderr, "\n\nParser didn't see 3 messages only %d\n", num_messages);
|
1010
|
+
goto error;
|
1011
|
+
}
|
1012
|
+
|
1013
|
+
if (!message_eq(0, r1)) {
|
1014
|
+
fprintf(stderr, "\n\nError matching messages[0] in test_scan.\n");
|
1015
|
+
goto error;
|
1016
|
+
}
|
1017
|
+
|
1018
|
+
if (!message_eq(1, r2)) {
|
1019
|
+
fprintf(stderr, "\n\nError matching messages[1] in test_scan.\n");
|
1020
|
+
goto error;
|
1021
|
+
}
|
1022
|
+
|
1023
|
+
if (!message_eq(2, r3)) {
|
1024
|
+
fprintf(stderr, "\n\nError matching messages[2] in test_scan.\n");
|
1025
|
+
goto error;
|
1026
|
+
}
|
1027
|
+
|
1028
|
+
parser_free();
|
1029
|
+
}
|
1030
|
+
}
|
1031
|
+
puts("\b\b\b\b100%");
|
1032
|
+
return;
|
1033
|
+
|
1034
|
+
error:
|
1035
|
+
fprintf(stderr, "i=%d j=%d\n", i, j);
|
1036
|
+
fprintf(stderr, "buf1 (%d) %s\n\n", buf1_len, buf1);
|
1037
|
+
fprintf(stderr, "buf2 (%d) %s\n\n", buf2_len , buf2);
|
1038
|
+
fprintf(stderr, "buf3 (%d) %s\n", buf3_len, buf3);
|
1039
|
+
exit(1);
|
1040
|
+
}
|
1041
|
+
|
1042
|
+
int
|
1043
|
+
main (void)
|
1044
|
+
{
|
1045
|
+
parser = NULL;
|
1046
|
+
int i, j, k;
|
1047
|
+
int request_count;
|
1048
|
+
int response_count;
|
1049
|
+
|
1050
|
+
printf("sizeof(http_parser) = %d\n", sizeof(http_parser));
|
1051
|
+
|
1052
|
+
for (request_count = 0; requests[request_count].name; request_count++);
|
1053
|
+
for (response_count = 0; responses[response_count].name; response_count++);
|
1054
|
+
|
1055
|
+
//// RESPONSES
|
1056
|
+
|
1057
|
+
for (i = 0; i < response_count; i++) {
|
1058
|
+
test_message(&responses[i]);
|
1059
|
+
}
|
1060
|
+
|
1061
|
+
for (i = 0; i < response_count; i++) {
|
1062
|
+
if (!responses[i].should_keep_alive) continue;
|
1063
|
+
for (j = 0; j < response_count; j++) {
|
1064
|
+
if (!responses[j].should_keep_alive) continue;
|
1065
|
+
for (k = 0; k < response_count; k++) {
|
1066
|
+
test_multiple3(&responses[i], &responses[j], &responses[k]);
|
1067
|
+
}
|
1068
|
+
}
|
1069
|
+
}
|
1070
|
+
|
1071
|
+
printf("response scan 1/1 ");
|
1072
|
+
test_scan( &responses[TRAILING_SPACE_ON_CHUNKED_BODY]
|
1073
|
+
, &responses[NO_HEADERS_NO_BODY_404]
|
1074
|
+
, &responses[NO_REASON_PHRASE]
|
1075
|
+
);
|
1076
|
+
|
1077
|
+
puts("responses okay");
|
1078
|
+
|
1079
|
+
|
1080
|
+
/// REQUESTS
|
1081
|
+
|
1082
|
+
|
1083
|
+
test_error("hello world");
|
1084
|
+
test_error("GET / HTP/1.1\r\n\r\n");
|
1085
|
+
|
1086
|
+
const char *dumbfuck2 =
|
1087
|
+
"GET / HTTP/1.1\r\n"
|
1088
|
+
"X-SSL-Bullshit: -----BEGIN CERTIFICATE-----\r\n"
|
1089
|
+
"\tMIIFbTCCBFWgAwIBAgICH4cwDQYJKoZIhvcNAQEFBQAwcDELMAkGA1UEBhMCVUsx\r\n"
|
1090
|
+
"\tETAPBgNVBAoTCGVTY2llbmNlMRIwEAYDVQQLEwlBdXRob3JpdHkxCzAJBgNVBAMT\r\n"
|
1091
|
+
"\tAkNBMS0wKwYJKoZIhvcNAQkBFh5jYS1vcGVyYXRvckBncmlkLXN1cHBvcnQuYWMu\r\n"
|
1092
|
+
"\tdWswHhcNMDYwNzI3MTQxMzI4WhcNMDcwNzI3MTQxMzI4WjBbMQswCQYDVQQGEwJV\r\n"
|
1093
|
+
"\tSzERMA8GA1UEChMIZVNjaWVuY2UxEzARBgNVBAsTCk1hbmNoZXN0ZXIxCzAJBgNV\r\n"
|
1094
|
+
"\tBAcTmrsogriqMWLAk1DMRcwFQYDVQQDEw5taWNoYWVsIHBhcmQYJKoZIhvcNAQEB\r\n"
|
1095
|
+
"\tBQADggEPADCCAQoCggEBANPEQBgl1IaKdSS1TbhF3hEXSl72G9J+WC/1R64fAcEF\r\n"
|
1096
|
+
"\tW51rEyFYiIeZGx/BVzwXbeBoNUK41OK65sxGuflMo5gLflbwJtHBRIEKAfVVp3YR\r\n"
|
1097
|
+
"\tgW7cMA/s/XKgL1GEC7rQw8lIZT8RApukCGqOVHSi/F1SiFlPDxuDfmdiNzL31+sL\r\n"
|
1098
|
+
"\t0iwHDdNkGjy5pyBSB8Y79dsSJtCW/iaLB0/n8Sj7HgvvZJ7x0fr+RQjYOUUfrePP\r\n"
|
1099
|
+
"\tu2MSpFyf+9BbC/aXgaZuiCvSR+8Snv3xApQY+fULK/xY8h8Ua51iXoQ5jrgu2SqR\r\n"
|
1100
|
+
"\twgA7BUi3G8LFzMBl8FRCDYGUDy7M6QaHXx1ZWIPWNKsCAwEAAaOCAiQwggIgMAwG\r\n"
|
1101
|
+
"\tA1UdEwEB/wQCMAAwEQYJYIZIAYb4QgHTTPAQDAgWgMA4GA1UdDwEB/wQEAwID6DAs\r\n"
|
1102
|
+
"\tBglghkgBhvhCAQ0EHxYdVUsgZS1TY2llbmNlIFVzZXIgQ2VydGlmaWNhdGUwHQYD\r\n"
|
1103
|
+
"\tVR0OBBYEFDTt/sf9PeMaZDHkUIldrDYMNTBZMIGaBgNVHSMEgZIwgY+AFAI4qxGj\r\n"
|
1104
|
+
"\tloCLDdMVKwiljjDastqooXSkcjBwMQswCQYDVQQGEwJVSzERMA8GA1UEChMIZVNj\r\n"
|
1105
|
+
"\taWVuY2UxEjAQBgNVBAsTCUF1dGhvcml0eTELMAkGA1UEAxMCQ0ExLTArBgkqhkiG\r\n"
|
1106
|
+
"\t9w0BCQEWHmNhLW9wZXJhdG9yQGdyaWQtc3VwcG9ydC5hYy51a4IBADApBgNVHRIE\r\n"
|
1107
|
+
"\tIjAggR5jYS1vcGVyYXRvckBncmlkLXN1cHBvcnQuYWMudWswGQYDVR0gBBIwEDAO\r\n"
|
1108
|
+
"\tBgwrBgEEAdkvAQEBAQYwPQYJYIZIAYb4QgEEBDAWLmh0dHA6Ly9jYS5ncmlkLXN1\r\n"
|
1109
|
+
"\tcHBvcnQuYWMudmT4sopwqlBWsvcHViL2NybC9jYWNybC5jcmwwPQYJYIZIAYb4QgEDBDAWLmh0\r\n"
|
1110
|
+
"\tdHA6Ly9jYS5ncmlkLXN1cHBvcnQuYWMudWsvcHViL2NybC9jYWNybC5jcmwwPwYD\r\n"
|
1111
|
+
"\tVR0fBDgwNjA0oDKgMIYuaHR0cDovL2NhLmdyaWQt5hYy51ay9wdWIv\r\n"
|
1112
|
+
"\tY3JsL2NhY3JsLmNybDANBgkqhkiG9w0BAQUFAAOCAQEAS/U4iiooBENGW/Hwmmd3\r\n"
|
1113
|
+
"\tXCy6Zrt08YjKCzGNjorT98g8uGsqYjSxv/hmi0qlnlHs+k/3Iobc3LjS5AMYr5L8\r\n"
|
1114
|
+
"\tUO7OSkgFFlLHQyC9JzPfmLCAugvzEbyv4Olnsr8hbxF1MbKZoQxUZtMVu29wjfXk\r\n"
|
1115
|
+
"\thTeApBv7eaKCWpSp7MCbvgzm74izKhu3vlDk9w6qVrxePfGgpKPqfHiOoGhFnbTK\r\n"
|
1116
|
+
"\twTC6o2xq5y0qZ03JonF7OJspEd3I5zKY3E+ov7/ZhW6DqT8UFvsAdjvQbXyhV8Eu\r\n"
|
1117
|
+
"\tYhixw1aKEPzNjNowuIseVogKOLXxWI5vAi5HgXdS0/ES5gDGsABo4fqovUKlgop3\r\n"
|
1118
|
+
"\tRA==\r\n"
|
1119
|
+
"\t-----END CERTIFICATE-----\r\n"
|
1120
|
+
"\r\n";
|
1121
|
+
test_error(dumbfuck2);
|
1122
|
+
|
1123
|
+
#if 0
|
1124
|
+
// NOTE(Wed Nov 18 11:57:27 CET 2009) this seems okay. we just read body
|
1125
|
+
// until EOF.
|
1126
|
+
//
|
1127
|
+
// no content-length
|
1128
|
+
// error if there is a body without content length
|
1129
|
+
const char *bad_get_no_headers_no_body = "GET /bad_get_no_headers_no_body/world HTTP/1.1\r\n"
|
1130
|
+
"Accept: */*\r\n"
|
1131
|
+
"\r\n"
|
1132
|
+
"HELLO";
|
1133
|
+
test_error(bad_get_no_headers_no_body);
|
1134
|
+
#endif
|
1135
|
+
/* TODO sending junk and large headers gets rejected */
|
1136
|
+
|
1137
|
+
|
1138
|
+
/* check to make sure our predefined requests are okay */
|
1139
|
+
for (i = 0; requests[i].name; i++) {
|
1140
|
+
test_message(&requests[i]);
|
1141
|
+
}
|
1142
|
+
|
1143
|
+
|
1144
|
+
|
1145
|
+
for (i = 0; i < request_count; i++) {
|
1146
|
+
if (!requests[i].should_keep_alive) continue;
|
1147
|
+
for (j = 0; j < request_count; j++) {
|
1148
|
+
if (!requests[j].should_keep_alive) continue;
|
1149
|
+
for (k = 0; k < request_count; k++) {
|
1150
|
+
test_multiple3(&requests[i], &requests[j], &requests[k]);
|
1151
|
+
}
|
1152
|
+
}
|
1153
|
+
}
|
1154
|
+
|
1155
|
+
printf("request scan 1/3 ");
|
1156
|
+
test_scan( &requests[GET_NO_HEADERS_NO_BODY]
|
1157
|
+
, &requests[GET_ONE_HEADER_NO_BODY]
|
1158
|
+
, &requests[GET_NO_HEADERS_NO_BODY]
|
1159
|
+
);
|
1160
|
+
|
1161
|
+
printf("request scan 2/3 ");
|
1162
|
+
test_scan( &requests[POST_CHUNKED_ALL_YOUR_BASE]
|
1163
|
+
, &requests[POST_IDENTITY_BODY_WORLD]
|
1164
|
+
, &requests[GET_FUNKY_CONTENT_LENGTH]
|
1165
|
+
);
|
1166
|
+
|
1167
|
+
printf("request scan 3/3 ");
|
1168
|
+
test_scan( &requests[TWO_CHUNKS_MULT_ZERO_END]
|
1169
|
+
, &requests[CHUNKED_W_TRAILING_HEADERS]
|
1170
|
+
, &requests[CHUNKED_W_BULLSHIT_AFTER_LENGTH]
|
1171
|
+
);
|
1172
|
+
|
1173
|
+
puts("requests okay");
|
1174
|
+
|
1175
|
+
return 0;
|
1176
|
+
}
|