http_parser.rb 0.5.0
Sign up to get free protection for your applications and to get access to all the features.
- data/.gitignore +11 -0
- data/.gitmodules +6 -0
- data/README.md +45 -0
- data/Rakefile +6 -0
- data/bench/thin.rb +57 -0
- data/ext/ruby_http_parser/.gitignore +1 -0
- data/ext/ruby_http_parser/RubyHttpParserService.java +18 -0
- data/ext/ruby_http_parser/ext_help.h +18 -0
- data/ext/ruby_http_parser/extconf.rb +16 -0
- data/ext/ruby_http_parser/org/ruby_http_parser/RubyHttpParser.java +403 -0
- data/ext/ruby_http_parser/ruby_http_parser.c +474 -0
- data/ext/ruby_http_parser/vendor/.gitkeep +0 -0
- data/ext/ruby_http_parser/vendor/http-parser-java/CONTRIBUTIONS +4 -0
- data/ext/ruby_http_parser/vendor/http-parser-java/LICENSE-MIT +19 -0
- data/ext/ruby_http_parser/vendor/http-parser-java/README.md +171 -0
- data/ext/ruby_http_parser/vendor/http-parser-java/TODO +19 -0
- data/ext/ruby_http_parser/vendor/http-parser-java/compile +1 -0
- data/ext/ruby_http_parser/vendor/http-parser-java/http_parser.c +1590 -0
- data/ext/ruby_http_parser/vendor/http-parser-java/http_parser.h +167 -0
- data/ext/ruby_http_parser/vendor/http-parser-java/src/impl/http_parser/HTTPException.java +7 -0
- data/ext/ruby_http_parser/vendor/http-parser-java/src/impl/http_parser/HTTPMethod.java +90 -0
- data/ext/ruby_http_parser/vendor/http-parser-java/src/impl/http_parser/HTTPParser.java +31 -0
- data/ext/ruby_http_parser/vendor/http-parser-java/src/impl/http_parser/ParserType.java +13 -0
- data/ext/ruby_http_parser/vendor/http-parser-java/src/impl/http_parser/lolevel/HTTPCallback.java +5 -0
- data/ext/ruby_http_parser/vendor/http-parser-java/src/impl/http_parser/lolevel/HTTPDataCallback.java +25 -0
- data/ext/ruby_http_parser/vendor/http-parser-java/src/impl/http_parser/lolevel/HTTPErrorCallback.java +7 -0
- data/ext/ruby_http_parser/vendor/http-parser-java/src/impl/http_parser/lolevel/HTTPParser.java +1894 -0
- data/ext/ruby_http_parser/vendor/http-parser-java/src/impl/http_parser/lolevel/ParserSettings.java +78 -0
- data/ext/ruby_http_parser/vendor/http-parser-java/src/impl/http_parser/lolevel/Util.java +112 -0
- data/ext/ruby_http_parser/vendor/http-parser-java/src/test/http_parser/lolevel/TestLoaderNG.java +487 -0
- data/ext/ruby_http_parser/vendor/http-parser-java/src/test/http_parser/lolevel/UnitTest.java +115 -0
- data/ext/ruby_http_parser/vendor/http-parser-java/test.c +1865 -0
- data/ext/ruby_http_parser/vendor/http-parser-java/test_permutations +1 -0
- data/ext/ruby_http_parser/vendor/http-parser-java/test_unit +1 -0
- data/ext/ruby_http_parser/vendor/http-parser-java/tests.dumped +539 -0
- data/ext/ruby_http_parser/vendor/http-parser-java/tools/byte_constants.rb +6 -0
- data/ext/ruby_http_parser/vendor/http-parser-java/tools/const_char.rb +13 -0
- data/ext/ruby_http_parser/vendor/http-parser-java/tools/lowcase.rb +15 -0
- data/ext/ruby_http_parser/vendor/http-parser-java/tools/parse_tests.rb +33 -0
- data/ext/ruby_http_parser/vendor/http-parser/CONTRIBUTIONS +4 -0
- data/ext/ruby_http_parser/vendor/http-parser/LICENSE-MIT +19 -0
- data/ext/ruby_http_parser/vendor/http-parser/README.md +171 -0
- data/ext/ruby_http_parser/vendor/http-parser/http_parser.c +1590 -0
- data/ext/ruby_http_parser/vendor/http-parser/http_parser.h +167 -0
- data/ext/ruby_http_parser/vendor/http-parser/test.c +1755 -0
- data/http_parser.rb.gemspec +15 -0
- data/lib/http/parser.rb +1 -0
- data/lib/http_parser.rb +4 -0
- data/spec/parser_spec.rb +187 -0
- data/spec/spec_helper.rb +2 -0
- data/spec/support/requests.json +381 -0
- data/spec/support/responses.json +186 -0
- data/tasks/compile.rake +39 -0
- data/tasks/spec.rake +5 -0
- data/tasks/submodules.rake +7 -0
- metadata +121 -0
@@ -0,0 +1,167 @@
|
|
1
|
+
/* Copyright 2009,2010 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
|
+
|
28
|
+
#include <sys/types.h>
|
29
|
+
#include <stdint.h>
|
30
|
+
|
31
|
+
#ifdef _WIN32
|
32
|
+
typedef unsigned int size_t;
|
33
|
+
typedef int ssize_t;
|
34
|
+
#endif
|
35
|
+
|
36
|
+
/* Compile with -DHTTP_PARSER_STRICT=0 to make less checks, but run
|
37
|
+
* faster
|
38
|
+
*/
|
39
|
+
#ifndef HTTP_PARSER_STRICT
|
40
|
+
# define HTTP_PARSER_STRICT 1
|
41
|
+
#else
|
42
|
+
# define HTTP_PARSER_STRICT 0
|
43
|
+
#endif
|
44
|
+
|
45
|
+
|
46
|
+
/* Maximium header size allowed */
|
47
|
+
#define HTTP_MAX_HEADER_SIZE (80*1024)
|
48
|
+
|
49
|
+
|
50
|
+
typedef struct http_parser http_parser;
|
51
|
+
typedef struct http_parser_settings http_parser_settings;
|
52
|
+
|
53
|
+
|
54
|
+
/* Callbacks should return non-zero to indicate an error. The parser will
|
55
|
+
* then halt execution.
|
56
|
+
*
|
57
|
+
* The one exception is on_headers_complete. In a HTTP_RESPONSE parser
|
58
|
+
* returning '1' from on_headers_complete will tell the parser that it
|
59
|
+
* should not expect a body. This is used when receiving a response to a
|
60
|
+
* HEAD request which may contain 'Content-Length' or 'Transfer-Encoding:
|
61
|
+
* chunked' headers that indicate the presence of a body.
|
62
|
+
*
|
63
|
+
* http_data_cb does not return data chunks. It will be call arbitrarally
|
64
|
+
* many times for each string. E.G. you might get 10 callbacks for "on_path"
|
65
|
+
* each providing just a few characters more data.
|
66
|
+
*/
|
67
|
+
typedef int (*http_data_cb) (http_parser*, const char *at, size_t length);
|
68
|
+
typedef int (*http_cb) (http_parser*);
|
69
|
+
|
70
|
+
|
71
|
+
/* Request Methods */
|
72
|
+
enum http_method
|
73
|
+
{ HTTP_DELETE = 0
|
74
|
+
, HTTP_GET
|
75
|
+
, HTTP_HEAD
|
76
|
+
, HTTP_POST
|
77
|
+
, HTTP_PUT
|
78
|
+
/* pathological */
|
79
|
+
, HTTP_CONNECT
|
80
|
+
, HTTP_OPTIONS
|
81
|
+
, HTTP_TRACE
|
82
|
+
/* webdav */
|
83
|
+
, HTTP_COPY
|
84
|
+
, HTTP_LOCK
|
85
|
+
, HTTP_MKCOL
|
86
|
+
, HTTP_MOVE
|
87
|
+
, HTTP_PROPFIND
|
88
|
+
, HTTP_PROPPATCH
|
89
|
+
, HTTP_UNLOCK
|
90
|
+
/* subversion */
|
91
|
+
, HTTP_REPORT
|
92
|
+
, HTTP_MKACTIVITY
|
93
|
+
, HTTP_CHECKOUT
|
94
|
+
, HTTP_MERGE
|
95
|
+
};
|
96
|
+
|
97
|
+
|
98
|
+
enum http_parser_type { HTTP_REQUEST, HTTP_RESPONSE, HTTP_BOTH };
|
99
|
+
|
100
|
+
|
101
|
+
struct http_parser {
|
102
|
+
/** PRIVATE **/
|
103
|
+
unsigned char type : 2;
|
104
|
+
unsigned char flags : 6;
|
105
|
+
unsigned char state;
|
106
|
+
unsigned char header_state;
|
107
|
+
unsigned char index;
|
108
|
+
|
109
|
+
uint32_t nread;
|
110
|
+
int64_t content_length;
|
111
|
+
|
112
|
+
/** READ-ONLY **/
|
113
|
+
unsigned short http_major;
|
114
|
+
unsigned short http_minor;
|
115
|
+
unsigned short status_code; /* responses only */
|
116
|
+
unsigned char method; /* requests only */
|
117
|
+
|
118
|
+
/* 1 = Upgrade header was present and the parser has exited because of that.
|
119
|
+
* 0 = No upgrade header present.
|
120
|
+
* Should be checked when http_parser_execute() returns in addition to
|
121
|
+
* error checking.
|
122
|
+
*/
|
123
|
+
char upgrade;
|
124
|
+
|
125
|
+
/** PUBLIC **/
|
126
|
+
void *data; /* A pointer to get hook to the "connection" or "socket" object */
|
127
|
+
};
|
128
|
+
|
129
|
+
|
130
|
+
struct http_parser_settings {
|
131
|
+
http_cb on_message_begin;
|
132
|
+
http_data_cb on_path;
|
133
|
+
http_data_cb on_query_string;
|
134
|
+
http_data_cb on_url;
|
135
|
+
http_data_cb on_fragment;
|
136
|
+
http_data_cb on_header_field;
|
137
|
+
http_data_cb on_header_value;
|
138
|
+
http_cb on_headers_complete;
|
139
|
+
http_data_cb on_body;
|
140
|
+
http_cb on_message_complete;
|
141
|
+
};
|
142
|
+
|
143
|
+
|
144
|
+
void http_parser_init(http_parser *parser, enum http_parser_type type);
|
145
|
+
|
146
|
+
|
147
|
+
size_t http_parser_execute(http_parser *parser,
|
148
|
+
const http_parser_settings *settings,
|
149
|
+
const char *data,
|
150
|
+
size_t len);
|
151
|
+
|
152
|
+
|
153
|
+
/* If http_should_keep_alive() in the on_headers_complete or
|
154
|
+
* on_message_complete callback returns true, then this will be should be
|
155
|
+
* the last message on the connection.
|
156
|
+
* If you are the server, respond with the "Connection: close" header.
|
157
|
+
* If you are the client, close the connection.
|
158
|
+
*/
|
159
|
+
int http_should_keep_alive(http_parser *parser);
|
160
|
+
|
161
|
+
/* Returns a string version of the HTTP method. */
|
162
|
+
const char *http_method_str(enum http_method);
|
163
|
+
|
164
|
+
#ifdef __cplusplus
|
165
|
+
}
|
166
|
+
#endif
|
167
|
+
#endif
|
@@ -0,0 +1,90 @@
|
|
1
|
+
package http_parser;
|
2
|
+
|
3
|
+
import java.nio.charset.Charset;
|
4
|
+
|
5
|
+
public enum HTTPMethod {
|
6
|
+
HTTP_DELETE("DELETE")// = 0
|
7
|
+
, HTTP_GET("GET")
|
8
|
+
, HTTP_HEAD("HEAD")
|
9
|
+
, HTTP_POST("POST")
|
10
|
+
, HTTP_PUT("PUT")
|
11
|
+
/* pathological */
|
12
|
+
, HTTP_CONNECT("CONNECT")
|
13
|
+
, HTTP_OPTIONS("OPTIONS")
|
14
|
+
, HTTP_TRACE("TRACE")
|
15
|
+
/* webdav */
|
16
|
+
, HTTP_COPY("COPY")
|
17
|
+
, HTTP_LOCK("LOCK")
|
18
|
+
, HTTP_MKCOL("MKCOL")
|
19
|
+
, HTTP_MOVE("MOVE")
|
20
|
+
, HTTP_PROPFIND("PROPFIND")
|
21
|
+
, HTTP_PROPPATCH("PROPPATCH")
|
22
|
+
, HTTP_UNLOCK("UNLOCK")
|
23
|
+
, HTTP_REPORT("REPORT")
|
24
|
+
, HTTP_MKACTIVITY("MKACTIVITY")
|
25
|
+
, HTTP_CHECKOUT("CHECKOUT")
|
26
|
+
, HTTP_MERGE("MERGE");
|
27
|
+
|
28
|
+
private static Charset ASCII;
|
29
|
+
static {
|
30
|
+
ASCII = Charset.forName("US-ASCII");;
|
31
|
+
}
|
32
|
+
public byte[] bytes;
|
33
|
+
|
34
|
+
HTTPMethod(String name) {
|
35
|
+
// good grief, Charlie Brown, the following is necessary because
|
36
|
+
// java is retarded:
|
37
|
+
// illegal reference to static field from initializer
|
38
|
+
// this.bytes = name.getBytes(ASCII);
|
39
|
+
// yet it's not illegal to reference static fields from
|
40
|
+
// methods called from initializer.
|
41
|
+
init(name);
|
42
|
+
}
|
43
|
+
public static HTTPMethod parse(String s) {
|
44
|
+
if ("HTTP_DELETE".equalsIgnoreCase(s)) {return HTTP_DELETE;}
|
45
|
+
else if ("DELETE".equalsIgnoreCase(s)) {return HTTP_DELETE;}
|
46
|
+
else if ("HTTP_GET".equalsIgnoreCase(s)) {return HTTP_GET;}
|
47
|
+
else if ("GET".equalsIgnoreCase(s)) {return HTTP_GET;}
|
48
|
+
else if ("HTTP_HEAD".equalsIgnoreCase(s)) {return HTTP_HEAD;}
|
49
|
+
else if ("HEAD".equalsIgnoreCase(s)) {return HTTP_HEAD;}
|
50
|
+
else if ("HTTP_POST".equalsIgnoreCase(s)) {return HTTP_POST;}
|
51
|
+
else if ("POST".equalsIgnoreCase(s)) {return HTTP_POST;}
|
52
|
+
else if ("HTTP_PUT".equalsIgnoreCase(s)) {return HTTP_PUT;}
|
53
|
+
else if ("PUT".equalsIgnoreCase(s)) {return HTTP_PUT;}
|
54
|
+
else if ("HTTP_CONNECT".equalsIgnoreCase(s)) {return HTTP_CONNECT;}
|
55
|
+
else if ("CONNECT".equalsIgnoreCase(s)) {return HTTP_CONNECT;}
|
56
|
+
else if ("HTTP_OPTIONS".equalsIgnoreCase(s)) {return HTTP_OPTIONS;}
|
57
|
+
else if ("OPTIONS".equalsIgnoreCase(s)) {return HTTP_OPTIONS;}
|
58
|
+
else if ("HTTP_TRACE".equalsIgnoreCase(s)) {return HTTP_TRACE;}
|
59
|
+
else if ("TRACE".equalsIgnoreCase(s)) {return HTTP_TRACE;}
|
60
|
+
else if ("HTTP_COPY".equalsIgnoreCase(s)) {return HTTP_COPY;}
|
61
|
+
else if ("COPY".equalsIgnoreCase(s)) {return HTTP_COPY;}
|
62
|
+
else if ("HTTP_LOCK".equalsIgnoreCase(s)) {return HTTP_LOCK;}
|
63
|
+
else if ("LOCK".equalsIgnoreCase(s)) {return HTTP_LOCK;}
|
64
|
+
else if ("HTTP_MKCOL".equalsIgnoreCase(s)) {return HTTP_MKCOL;}
|
65
|
+
else if ("MKCOL".equalsIgnoreCase(s)) {return HTTP_MKCOL;}
|
66
|
+
else if ("HTTP_MOVE".equalsIgnoreCase(s)) {return HTTP_MOVE;}
|
67
|
+
else if ("MOVE".equalsIgnoreCase(s)) {return HTTP_MOVE;}
|
68
|
+
else if ("HTTP_PROPFIND".equalsIgnoreCase(s)){return HTTP_PROPFIND;}
|
69
|
+
else if ("PROPFIND".equalsIgnoreCase(s)) {return HTTP_PROPFIND;}
|
70
|
+
else if ("HTTP_PROPPATCH".equalsIgnoreCase(s)){return HTTP_PROPPATCH;}
|
71
|
+
else if ("PROPPATCH".equalsIgnoreCase(s)) {return HTTP_PROPPATCH;}
|
72
|
+
else if ("HTTP_UNLOCK".equalsIgnoreCase(s)) {return HTTP_UNLOCK;}
|
73
|
+
else if ("UNLOCK".equalsIgnoreCase(s)) {return HTTP_UNLOCK;}
|
74
|
+
else if ("HTTP_REPORT".equalsIgnoreCase(s)) {return HTTP_REPORT;}
|
75
|
+
else if ("REPORT".equalsIgnoreCase(s)){return HTTP_REPORT;}
|
76
|
+
else if ("HTTP_MKACTIVITY".equalsIgnoreCase(s)) {return HTTP_MKACTIVITY;}
|
77
|
+
else if ("MKACTIVITY".equalsIgnoreCase(s)){return HTTP_MKACTIVITY;}
|
78
|
+
else if ("HTTP_CHECKOUT".equalsIgnoreCase(s)) {return HTTP_CHECKOUT;}
|
79
|
+
else if ("CHECKOUT".equalsIgnoreCase(s)){return HTTP_CHECKOUT;}
|
80
|
+
else if ("HTTP_MERGE".equalsIgnoreCase(s)) {return HTTP_MERGE;}
|
81
|
+
else if ("MERGE".equalsIgnoreCase(s)){return HTTP_MERGE;}
|
82
|
+
|
83
|
+
else {return null;}
|
84
|
+
}
|
85
|
+
void init (String name) {
|
86
|
+
ASCII = null == ASCII ? Charset.forName("US-ASCII") : ASCII;
|
87
|
+
this.bytes = name.getBytes(ASCII);
|
88
|
+
}
|
89
|
+
|
90
|
+
}
|
@@ -0,0 +1,31 @@
|
|
1
|
+
package http_parser;
|
2
|
+
|
3
|
+
public class HTTPParser extends http_parser.lolevel.HTTPParser {
|
4
|
+
|
5
|
+
public HTTPParser() { super(); }
|
6
|
+
public HTTPParser(ParserType type) { super(type); }
|
7
|
+
|
8
|
+
public int getMajor() {
|
9
|
+
return super.http_major;
|
10
|
+
}
|
11
|
+
|
12
|
+
public int getMinor() {
|
13
|
+
return super.http_minor;
|
14
|
+
}
|
15
|
+
|
16
|
+
public int getStatusCode() {
|
17
|
+
return super.status_code;
|
18
|
+
}
|
19
|
+
|
20
|
+
public HTTPMethod getHTTPMethod() {
|
21
|
+
return super.method;
|
22
|
+
}
|
23
|
+
|
24
|
+
public boolean getUpgrade() {
|
25
|
+
return super.upgrade;
|
26
|
+
}
|
27
|
+
|
28
|
+
public boolean shouldKeepAlive() {
|
29
|
+
return super.http_should_keep_alive();
|
30
|
+
}
|
31
|
+
}
|
@@ -0,0 +1,13 @@
|
|
1
|
+
package http_parser;
|
2
|
+
|
3
|
+
public enum ParserType {
|
4
|
+
HTTP_REQUEST, HTTP_RESPONSE, HTTP_BOTH;
|
5
|
+
|
6
|
+
public static ParserType parse(String s) {
|
7
|
+
if ("HTTP_REQUEST".equalsIgnoreCase(s)) { return HTTP_REQUEST; }
|
8
|
+
else if ("HTTP_RESPONSE".equalsIgnoreCase(s)) { return HTTP_RESPONSE; }
|
9
|
+
else if ("HTTP_BOTH".equalsIgnoreCase(s)) { return HTTP_BOTH; }
|
10
|
+
else { return null; }
|
11
|
+
}
|
12
|
+
}
|
13
|
+
|
data/ext/ruby_http_parser/vendor/http-parser-java/src/impl/http_parser/lolevel/HTTPDataCallback.java
ADDED
@@ -0,0 +1,25 @@
|
|
1
|
+
package http_parser.lolevel;
|
2
|
+
|
3
|
+
import java.nio.ByteBuffer;
|
4
|
+
|
5
|
+
public interface HTTPDataCallback {
|
6
|
+
/*
|
7
|
+
very raw and extremly foolhardy! DANGER!
|
8
|
+
The whole Buffer concept is difficult enough to grasp as it is,
|
9
|
+
we pass in a buffer with an arbitrary position.
|
10
|
+
|
11
|
+
The interesting data is located at position pos and is len
|
12
|
+
bytes long.
|
13
|
+
|
14
|
+
The contract of this callback is that the buffer is
|
15
|
+
returned in the state that it was passed in, so implementing
|
16
|
+
this require good citizenship, you'll need to remember the current
|
17
|
+
position, change the position to get at the data you're interested
|
18
|
+
in and then set the position back to how you found it...
|
19
|
+
|
20
|
+
//TODO: there should be an abstract implementation that implements
|
21
|
+
cb as described above, marks it final an provides a new callback
|
22
|
+
with signature cb(byte[], int, int)
|
23
|
+
*/
|
24
|
+
public int cb(HTTPParser p, ByteBuffer buf, int pos, int len);
|
25
|
+
}
|
data/ext/ruby_http_parser/vendor/http-parser-java/src/impl/http_parser/lolevel/HTTPParser.java
ADDED
@@ -0,0 +1,1894 @@
|
|
1
|
+
package http_parser.lolevel;
|
2
|
+
|
3
|
+
import java.nio.ByteBuffer;
|
4
|
+
import http_parser.HTTPException;
|
5
|
+
import http_parser.HTTPMethod;
|
6
|
+
import http_parser.ParserType;
|
7
|
+
import static http_parser.lolevel.HTTPParser.C.*;
|
8
|
+
import static http_parser.lolevel.HTTPParser.State.*;
|
9
|
+
|
10
|
+
public class HTTPParser {
|
11
|
+
/* lots of unsigned chars here, not sure what
|
12
|
+
to about them, `bytes` in java suck... */
|
13
|
+
|
14
|
+
ParserType type;
|
15
|
+
State state;
|
16
|
+
HState header_state;
|
17
|
+
boolean strict;
|
18
|
+
|
19
|
+
int index;
|
20
|
+
int flags; // TODO
|
21
|
+
|
22
|
+
int nread;
|
23
|
+
int content_length;
|
24
|
+
|
25
|
+
|
26
|
+
/** READ-ONLY **/
|
27
|
+
protected int http_major;
|
28
|
+
protected int http_minor;
|
29
|
+
protected int status_code; /* responses only */
|
30
|
+
protected HTTPMethod method; /* requests only */
|
31
|
+
|
32
|
+
/* true = Upgrade header was present and the parser has exited because of that.
|
33
|
+
* false = No upgrade header present.
|
34
|
+
* Should be checked when http_parser_execute() returns in addition to
|
35
|
+
* error checking.
|
36
|
+
*/
|
37
|
+
protected boolean upgrade;
|
38
|
+
|
39
|
+
/** PUBLIC **/
|
40
|
+
// TODO : this is used in c to maintain application state.
|
41
|
+
// is this even necessary? we have state in java ?
|
42
|
+
// consider
|
43
|
+
// Object data; /* A pointer to get hook to the "connection" or "socket" object */
|
44
|
+
|
45
|
+
|
46
|
+
/*
|
47
|
+
* technically we could combine all of these (except for url_mark) into one
|
48
|
+
* variable, saving stack space, but it seems more clear to have them
|
49
|
+
* separated.
|
50
|
+
*/
|
51
|
+
int header_field_mark = -1;
|
52
|
+
int header_value_mark = -1;
|
53
|
+
int fragment_mark = -1;
|
54
|
+
int query_string_mark = -1;
|
55
|
+
int path_mark = -1;
|
56
|
+
int url_mark = -1;
|
57
|
+
|
58
|
+
/**
|
59
|
+
* Construct a Parser for ParserType.HTTP_BOTH, meaning it
|
60
|
+
* determines whether it's parsing a request or a response.
|
61
|
+
*/
|
62
|
+
public HTTPParser() {
|
63
|
+
this(ParserType.HTTP_BOTH);
|
64
|
+
}
|
65
|
+
|
66
|
+
/**
|
67
|
+
* Construct a Parser and initialise it to parse either
|
68
|
+
* requests or responses.
|
69
|
+
*/
|
70
|
+
public HTTPParser(ParserType type) {
|
71
|
+
this.type = type;
|
72
|
+
switch(type) {
|
73
|
+
case HTTP_REQUEST:
|
74
|
+
this.state = State.start_req;
|
75
|
+
break;
|
76
|
+
case HTTP_RESPONSE:
|
77
|
+
this.state = State.start_res;
|
78
|
+
break;
|
79
|
+
case HTTP_BOTH:
|
80
|
+
this.state = State.start_res_or_res;
|
81
|
+
break;
|
82
|
+
default:
|
83
|
+
throw new HTTPException("can't happen, invalid ParserType enum");
|
84
|
+
}
|
85
|
+
}
|
86
|
+
|
87
|
+
/*
|
88
|
+
* Utility to facilitate System.out.println style debugging (the way god intended)
|
89
|
+
*/
|
90
|
+
static void p(Object o) {System.out.println(o);}
|
91
|
+
|
92
|
+
|
93
|
+
/** Execute the parser with the currently available data contained in
|
94
|
+
* the buffer. The buffers position() and limit() need to be set
|
95
|
+
* correctly (obviously) and a will be updated approriately when the
|
96
|
+
* method returns to reflect the consumed data.
|
97
|
+
*/
|
98
|
+
public void execute(ParserSettings settings, ByteBuffer data) {
|
99
|
+
|
100
|
+
int p = data.position();
|
101
|
+
int p_err = p; // this is used for pretty printing errors.
|
102
|
+
|
103
|
+
// In case the headers don't provide information about the content
|
104
|
+
// length, `execute` needs to be called with an empty buffer to
|
105
|
+
// indicate that all the data has been send be the client/server,
|
106
|
+
// else there is no way of knowing the message is complete.
|
107
|
+
int len = (data.limit() - data.position());
|
108
|
+
if (0 == len) {
|
109
|
+
if (State.body_identity_eof == state) {
|
110
|
+
settings.call_on_message_complete(this);
|
111
|
+
}
|
112
|
+
}
|
113
|
+
|
114
|
+
|
115
|
+
// in case the _previous_ call to the parser only has data to get to
|
116
|
+
// the middle of certain fields, we need to update marks to point at
|
117
|
+
// the beginning of the current buffer.
|
118
|
+
switch (state) {
|
119
|
+
case header_field:
|
120
|
+
header_field_mark = p;
|
121
|
+
break;
|
122
|
+
case header_value:
|
123
|
+
header_value_mark = p;
|
124
|
+
break;
|
125
|
+
case req_fragment:
|
126
|
+
fragment_mark = p;
|
127
|
+
url_mark = p;
|
128
|
+
break;
|
129
|
+
case req_query_string:
|
130
|
+
query_string_mark = p;
|
131
|
+
url_mark = p;
|
132
|
+
break;
|
133
|
+
case req_path:
|
134
|
+
path_mark = p;
|
135
|
+
|
136
|
+
case req_host:
|
137
|
+
case req_schema:
|
138
|
+
case req_schema_slash:
|
139
|
+
case req_schema_slash_slash:
|
140
|
+
case req_port:
|
141
|
+
case req_query_string_start:
|
142
|
+
case req_fragment_start:
|
143
|
+
url_mark = p;
|
144
|
+
break;
|
145
|
+
}
|
146
|
+
|
147
|
+
// this is where the work gets done, traverse the available data...
|
148
|
+
while (data.position() != data.limit()) {
|
149
|
+
|
150
|
+
p = data.position();
|
151
|
+
int pe = data.limit();
|
152
|
+
|
153
|
+
byte ch = data.get(); // the current character to process.
|
154
|
+
byte c = -1; // utility variably used for up- and downcasing etc.
|
155
|
+
int to_read = 0; // used to keep track of how much of body, etc. is left to read
|
156
|
+
|
157
|
+
if (parsing_header(state)) {
|
158
|
+
++nread;
|
159
|
+
if (nread > HTTP_MAX_HEADER_SIZE) {
|
160
|
+
settings.call_on_error(this, "possible buffer overflow", data, p_err);
|
161
|
+
}
|
162
|
+
}
|
163
|
+
|
164
|
+
switch (state) {
|
165
|
+
/*
|
166
|
+
* this state is used after a 'Connection: close' message
|
167
|
+
* the parser will error out if it reads another message
|
168
|
+
*/
|
169
|
+
case dead:
|
170
|
+
settings.call_on_error(this, "Connection already closed", data, p_err);
|
171
|
+
|
172
|
+
|
173
|
+
|
174
|
+
case start_res_or_res:
|
175
|
+
if (CR == ch || LF == ch){
|
176
|
+
break;
|
177
|
+
}
|
178
|
+
flags = 0;
|
179
|
+
content_length = -1;
|
180
|
+
|
181
|
+
settings.call_on_message_begin(this);
|
182
|
+
|
183
|
+
if (H == ch) {
|
184
|
+
state = State.res_or_resp_H;
|
185
|
+
} else {
|
186
|
+
type = ParserType.HTTP_REQUEST;
|
187
|
+
method = start_req_method_assign(ch);
|
188
|
+
if (null == method) {
|
189
|
+
settings.call_on_error(this, "invalid method", data, p_err);
|
190
|
+
}
|
191
|
+
index = 1;
|
192
|
+
state = State.req_method;
|
193
|
+
}
|
194
|
+
break;
|
195
|
+
|
196
|
+
|
197
|
+
|
198
|
+
case res_or_resp_H:
|
199
|
+
if (T == ch) {
|
200
|
+
type = ParserType.HTTP_RESPONSE;
|
201
|
+
state = State.res_HT;
|
202
|
+
} else {
|
203
|
+
if (E != ch) {
|
204
|
+
settings.call_on_error(this, "not E", data, p_err);
|
205
|
+
}
|
206
|
+
type = ParserType.HTTP_REQUEST;
|
207
|
+
method = HTTPMethod.HTTP_HEAD;
|
208
|
+
index = 2;
|
209
|
+
state = State.req_method;
|
210
|
+
}
|
211
|
+
break;
|
212
|
+
|
213
|
+
|
214
|
+
|
215
|
+
case start_res:
|
216
|
+
flags = 0;
|
217
|
+
content_length = -1;
|
218
|
+
|
219
|
+
settings.call_on_message_begin(this);
|
220
|
+
|
221
|
+
switch(ch) {
|
222
|
+
case H:
|
223
|
+
state = State.res_H;
|
224
|
+
break;
|
225
|
+
case CR:
|
226
|
+
case LF:
|
227
|
+
break;
|
228
|
+
default:
|
229
|
+
settings.call_on_error(this, "Not H or CR/LF", data, p_err);
|
230
|
+
}
|
231
|
+
break;
|
232
|
+
|
233
|
+
|
234
|
+
|
235
|
+
case res_H:
|
236
|
+
if (strict && T != ch) {
|
237
|
+
settings.call_on_error(this, "Not T", data, p_err);
|
238
|
+
}
|
239
|
+
state = State.res_HT;
|
240
|
+
break;
|
241
|
+
case res_HT:
|
242
|
+
if (strict && T != ch) {
|
243
|
+
settings.call_on_error(this, "Not T2", data, p_err);
|
244
|
+
}
|
245
|
+
state = State.res_HTT;
|
246
|
+
break;
|
247
|
+
case res_HTT:
|
248
|
+
if (strict && P != ch) {
|
249
|
+
settings.call_on_error(this, "Not P", data, p_err);
|
250
|
+
}
|
251
|
+
state = State.res_HTTP;
|
252
|
+
break;
|
253
|
+
case res_HTTP:
|
254
|
+
if (strict && SLASH != ch) {
|
255
|
+
settings.call_on_error(this, "Not '/'", data, p_err);
|
256
|
+
}
|
257
|
+
state = State.res_first_http_major;
|
258
|
+
break;
|
259
|
+
|
260
|
+
|
261
|
+
|
262
|
+
case res_first_http_major:
|
263
|
+
if (!isDigit(ch)) {
|
264
|
+
settings.call_on_error(this, "Not a digit", data, p_err);
|
265
|
+
}
|
266
|
+
http_major = (int) ch - 0x30;
|
267
|
+
state = State.res_http_major;
|
268
|
+
break;
|
269
|
+
|
270
|
+
/* major HTTP version or dot */
|
271
|
+
case res_http_major:
|
272
|
+
if (DOT == ch) {
|
273
|
+
state = State.res_http_minor;
|
274
|
+
break;
|
275
|
+
}
|
276
|
+
if (!isDigit(ch)) {
|
277
|
+
settings.call_on_error(this, "Not a digit", data, p_err);
|
278
|
+
}
|
279
|
+
http_major *= 10;
|
280
|
+
http_major += (ch - 0x30);
|
281
|
+
|
282
|
+
if (http_major > 999) {
|
283
|
+
settings.call_on_error(this, "invalid http major version: "+http_major, data, p_err);
|
284
|
+
}
|
285
|
+
break;
|
286
|
+
|
287
|
+
/* first digit of minor HTTP version */
|
288
|
+
case res_first_http_minor:
|
289
|
+
if (!isDigit(ch)) {
|
290
|
+
settings.call_on_error(this, "Not a digit", data, p_err);
|
291
|
+
}
|
292
|
+
http_minor = (int)ch - 0x30;
|
293
|
+
state = State.res_http_minor;
|
294
|
+
break;
|
295
|
+
|
296
|
+
/* minor HTTP version or end of request line */
|
297
|
+
case res_http_minor:
|
298
|
+
if (SPACE == ch) {
|
299
|
+
state = State.res_first_status_code;
|
300
|
+
break;
|
301
|
+
}
|
302
|
+
if (!isDigit(ch)) {
|
303
|
+
settings.call_on_error(this, "Not a digit", data, p_err);
|
304
|
+
}
|
305
|
+
http_minor *= 10;
|
306
|
+
http_minor += (ch - 0x30);
|
307
|
+
|
308
|
+
if (http_minor > 999) {
|
309
|
+
settings.call_on_error(this, "invalid http minor version: "+http_minor, data, p_err);
|
310
|
+
}
|
311
|
+
break;
|
312
|
+
|
313
|
+
|
314
|
+
|
315
|
+
case res_first_status_code:
|
316
|
+
if (!isDigit(ch)) {
|
317
|
+
if (SPACE == ch) {
|
318
|
+
break;
|
319
|
+
}
|
320
|
+
settings.call_on_error(this, "Not a digit (status code)", data, p_err);
|
321
|
+
}
|
322
|
+
status_code = (int)ch - 0x30;
|
323
|
+
state = State.res_status_code;
|
324
|
+
break;
|
325
|
+
|
326
|
+
case res_status_code:
|
327
|
+
if (!isDigit(ch)) {
|
328
|
+
switch(ch) {
|
329
|
+
case SPACE:
|
330
|
+
state = State.res_status;
|
331
|
+
break;
|
332
|
+
case CR:
|
333
|
+
state = State.res_line_almost_done;
|
334
|
+
break;
|
335
|
+
case LF:
|
336
|
+
state = State.header_field_start;
|
337
|
+
break;
|
338
|
+
default:
|
339
|
+
settings.call_on_error(this, "not a valid status code", data, p_err);
|
340
|
+
}
|
341
|
+
break;
|
342
|
+
}
|
343
|
+
status_code *= 10;
|
344
|
+
status_code += (int)ch - 0x30;
|
345
|
+
if (status_code > 999) {
|
346
|
+
settings.call_on_error(this, "ridiculous status code:"+status_code, data, p_err);
|
347
|
+
}
|
348
|
+
break;
|
349
|
+
|
350
|
+
case res_status:
|
351
|
+
/* the human readable status. e.g. "NOT FOUND"
|
352
|
+
* we are not humans so just ignore this
|
353
|
+
* we are not men, we are devo. */
|
354
|
+
|
355
|
+
if (CR == ch) {
|
356
|
+
state = State.res_line_almost_done;
|
357
|
+
break;
|
358
|
+
}
|
359
|
+
if (LF == ch) {
|
360
|
+
state = State.header_field_start;
|
361
|
+
break;
|
362
|
+
}
|
363
|
+
break;
|
364
|
+
|
365
|
+
case res_line_almost_done:
|
366
|
+
if (strict && LF != ch) {
|
367
|
+
settings.call_on_error(this, "not LF", data, p_err);
|
368
|
+
}
|
369
|
+
state = State.header_field_start;
|
370
|
+
break;
|
371
|
+
|
372
|
+
|
373
|
+
|
374
|
+
case start_req:
|
375
|
+
if (CR==ch || LF == LF) {
|
376
|
+
break;
|
377
|
+
}
|
378
|
+
flags = 0;
|
379
|
+
content_length = -1;
|
380
|
+
settings.call_on_message_begin(this);
|
381
|
+
method = start_req_method_assign(ch);
|
382
|
+
if (null == method) {
|
383
|
+
settings.call_on_error(this, "invalid method", data, p_err);
|
384
|
+
}
|
385
|
+
index = 1;
|
386
|
+
state = State.req_method;
|
387
|
+
break;
|
388
|
+
|
389
|
+
|
390
|
+
|
391
|
+
case req_method:
|
392
|
+
if (0 == ch) {
|
393
|
+
settings.call_on_error(this, "NULL in method", data, p_err);
|
394
|
+
}
|
395
|
+
|
396
|
+
byte [] arr = method.bytes;
|
397
|
+
|
398
|
+
if (SPACE == ch && index == arr.length) {
|
399
|
+
state = State.req_spaces_before_url;
|
400
|
+
} else if (arr[index] == ch) {
|
401
|
+
// wuhu!
|
402
|
+
} else if (HTTPMethod.HTTP_CONNECT == method) {
|
403
|
+
if (1 == index && H == ch) {
|
404
|
+
method = HTTPMethod.HTTP_CHECKOUT;
|
405
|
+
} else if (2 == index && P == ch) {
|
406
|
+
method = HTTPMethod.HTTP_COPY;
|
407
|
+
}
|
408
|
+
} else if (HTTPMethod.HTTP_MKCOL == method) {
|
409
|
+
if (1 == index && O == ch) {
|
410
|
+
method = HTTPMethod.HTTP_MOVE;
|
411
|
+
} else if (1 == index && E == ch) {
|
412
|
+
method = HTTPMethod.HTTP_MERGE;
|
413
|
+
} else if (2 == index && A == ch) {
|
414
|
+
method = HTTPMethod.HTTP_MKACTIVITY;
|
415
|
+
}
|
416
|
+
} else if (1 == index && HTTPMethod.HTTP_POST == method && R == ch) {
|
417
|
+
method = HTTPMethod.HTTP_PROPFIND;
|
418
|
+
} else if (1 == index && HTTPMethod.HTTP_POST == method && U == ch) {
|
419
|
+
method = HTTPMethod.HTTP_PUT;
|
420
|
+
} else if (4 == index && HTTPMethod.HTTP_PROPFIND == method && P == ch) {
|
421
|
+
method = HTTPMethod.HTTP_PROPPATCH;
|
422
|
+
} else {
|
423
|
+
settings.call_on_error(this, "Invalid HTTP method", data, p_err);
|
424
|
+
}
|
425
|
+
|
426
|
+
++index;
|
427
|
+
break;
|
428
|
+
|
429
|
+
|
430
|
+
|
431
|
+
/******************* URL *******************/
|
432
|
+
case req_spaces_before_url:
|
433
|
+
if (SPACE == ch) {
|
434
|
+
break;
|
435
|
+
}
|
436
|
+
if (SLASH == ch) {
|
437
|
+
url_mark = p;
|
438
|
+
path_mark = p;
|
439
|
+
state = State.req_path;
|
440
|
+
break;
|
441
|
+
}
|
442
|
+
if (isAtoZ(ch)) {
|
443
|
+
url_mark = p;
|
444
|
+
state = State.req_schema;
|
445
|
+
break;
|
446
|
+
}
|
447
|
+
settings.call_on_error(this, "Invalid something", data, p_err);
|
448
|
+
|
449
|
+
case req_schema:
|
450
|
+
if (isAtoZ(ch)){
|
451
|
+
break;
|
452
|
+
}
|
453
|
+
if (COLON == ch) {
|
454
|
+
state = State.req_schema_slash;
|
455
|
+
break;
|
456
|
+
} else if (DOT == ch) {
|
457
|
+
state = State.req_host;
|
458
|
+
break;
|
459
|
+
}
|
460
|
+
settings.call_on_error(this, "invalid char in schema: "+ch, data, p_err);
|
461
|
+
|
462
|
+
case req_schema_slash:
|
463
|
+
if (strict && SLASH != ch) {
|
464
|
+
settings.call_on_error(this, "invalid char in schema, not /", data, p_err);
|
465
|
+
}
|
466
|
+
state = State.req_schema_slash_slash;
|
467
|
+
break;
|
468
|
+
|
469
|
+
case req_schema_slash_slash:
|
470
|
+
if (strict && SLASH != ch) {
|
471
|
+
settings.call_on_error(this, "invalid char in schema, not /", data, p_err);
|
472
|
+
}
|
473
|
+
state = State.req_host;
|
474
|
+
break;
|
475
|
+
|
476
|
+
case req_host:
|
477
|
+
if (isAtoZ(ch)) {
|
478
|
+
break;
|
479
|
+
}
|
480
|
+
if (isDigit(ch) || DOT == ch || DASH == ch) break;
|
481
|
+
switch (ch) {
|
482
|
+
case COLON:
|
483
|
+
state = State.req_port;
|
484
|
+
break;
|
485
|
+
case SLASH:
|
486
|
+
path_mark = p;
|
487
|
+
break;
|
488
|
+
case SPACE:
|
489
|
+
/* The request line looks like:
|
490
|
+
* "GET http://foo.bar.com HTTP/1.1"
|
491
|
+
* That is, there is no path.
|
492
|
+
*/
|
493
|
+
settings.call_on_url(this, data, url_mark, p-url_mark);
|
494
|
+
url_mark = -1;
|
495
|
+
state = State.req_http_start;
|
496
|
+
break;
|
497
|
+
default:
|
498
|
+
settings.call_on_error(this, "host error in method line", data, p_err);
|
499
|
+
}
|
500
|
+
break;
|
501
|
+
|
502
|
+
case req_port:
|
503
|
+
if (isDigit(ch)) break;
|
504
|
+
switch (ch) {
|
505
|
+
case SLASH:
|
506
|
+
path_mark = p;
|
507
|
+
state = State.req_path;
|
508
|
+
break;
|
509
|
+
case SPACE:
|
510
|
+
/* The request line looks like:
|
511
|
+
* "GET http://foo.bar.com:1234 HTTP/1.1"
|
512
|
+
* That is, there is no path.
|
513
|
+
*/
|
514
|
+
settings.call_on_url(this,data,url_mark,p-url_mark);
|
515
|
+
url_mark = -1;
|
516
|
+
state = State.req_http_start;
|
517
|
+
break;
|
518
|
+
default:
|
519
|
+
settings.call_on_error(this, "invalid port", data, p_err);
|
520
|
+
}
|
521
|
+
break;
|
522
|
+
|
523
|
+
case req_path:
|
524
|
+
if (normal_url_char[ch]) break;
|
525
|
+
switch (ch) {
|
526
|
+
case SPACE:
|
527
|
+
settings.call_on_url(this,data,url_mark, p-url_mark);
|
528
|
+
url_mark = -1;
|
529
|
+
|
530
|
+
settings.call_on_path(this,data,path_mark, p-path_mark);
|
531
|
+
path_mark = -1;
|
532
|
+
|
533
|
+
state = State.req_http_start;
|
534
|
+
break;
|
535
|
+
|
536
|
+
case CR:
|
537
|
+
settings.call_on_url(this,data,url_mark, p-url_mark);
|
538
|
+
url_mark = -1;
|
539
|
+
|
540
|
+
settings.call_on_path(this,data,path_mark, p-path_mark);
|
541
|
+
path_mark = -1;
|
542
|
+
|
543
|
+
http_minor = 9;
|
544
|
+
state = State.res_line_almost_done;
|
545
|
+
break;
|
546
|
+
|
547
|
+
case LF:
|
548
|
+
settings.call_on_url(this,data,url_mark, p-url_mark);
|
549
|
+
url_mark = -1;
|
550
|
+
|
551
|
+
settings.call_on_path(this,data,path_mark, p-path_mark);
|
552
|
+
path_mark = -1;
|
553
|
+
|
554
|
+
http_minor = 9;
|
555
|
+
state = State.header_field_start;
|
556
|
+
break;
|
557
|
+
|
558
|
+
case QMARK:
|
559
|
+
settings.call_on_path(this,data,path_mark, p-path_mark);
|
560
|
+
path_mark = -1;
|
561
|
+
|
562
|
+
state = State.req_query_string_start;
|
563
|
+
break;
|
564
|
+
|
565
|
+
case HASH:
|
566
|
+
settings.call_on_path(this,data,path_mark, p-path_mark);
|
567
|
+
path_mark = -1;
|
568
|
+
|
569
|
+
state = State.req_fragment_start;
|
570
|
+
break;
|
571
|
+
|
572
|
+
default:
|
573
|
+
settings.call_on_error(this, "unexpected char in path", data, p_err);
|
574
|
+
}
|
575
|
+
break;
|
576
|
+
|
577
|
+
case req_query_string_start:
|
578
|
+
if (normal_url_char[ch]) {
|
579
|
+
query_string_mark = p;
|
580
|
+
state = State.req_query_string;
|
581
|
+
break;
|
582
|
+
}
|
583
|
+
|
584
|
+
switch (ch) {
|
585
|
+
case QMARK: break;
|
586
|
+
case SPACE:
|
587
|
+
settings.call_on_url(this, data, url_mark, p-url_mark);
|
588
|
+
url_mark = -1;
|
589
|
+
state = State.req_http_start;
|
590
|
+
break;
|
591
|
+
case CR:
|
592
|
+
settings.call_on_url(this,data,url_mark, p-url_mark);
|
593
|
+
url_mark = -1;
|
594
|
+
http_minor = 9;
|
595
|
+
state = State.res_line_almost_done;
|
596
|
+
break;
|
597
|
+
case LF:
|
598
|
+
settings.call_on_url(this,data,url_mark, p-url_mark);
|
599
|
+
url_mark = -1;
|
600
|
+
http_minor = 9;
|
601
|
+
state = State.header_field_start;
|
602
|
+
break;
|
603
|
+
case HASH:
|
604
|
+
state = State.req_fragment_start;
|
605
|
+
break;
|
606
|
+
default:
|
607
|
+
settings.call_on_error(this, "unexpected char in path", data, p_err);
|
608
|
+
}
|
609
|
+
break;
|
610
|
+
|
611
|
+
case req_query_string:
|
612
|
+
if (normal_url_char[ch]) {
|
613
|
+
break;
|
614
|
+
}
|
615
|
+
|
616
|
+
switch (ch) {
|
617
|
+
case QMARK: break; // allow extra '?' in query string
|
618
|
+
case SPACE:
|
619
|
+
settings.call_on_url(this, data, url_mark, p-url_mark);
|
620
|
+
url_mark = -1;
|
621
|
+
|
622
|
+
settings.call_on_query_string(this, data, query_string_mark, p-query_string_mark);
|
623
|
+
query_string_mark = -1;
|
624
|
+
|
625
|
+
state = State.req_http_start;
|
626
|
+
break;
|
627
|
+
case CR:
|
628
|
+
settings.call_on_url(this,data,url_mark, p-url_mark);
|
629
|
+
url_mark = -1;
|
630
|
+
|
631
|
+
settings.call_on_query_string(this, data, query_string_mark, p-query_string_mark);
|
632
|
+
query_string_mark = -1;
|
633
|
+
|
634
|
+
http_minor = 9;
|
635
|
+
state = State.res_line_almost_done;
|
636
|
+
break;
|
637
|
+
case LF:
|
638
|
+
settings.call_on_url(this,data,url_mark, p-url_mark);
|
639
|
+
url_mark = -1;
|
640
|
+
|
641
|
+
settings.call_on_query_string(this, data, query_string_mark, p-query_string_mark);
|
642
|
+
query_string_mark = -1;
|
643
|
+
http_minor = 9;
|
644
|
+
|
645
|
+
state = State.header_field_start;
|
646
|
+
break;
|
647
|
+
case HASH:
|
648
|
+
settings.call_on_query_string(this, data, query_string_mark, p-query_string_mark);
|
649
|
+
query_string_mark = -1;
|
650
|
+
|
651
|
+
state = State.req_fragment_start;
|
652
|
+
break;
|
653
|
+
default:
|
654
|
+
settings.call_on_error(this, "unexpected char in path", data, p_err);
|
655
|
+
}
|
656
|
+
break;
|
657
|
+
|
658
|
+
case req_fragment_start:
|
659
|
+
if (normal_url_char[ch]) {
|
660
|
+
fragment_mark = p;
|
661
|
+
state = State.req_fragment;
|
662
|
+
break;
|
663
|
+
}
|
664
|
+
|
665
|
+
switch (ch) {
|
666
|
+
case SPACE:
|
667
|
+
settings.call_on_url(this, data, url_mark, p-url_mark);
|
668
|
+
url_mark = -1;
|
669
|
+
|
670
|
+
state = State.req_http_start;
|
671
|
+
break;
|
672
|
+
case CR:
|
673
|
+
settings.call_on_url(this,data,url_mark, p-url_mark);
|
674
|
+
url_mark = -1;
|
675
|
+
|
676
|
+
http_minor = 9;
|
677
|
+
state = State.res_line_almost_done;
|
678
|
+
break;
|
679
|
+
case LF:
|
680
|
+
settings.call_on_url(this,data,url_mark, p-url_mark);
|
681
|
+
url_mark = -1;
|
682
|
+
|
683
|
+
http_minor = 9;
|
684
|
+
state = State.header_field_start;
|
685
|
+
break;
|
686
|
+
case QMARK:
|
687
|
+
fragment_mark = p;
|
688
|
+
state = State.req_fragment;
|
689
|
+
break;
|
690
|
+
case HASH:
|
691
|
+
break;
|
692
|
+
default:
|
693
|
+
settings.call_on_error(this, "unexpected char in path", data, p_err);
|
694
|
+
}
|
695
|
+
break;
|
696
|
+
|
697
|
+
case req_fragment:
|
698
|
+
if (normal_url_char[ch]) {
|
699
|
+
break;
|
700
|
+
}
|
701
|
+
|
702
|
+
switch (ch) {
|
703
|
+
case SPACE:
|
704
|
+
settings.call_on_url(this, data, url_mark, p-url_mark);
|
705
|
+
url_mark = -1;
|
706
|
+
|
707
|
+
settings.call_on_fragment(this, data, fragment_mark, p-fragment_mark);
|
708
|
+
fragment_mark = -1;
|
709
|
+
|
710
|
+
state = State.req_http_start;
|
711
|
+
break;
|
712
|
+
case CR:
|
713
|
+
settings.call_on_url(this,data,url_mark, p-url_mark);
|
714
|
+
url_mark = -1;
|
715
|
+
|
716
|
+
settings.call_on_fragment(this, data, query_string_mark, p-query_string_mark);
|
717
|
+
fragment_mark = -1;
|
718
|
+
|
719
|
+
http_minor = 9;
|
720
|
+
state = State.res_line_almost_done;
|
721
|
+
break;
|
722
|
+
case LF:
|
723
|
+
settings.call_on_url(this,data,url_mark, p-url_mark);
|
724
|
+
url_mark = -1;
|
725
|
+
|
726
|
+
settings.call_on_fragment(this, data, query_string_mark, p-query_string_mark);
|
727
|
+
fragment_mark = -1;
|
728
|
+
|
729
|
+
http_minor = 9;
|
730
|
+
state = State.header_field_start;
|
731
|
+
break;
|
732
|
+
case QMARK:
|
733
|
+
case HASH:
|
734
|
+
break;
|
735
|
+
default:
|
736
|
+
settings.call_on_error(this, "unexpected char in path", data, p_err);
|
737
|
+
}
|
738
|
+
break;
|
739
|
+
/******************* URL *******************/
|
740
|
+
|
741
|
+
|
742
|
+
|
743
|
+
/******************* HTTP 1.1 *******************/
|
744
|
+
case req_http_start:
|
745
|
+
switch (ch) {
|
746
|
+
case H:
|
747
|
+
state = State.req_http_H;
|
748
|
+
break;
|
749
|
+
case SPACE:
|
750
|
+
break;
|
751
|
+
default:
|
752
|
+
settings.call_on_error(this, "error in req_http_H", data, p_err);
|
753
|
+
}
|
754
|
+
break;
|
755
|
+
|
756
|
+
case req_http_H:
|
757
|
+
if (strict && T != ch) {
|
758
|
+
settings.call_on_error(this, "unexpected char", data, p_err);
|
759
|
+
}
|
760
|
+
state = State.req_http_HT;
|
761
|
+
break;
|
762
|
+
|
763
|
+
case req_http_HT:
|
764
|
+
if (strict && T != ch) {
|
765
|
+
settings.call_on_error(this, "unexpected char", data, p_err);
|
766
|
+
}
|
767
|
+
state = State.req_http_HTT;
|
768
|
+
break;
|
769
|
+
|
770
|
+
case req_http_HTT:
|
771
|
+
if (strict && P != ch) {
|
772
|
+
settings.call_on_error(this, "unexpected char", data, p_err);
|
773
|
+
}
|
774
|
+
state = State.req_http_HTTP;
|
775
|
+
break;
|
776
|
+
|
777
|
+
case req_http_HTTP:
|
778
|
+
if (strict && SLASH != ch) {
|
779
|
+
settings.call_on_error(this, "unexpected char", data, p_err);
|
780
|
+
}
|
781
|
+
state = req_first_http_major;
|
782
|
+
break;
|
783
|
+
|
784
|
+
/* first digit of major HTTP version */
|
785
|
+
case req_first_http_major:
|
786
|
+
if (!isDigit(ch)) {
|
787
|
+
settings.call_on_error(this, "non digit in http major", data, p_err);
|
788
|
+
}
|
789
|
+
http_major = (int)ch - 0x30;
|
790
|
+
state = State.req_http_major;
|
791
|
+
break;
|
792
|
+
|
793
|
+
/* major HTTP version or dot */
|
794
|
+
case req_http_major:
|
795
|
+
if (DOT == ch) {
|
796
|
+
state = State.req_first_http_minor;
|
797
|
+
break;
|
798
|
+
}
|
799
|
+
|
800
|
+
if (!isDigit(ch)) {
|
801
|
+
settings.call_on_error(this, "non digit in http major", data, p_err);
|
802
|
+
}
|
803
|
+
|
804
|
+
http_major *= 10;
|
805
|
+
http_major += (int)ch - 0x30;
|
806
|
+
|
807
|
+
if (http_major > 999) {
|
808
|
+
settings.call_on_error(this, "ridiculous http major", data, p_err);
|
809
|
+
};
|
810
|
+
break;
|
811
|
+
|
812
|
+
/* first digit of minor HTTP version */
|
813
|
+
case req_first_http_minor:
|
814
|
+
if (!isDigit(ch)) {
|
815
|
+
settings.call_on_error(this, "non digit in http minor", data, p_err);
|
816
|
+
}
|
817
|
+
http_minor = (int)ch - 0x30;
|
818
|
+
state = State.req_http_minor;
|
819
|
+
break;
|
820
|
+
|
821
|
+
case req_http_minor:
|
822
|
+
if (ch == CR) {
|
823
|
+
state = State.req_line_almost_done;
|
824
|
+
break;
|
825
|
+
}
|
826
|
+
|
827
|
+
if (ch == LF) {
|
828
|
+
state = State.header_field_start;
|
829
|
+
break;
|
830
|
+
}
|
831
|
+
|
832
|
+
/* XXX allow spaces after digit? */
|
833
|
+
|
834
|
+
if (!isDigit(ch)) {
|
835
|
+
settings.call_on_error(this, "non digit in http minor", data, p_err);
|
836
|
+
}
|
837
|
+
|
838
|
+
http_minor *= 10;
|
839
|
+
http_minor += (int)ch - 0x30;
|
840
|
+
|
841
|
+
|
842
|
+
if (http_minor > 999) {
|
843
|
+
settings.call_on_error(this, "ridiculous http minor", data, p_err);
|
844
|
+
};
|
845
|
+
|
846
|
+
break;
|
847
|
+
|
848
|
+
/* end of request line */
|
849
|
+
case req_line_almost_done:
|
850
|
+
{
|
851
|
+
if (ch != LF) {
|
852
|
+
settings.call_on_error(this, "missing LF after request line", data, p_err);
|
853
|
+
}
|
854
|
+
state = State.header_field_start;
|
855
|
+
break;
|
856
|
+
}
|
857
|
+
|
858
|
+
/******************* HTTP 1.1 *******************/
|
859
|
+
|
860
|
+
|
861
|
+
|
862
|
+
/******************* Header *******************/
|
863
|
+
case header_field_start:
|
864
|
+
{
|
865
|
+
if (ch == CR) {
|
866
|
+
state = State.headers_almost_done;
|
867
|
+
break;
|
868
|
+
}
|
869
|
+
|
870
|
+
if (ch == LF) {
|
871
|
+
/* they might be just sending \n instead of \r\n so this would be
|
872
|
+
* the second \n to denote the end of headers*/
|
873
|
+
state = State.headers_almost_done;
|
874
|
+
if (!headers_almost_done(ch, settings)) {
|
875
|
+
settings.call_on_error(this, "header not properly completed", data, p_err);
|
876
|
+
}
|
877
|
+
break;
|
878
|
+
}
|
879
|
+
|
880
|
+
c = upper(ch);
|
881
|
+
|
882
|
+
if (c < A || Z < c) {
|
883
|
+
settings.call_on_error(this, "invalid char in header", data, p_err);
|
884
|
+
};
|
885
|
+
|
886
|
+
header_field_mark = p;
|
887
|
+
|
888
|
+
index = 0;
|
889
|
+
state = State.header_field;
|
890
|
+
|
891
|
+
switch (c) {
|
892
|
+
case C:
|
893
|
+
header_state = HState.C;
|
894
|
+
break;
|
895
|
+
|
896
|
+
case P:
|
897
|
+
header_state = HState.matching_proxy_connection;
|
898
|
+
break;
|
899
|
+
|
900
|
+
case T:
|
901
|
+
header_state = HState.matching_transfer_encoding;
|
902
|
+
break;
|
903
|
+
|
904
|
+
case U:
|
905
|
+
header_state = HState.matching_upgrade;
|
906
|
+
break;
|
907
|
+
|
908
|
+
default:
|
909
|
+
header_state = HState.general;
|
910
|
+
break;
|
911
|
+
}
|
912
|
+
break;
|
913
|
+
}
|
914
|
+
|
915
|
+
|
916
|
+
|
917
|
+
case header_field:
|
918
|
+
{
|
919
|
+
c = (byte) acceptable_header[ch];
|
920
|
+
if (0 != c) {
|
921
|
+
switch (header_state) {
|
922
|
+
case general:
|
923
|
+
break;
|
924
|
+
|
925
|
+
case C:
|
926
|
+
index++;
|
927
|
+
header_state = (O == c ? HState.CO : HState.general);
|
928
|
+
break;
|
929
|
+
|
930
|
+
case CO:
|
931
|
+
index++;
|
932
|
+
header_state = (N == c ? HState.CON : HState.general);
|
933
|
+
break;
|
934
|
+
|
935
|
+
case CON:
|
936
|
+
index++;
|
937
|
+
switch (c) {
|
938
|
+
case N:
|
939
|
+
header_state = HState.matching_connection;
|
940
|
+
break;
|
941
|
+
case T:
|
942
|
+
header_state = HState.matching_content_length;
|
943
|
+
break;
|
944
|
+
default:
|
945
|
+
header_state = HState.general;
|
946
|
+
break;
|
947
|
+
}
|
948
|
+
break;
|
949
|
+
|
950
|
+
/* connection */
|
951
|
+
|
952
|
+
case matching_connection:
|
953
|
+
index++;
|
954
|
+
if (index > CONNECTION.length || c != CONNECTION[index]) {
|
955
|
+
header_state = HState.general;
|
956
|
+
} else if (index == CONNECTION.length-1) {
|
957
|
+
header_state = HState.connection;
|
958
|
+
}
|
959
|
+
break;
|
960
|
+
|
961
|
+
/* proxy-connection */
|
962
|
+
|
963
|
+
case matching_proxy_connection:
|
964
|
+
index++;
|
965
|
+
if (index > PROXY_CONNECTION.length || c != PROXY_CONNECTION[index]) {
|
966
|
+
header_state = HState.general;
|
967
|
+
} else if (index == PROXY_CONNECTION.length-1) {
|
968
|
+
header_state = HState.connection;
|
969
|
+
}
|
970
|
+
break;
|
971
|
+
|
972
|
+
/* content-length */
|
973
|
+
|
974
|
+
case matching_content_length:
|
975
|
+
index++;
|
976
|
+
if (index > CONTENT_LENGTH.length || c != CONTENT_LENGTH[index]) {
|
977
|
+
header_state = HState.general;
|
978
|
+
} else if (index == CONTENT_LENGTH.length-1) {
|
979
|
+
header_state = HState.content_length;
|
980
|
+
}
|
981
|
+
break;
|
982
|
+
|
983
|
+
/* transfer-encoding */
|
984
|
+
|
985
|
+
case matching_transfer_encoding:
|
986
|
+
index++;
|
987
|
+
if (index > TRANSFER_ENCODING.length || c != TRANSFER_ENCODING[index]) {
|
988
|
+
header_state = HState.general;
|
989
|
+
} else if (index == TRANSFER_ENCODING.length-1) {
|
990
|
+
header_state = HState.transfer_encoding;
|
991
|
+
}
|
992
|
+
break;
|
993
|
+
|
994
|
+
/* upgrade */
|
995
|
+
|
996
|
+
case matching_upgrade:
|
997
|
+
index++;
|
998
|
+
if (index > UPGRADE.length || c != UPGRADE[index]) {
|
999
|
+
header_state = HState.general;
|
1000
|
+
} else if (index == UPGRADE.length-1) {
|
1001
|
+
header_state = HState.upgrade;
|
1002
|
+
}
|
1003
|
+
break;
|
1004
|
+
|
1005
|
+
case connection:
|
1006
|
+
case content_length:
|
1007
|
+
case transfer_encoding:
|
1008
|
+
case upgrade:
|
1009
|
+
if (SPACE != ch) header_state = HState.general;
|
1010
|
+
break;
|
1011
|
+
|
1012
|
+
default:
|
1013
|
+
settings.call_on_error(this, "Unknown Header State", data, p_err);
|
1014
|
+
} // switch: header_state
|
1015
|
+
break;
|
1016
|
+
} // 0 != c
|
1017
|
+
|
1018
|
+
if (COLON == ch) {
|
1019
|
+
settings.call_on_header_field(this, data, header_field_mark, p-header_field_mark);
|
1020
|
+
header_field_mark = -1;
|
1021
|
+
|
1022
|
+
state = State.header_value_start;
|
1023
|
+
break;
|
1024
|
+
}
|
1025
|
+
|
1026
|
+
if (CR == ch) {
|
1027
|
+
state = State.header_almost_done;
|
1028
|
+
settings.call_on_header_field(this, data, header_field_mark, p-header_field_mark);
|
1029
|
+
|
1030
|
+
header_field_mark = -1;
|
1031
|
+
break;
|
1032
|
+
}
|
1033
|
+
|
1034
|
+
if (ch == LF) {
|
1035
|
+
settings.call_on_header_field(this, data, header_field_mark, p-header_field_mark);
|
1036
|
+
header_field_mark = -1;
|
1037
|
+
|
1038
|
+
state = State.header_field_start;
|
1039
|
+
break;
|
1040
|
+
}
|
1041
|
+
|
1042
|
+
settings.call_on_error(this, "invalid header field", data, p_err);
|
1043
|
+
}
|
1044
|
+
|
1045
|
+
|
1046
|
+
|
1047
|
+
case header_value_start:
|
1048
|
+
{
|
1049
|
+
if (SPACE == ch) break;
|
1050
|
+
|
1051
|
+
header_value_mark = p;
|
1052
|
+
|
1053
|
+
state = State.header_value;
|
1054
|
+
index = 0;
|
1055
|
+
|
1056
|
+
c = (byte)acceptable_header[ch];
|
1057
|
+
|
1058
|
+
if (c == 0) {
|
1059
|
+
if (CR == ch) {
|
1060
|
+
settings.call_on_header_value(this, data, header_value_mark, p-header_value_mark);
|
1061
|
+
header_value_mark = -1;
|
1062
|
+
|
1063
|
+
header_state = HState.general;
|
1064
|
+
state = State.header_almost_done;
|
1065
|
+
break;
|
1066
|
+
}
|
1067
|
+
|
1068
|
+
if (LF == ch) {
|
1069
|
+
settings.call_on_header_value(this, data, header_value_mark, p-header_value_mark);
|
1070
|
+
header_value_mark = -1;
|
1071
|
+
|
1072
|
+
state = State.header_field_start;
|
1073
|
+
break;
|
1074
|
+
}
|
1075
|
+
|
1076
|
+
header_state = HState.general;
|
1077
|
+
break;
|
1078
|
+
}
|
1079
|
+
|
1080
|
+
switch (header_state) {
|
1081
|
+
case upgrade:
|
1082
|
+
flags |= F_UPGRADE;
|
1083
|
+
header_state = HState.general;
|
1084
|
+
break;
|
1085
|
+
|
1086
|
+
case transfer_encoding:
|
1087
|
+
/* looking for 'Transfer-Encoding: chunked' */
|
1088
|
+
if (C == c) {
|
1089
|
+
header_state = HState.matching_transfer_encoding_chunked;
|
1090
|
+
} else {
|
1091
|
+
header_state = HState.general;
|
1092
|
+
}
|
1093
|
+
break;
|
1094
|
+
|
1095
|
+
case content_length:
|
1096
|
+
if (!isDigit(ch)) {
|
1097
|
+
settings.call_on_error(this, "Content-Length not numeric", data, p_err);
|
1098
|
+
}
|
1099
|
+
content_length = (int)ch - 0x30;
|
1100
|
+
break;
|
1101
|
+
|
1102
|
+
case connection:
|
1103
|
+
/* looking for 'Connection: keep-alive' */
|
1104
|
+
if (K == c) {
|
1105
|
+
header_state = HState.matching_connection_keep_alive;
|
1106
|
+
/* looking for 'Connection: close' */
|
1107
|
+
} else if (C == c) {
|
1108
|
+
header_state = HState.matching_connection_close;
|
1109
|
+
} else {
|
1110
|
+
header_state = HState.general;
|
1111
|
+
}
|
1112
|
+
break;
|
1113
|
+
|
1114
|
+
default:
|
1115
|
+
header_state = HState.general;
|
1116
|
+
break;
|
1117
|
+
}
|
1118
|
+
break;
|
1119
|
+
} // header value start
|
1120
|
+
|
1121
|
+
|
1122
|
+
|
1123
|
+
case header_value:
|
1124
|
+
{
|
1125
|
+
c = (byte)acceptable_header[ch];
|
1126
|
+
|
1127
|
+
if (c == 0) {
|
1128
|
+
if (CR == ch) {
|
1129
|
+
settings.call_on_header_value(this, data, header_value_mark, p-header_value_mark);
|
1130
|
+
header_value_mark = -1;
|
1131
|
+
|
1132
|
+
state = State.header_almost_done;
|
1133
|
+
break;
|
1134
|
+
}
|
1135
|
+
|
1136
|
+
if (LF == ch) {
|
1137
|
+
settings.call_on_header_value(this, data, header_value_mark, p-header_value_mark);
|
1138
|
+
header_value_mark = -1;
|
1139
|
+
|
1140
|
+
if (!header_almost_done(ch)) {
|
1141
|
+
settings.call_on_error(this,"incorrect header ending, expection LF", data, p_err);
|
1142
|
+
}
|
1143
|
+
break;
|
1144
|
+
}
|
1145
|
+
break;
|
1146
|
+
}
|
1147
|
+
|
1148
|
+
switch (header_state) {
|
1149
|
+
case general:
|
1150
|
+
break;
|
1151
|
+
|
1152
|
+
case connection:
|
1153
|
+
case transfer_encoding:
|
1154
|
+
settings.call_on_error(this, "Shouldn't be here", data, p_err);
|
1155
|
+
|
1156
|
+
case content_length:
|
1157
|
+
if (SPACE == ch) {
|
1158
|
+
break;
|
1159
|
+
}
|
1160
|
+
if (!isDigit(ch)) {
|
1161
|
+
settings.call_on_error(this, "Content-Length not numeric", data, p_err);
|
1162
|
+
}
|
1163
|
+
|
1164
|
+
content_length *= 10;
|
1165
|
+
content_length += (int)ch - 0x30;
|
1166
|
+
break;
|
1167
|
+
|
1168
|
+
/* Transfer-Encoding: chunked */
|
1169
|
+
case matching_transfer_encoding_chunked:
|
1170
|
+
index++;
|
1171
|
+
if (index > CHUNKED.length || c != CHUNKED[index]) {
|
1172
|
+
header_state = HState.general;
|
1173
|
+
} else if (index == CHUNKED.length-1) {
|
1174
|
+
header_state = HState.transfer_encoding_chunked;
|
1175
|
+
}
|
1176
|
+
break;
|
1177
|
+
|
1178
|
+
/* looking for 'Connection: keep-alive' */
|
1179
|
+
case matching_connection_keep_alive:
|
1180
|
+
index++;
|
1181
|
+
if (index > KEEP_ALIVE.length || c != KEEP_ALIVE[index]) {
|
1182
|
+
header_state = HState.general;
|
1183
|
+
} else if (index == KEEP_ALIVE.length-1) {
|
1184
|
+
header_state = HState.connection_keep_alive;
|
1185
|
+
}
|
1186
|
+
break;
|
1187
|
+
|
1188
|
+
/* looking for 'Connection: close' */
|
1189
|
+
case matching_connection_close:
|
1190
|
+
index++;
|
1191
|
+
if (index > CLOSE.length || c != CLOSE[index]) {
|
1192
|
+
header_state = HState.general;
|
1193
|
+
} else if (index == CLOSE.length-1) {
|
1194
|
+
header_state = HState.connection_close;
|
1195
|
+
}
|
1196
|
+
break;
|
1197
|
+
|
1198
|
+
case transfer_encoding_chunked:
|
1199
|
+
case connection_keep_alive:
|
1200
|
+
case connection_close:
|
1201
|
+
if (SPACE != ch) header_state = HState.general;
|
1202
|
+
break;
|
1203
|
+
|
1204
|
+
default:
|
1205
|
+
state = State.header_value;
|
1206
|
+
header_state = HState.general;
|
1207
|
+
break;
|
1208
|
+
}
|
1209
|
+
break;
|
1210
|
+
} // header_value
|
1211
|
+
|
1212
|
+
|
1213
|
+
|
1214
|
+
case header_almost_done:
|
1215
|
+
if (!header_almost_done(ch)) {
|
1216
|
+
settings.call_on_error(this,"incorrect header ending, expection LF", data, p_err);
|
1217
|
+
}
|
1218
|
+
break;
|
1219
|
+
|
1220
|
+
case headers_almost_done:
|
1221
|
+
if (!headers_almost_done(ch, settings)) {
|
1222
|
+
settings.call_on_error(this, "header not properly completed", data, p_err);
|
1223
|
+
}
|
1224
|
+
break;
|
1225
|
+
|
1226
|
+
/******************* Header *******************/
|
1227
|
+
|
1228
|
+
|
1229
|
+
|
1230
|
+
|
1231
|
+
/******************* Body *******************/
|
1232
|
+
case body_identity:
|
1233
|
+
to_read = min(pe - p, content_length); //TODO change to use buffer?
|
1234
|
+
|
1235
|
+
if (to_read > 0) {
|
1236
|
+
settings.call_on_body(this, data, p, to_read);
|
1237
|
+
data.position(p+to_read);
|
1238
|
+
content_length -= to_read;
|
1239
|
+
if (content_length == 0) {
|
1240
|
+
settings.call_on_message_complete(this);
|
1241
|
+
state = new_message();
|
1242
|
+
}
|
1243
|
+
}
|
1244
|
+
break;
|
1245
|
+
|
1246
|
+
|
1247
|
+
|
1248
|
+
case body_identity_eof:
|
1249
|
+
to_read = pe - p; // TODO change to use buffer ?
|
1250
|
+
if (to_read > 0) {
|
1251
|
+
settings.call_on_body(this, data, p, to_read);
|
1252
|
+
data.position(p+to_read);
|
1253
|
+
}
|
1254
|
+
break;
|
1255
|
+
/******************* Body *******************/
|
1256
|
+
|
1257
|
+
|
1258
|
+
|
1259
|
+
/******************* Chunk *******************/
|
1260
|
+
case chunk_size_start:
|
1261
|
+
if (0 == (flags & F_CHUNKED)) {
|
1262
|
+
settings.call_on_error(this, "not chunked", data, p_err);
|
1263
|
+
}
|
1264
|
+
|
1265
|
+
c = UNHEX[ch];
|
1266
|
+
if (c == -1) {
|
1267
|
+
settings.call_on_error(this, "invalid hex char in chunk content length", data, p_err);
|
1268
|
+
}
|
1269
|
+
content_length = c;
|
1270
|
+
state = State.chunk_size;
|
1271
|
+
break;
|
1272
|
+
|
1273
|
+
|
1274
|
+
|
1275
|
+
case chunk_size:
|
1276
|
+
if (0 == (flags & F_CHUNKED)) {
|
1277
|
+
settings.call_on_error(this, "not chunked", data, p_err);
|
1278
|
+
}
|
1279
|
+
|
1280
|
+
if (CR == ch) {
|
1281
|
+
state = State.chunk_size_almost_done;
|
1282
|
+
break;
|
1283
|
+
}
|
1284
|
+
|
1285
|
+
c = UNHEX[ch];
|
1286
|
+
|
1287
|
+
if (c == -1) {
|
1288
|
+
if (SEMI == ch || SPACE == ch) {
|
1289
|
+
state = State.chunk_parameters;
|
1290
|
+
break;
|
1291
|
+
}
|
1292
|
+
settings.call_on_error(this, "invalid hex char in chunk content length", data, p_err);
|
1293
|
+
}
|
1294
|
+
|
1295
|
+
content_length *= 16;
|
1296
|
+
content_length += c;
|
1297
|
+
break;
|
1298
|
+
|
1299
|
+
|
1300
|
+
|
1301
|
+
case chunk_parameters:
|
1302
|
+
if (0 == (flags & F_CHUNKED)) {
|
1303
|
+
settings.call_on_error(this, "not chunked", data, p_err);
|
1304
|
+
}
|
1305
|
+
/* just ignore this shit. TODO check for overflow */
|
1306
|
+
if (CR == ch) {
|
1307
|
+
state = State.chunk_size_almost_done;
|
1308
|
+
break;
|
1309
|
+
}
|
1310
|
+
break;
|
1311
|
+
|
1312
|
+
|
1313
|
+
|
1314
|
+
case chunk_size_almost_done:
|
1315
|
+
if (0 == (flags & F_CHUNKED)) {
|
1316
|
+
settings.call_on_error(this, "not chunked", data, p_err);
|
1317
|
+
}
|
1318
|
+
if (strict && LF != ch) {
|
1319
|
+
settings.call_on_error(this, "expected LF at end of chunk size", data, p_err);
|
1320
|
+
}
|
1321
|
+
|
1322
|
+
if (0 == content_length) {
|
1323
|
+
flags |= F_TRAILING;
|
1324
|
+
state = State.header_field_start;
|
1325
|
+
} else {
|
1326
|
+
state = State.chunk_data;
|
1327
|
+
}
|
1328
|
+
break;
|
1329
|
+
|
1330
|
+
|
1331
|
+
|
1332
|
+
case chunk_data:
|
1333
|
+
{
|
1334
|
+
if (0 == (flags & F_CHUNKED)) {
|
1335
|
+
settings.call_on_error(this, "not chunked", data, p_err);
|
1336
|
+
}
|
1337
|
+
|
1338
|
+
to_read = min(pe-p, content_length);
|
1339
|
+
if (to_read > 0) {
|
1340
|
+
settings.call_on_body(this, data, p, to_read);
|
1341
|
+
data.position(p+to_read);
|
1342
|
+
}
|
1343
|
+
|
1344
|
+
if (to_read == content_length) {
|
1345
|
+
state = State.chunk_data_almost_done;
|
1346
|
+
}
|
1347
|
+
|
1348
|
+
content_length -= to_read;
|
1349
|
+
break;
|
1350
|
+
}
|
1351
|
+
|
1352
|
+
|
1353
|
+
|
1354
|
+
case chunk_data_almost_done:
|
1355
|
+
if (0 == (flags & F_CHUNKED)) {
|
1356
|
+
settings.call_on_error(this, "not chunked", data, p_err);
|
1357
|
+
}
|
1358
|
+
if (strict && CR != ch) {
|
1359
|
+
settings.call_on_error(this, "chunk data terminated incorrectly, expected CR", data, p_err);
|
1360
|
+
}
|
1361
|
+
state = State.chunk_data_done;
|
1362
|
+
break;
|
1363
|
+
|
1364
|
+
|
1365
|
+
|
1366
|
+
case chunk_data_done:
|
1367
|
+
if (0 == (flags & F_CHUNKED)) {
|
1368
|
+
settings.call_on_error(this, "not chunked", data, p_err);
|
1369
|
+
}
|
1370
|
+
if (strict && LF != ch) {
|
1371
|
+
settings.call_on_error(this, "chunk data terminated incorrectly, expected LF", data, p_err);
|
1372
|
+
}
|
1373
|
+
state = State.chunk_size_start;
|
1374
|
+
break;
|
1375
|
+
/******************* Chunk *******************/
|
1376
|
+
|
1377
|
+
|
1378
|
+
|
1379
|
+
default:
|
1380
|
+
settings.call_on_error(this, "unhandled state", data, p_err);
|
1381
|
+
|
1382
|
+
} // switch
|
1383
|
+
} // while
|
1384
|
+
|
1385
|
+
p = data.position();
|
1386
|
+
|
1387
|
+
|
1388
|
+
/* Reaching this point assumes that we only received part of a
|
1389
|
+
* message, inform the callbacks about the progress made so far*/
|
1390
|
+
|
1391
|
+
settings.call_on_header_field(this, data, header_field_mark, p-header_field_mark);
|
1392
|
+
settings.call_on_header_value(this, data, header_value_mark, p-header_value_mark);
|
1393
|
+
settings.call_on_fragment (this, data, fragment_mark, p-fragment_mark);
|
1394
|
+
settings.call_on_query_string(this, data, query_string_mark, p-query_string_mark);
|
1395
|
+
settings.call_on_path (this, data, path_mark, p-path_mark);
|
1396
|
+
settings.call_on_url (this, data, url_mark, p-url_mark);
|
1397
|
+
|
1398
|
+
} // execute
|
1399
|
+
|
1400
|
+
/* If http_should_keep_alive() in the on_headers_complete or
|
1401
|
+
* on_message_complete callback returns true, then this will be should be
|
1402
|
+
* the last message on the connection.
|
1403
|
+
* If you are the server, respond with the "Connection: close" header.
|
1404
|
+
* If you are the client, close the connection.
|
1405
|
+
*/
|
1406
|
+
public boolean http_should_keep_alive() {
|
1407
|
+
if (http_major > 0 && http_minor > 0) {
|
1408
|
+
/* HTTP/1.1 */
|
1409
|
+
if ( 0 != (flags & F_CONNECTION_CLOSE) ) {
|
1410
|
+
return false;
|
1411
|
+
} else {
|
1412
|
+
return true;
|
1413
|
+
}
|
1414
|
+
} else {
|
1415
|
+
/* HTTP/1.0 or earlier */
|
1416
|
+
if ( 0 != (flags & F_CONNECTION_KEEP_ALIVE) ) {
|
1417
|
+
return true;
|
1418
|
+
} else {
|
1419
|
+
return false;
|
1420
|
+
}
|
1421
|
+
}
|
1422
|
+
}
|
1423
|
+
|
1424
|
+
boolean isDigit(byte b) {
|
1425
|
+
if (b >= 0x30 && b <=0x39) {
|
1426
|
+
return true;
|
1427
|
+
}
|
1428
|
+
return false;
|
1429
|
+
}
|
1430
|
+
|
1431
|
+
boolean isAtoZ(byte b) {
|
1432
|
+
byte c = lower(b);
|
1433
|
+
return (c>= 0x61 /*a*/ && c <= 0x7a /*z*/);
|
1434
|
+
}
|
1435
|
+
|
1436
|
+
|
1437
|
+
byte lower (byte b) {
|
1438
|
+
return (byte)(b|0x20);
|
1439
|
+
}
|
1440
|
+
|
1441
|
+
byte upper(byte b) {
|
1442
|
+
char c = (char)(b);
|
1443
|
+
return (byte)Character.toUpperCase(c);
|
1444
|
+
}
|
1445
|
+
|
1446
|
+
|
1447
|
+
HTTPMethod start_req_method_assign(byte c){
|
1448
|
+
switch (c) {
|
1449
|
+
case C: return HTTPMethod.HTTP_CONNECT; /* or COPY, CHECKOUT */
|
1450
|
+
case D: return HTTPMethod.HTTP_DELETE;
|
1451
|
+
case G: return HTTPMethod.HTTP_GET;
|
1452
|
+
case H: return HTTPMethod.HTTP_HEAD;
|
1453
|
+
case L: return HTTPMethod.HTTP_LOCK;
|
1454
|
+
case M: return HTTPMethod.HTTP_MKCOL; /* or MOVE, MKACTIVITY, MERGE */
|
1455
|
+
case O: return HTTPMethod.HTTP_OPTIONS;
|
1456
|
+
case P: return HTTPMethod.HTTP_POST; /* or PROPFIND, PROPPATH, PUT */
|
1457
|
+
case R: return HTTPMethod.HTTP_REPORT;
|
1458
|
+
case T: return HTTPMethod.HTTP_TRACE;
|
1459
|
+
case U: return HTTPMethod.HTTP_UNLOCK;
|
1460
|
+
}
|
1461
|
+
return null; // ugh.
|
1462
|
+
}
|
1463
|
+
|
1464
|
+
boolean header_almost_done(byte ch) {
|
1465
|
+
if (strict && LF != ch) {
|
1466
|
+
return false;
|
1467
|
+
}
|
1468
|
+
|
1469
|
+
state = State.header_field_start;
|
1470
|
+
// TODO java enums support some sort of bitflag mechanism !?
|
1471
|
+
switch (header_state) {
|
1472
|
+
case connection_keep_alive:
|
1473
|
+
flags |= F_CONNECTION_KEEP_ALIVE;
|
1474
|
+
break;
|
1475
|
+
case connection_close:
|
1476
|
+
flags |= F_CONNECTION_CLOSE;
|
1477
|
+
break;
|
1478
|
+
case transfer_encoding_chunked:
|
1479
|
+
flags |= F_CHUNKED;
|
1480
|
+
break;
|
1481
|
+
default:
|
1482
|
+
break;
|
1483
|
+
}
|
1484
|
+
return true;
|
1485
|
+
}
|
1486
|
+
|
1487
|
+
boolean headers_almost_done (byte ch, ParserSettings settings) {
|
1488
|
+
|
1489
|
+
if (LF != ch) {
|
1490
|
+
return false;
|
1491
|
+
}
|
1492
|
+
|
1493
|
+
if (0 != (flags & F_TRAILING)) {
|
1494
|
+
/* End of a chunked request */
|
1495
|
+
|
1496
|
+
settings.call_on_headers_complete(this);
|
1497
|
+
settings.call_on_message_complete(this);
|
1498
|
+
|
1499
|
+
state = new_message();
|
1500
|
+
|
1501
|
+
return true;
|
1502
|
+
}
|
1503
|
+
|
1504
|
+
nread = 0;
|
1505
|
+
|
1506
|
+
if (0 != (flags & F_UPGRADE) || HTTPMethod.HTTP_CONNECT == method) {
|
1507
|
+
upgrade = true;
|
1508
|
+
}
|
1509
|
+
|
1510
|
+
|
1511
|
+
/* Here we call the headers_complete callback. This is somewhat
|
1512
|
+
* different than other callbacks because if the user returns 1, we
|
1513
|
+
* will interpret that as saying that this message has no body. This
|
1514
|
+
* is needed for the annoying case of recieving a response to a HEAD
|
1515
|
+
* request.
|
1516
|
+
*/
|
1517
|
+
|
1518
|
+
/* (responses to HEAD request contain a CONTENT-LENGTH header
|
1519
|
+
* but no content)
|
1520
|
+
*
|
1521
|
+
* Consider what to do here: I don't like the idea of the callback
|
1522
|
+
* interface having a different contract in the case of HEAD
|
1523
|
+
* responses. The alternatives would be either to:
|
1524
|
+
*
|
1525
|
+
* a.) require the header_complete callback to implement a different
|
1526
|
+
* interface or
|
1527
|
+
*
|
1528
|
+
* b.) provide an overridden execute(bla, bla, boolean
|
1529
|
+
* parsingHeader) implementation ...
|
1530
|
+
*/
|
1531
|
+
|
1532
|
+
/*TODO TODO TODO TODO TODO TODO TODO TODO TODO TODO */
|
1533
|
+
if (null != settings.on_headers_complete) {
|
1534
|
+
settings.call_on_headers_complete(this);
|
1535
|
+
//return;
|
1536
|
+
}
|
1537
|
+
|
1538
|
+
// if (null != settings.on_headers_complete) {
|
1539
|
+
// switch (settings.on_headers_complete.cb(parser)) {
|
1540
|
+
// case 0:
|
1541
|
+
// break;
|
1542
|
+
//
|
1543
|
+
// case 1:
|
1544
|
+
// flags |= F_SKIPBODY;
|
1545
|
+
// break;
|
1546
|
+
//
|
1547
|
+
// default:
|
1548
|
+
// return p - data; /* Error */ // TODO // RuntimeException ?
|
1549
|
+
// }
|
1550
|
+
// }
|
1551
|
+
|
1552
|
+
|
1553
|
+
// Exit, the rest of the connect is in a different protocol.
|
1554
|
+
if (upgrade) {
|
1555
|
+
settings.call_on_message_complete(this);
|
1556
|
+
return true;
|
1557
|
+
}
|
1558
|
+
|
1559
|
+
if (0 != (flags & F_SKIPBODY)) {
|
1560
|
+
settings.call_on_message_complete(this);
|
1561
|
+
state = new_message();
|
1562
|
+
} else if (0 != (flags & F_CHUNKED)) {
|
1563
|
+
/* chunked encoding - ignore Content-Length header */
|
1564
|
+
state = State.chunk_size_start;
|
1565
|
+
} else {
|
1566
|
+
if (content_length == 0) {
|
1567
|
+
/* Content-Length header given but zero: Content-Length: 0\r\n */
|
1568
|
+
settings.call_on_message_complete(this);
|
1569
|
+
state = new_message();
|
1570
|
+
} else if (content_length > 0) {
|
1571
|
+
/* Content-Length header given and non-zero */
|
1572
|
+
state = State.body_identity;
|
1573
|
+
} else {
|
1574
|
+
if (type == ParserType.HTTP_REQUEST || http_should_keep_alive()) {
|
1575
|
+
/* Assume content-length 0 - read the next */
|
1576
|
+
settings.call_on_message_complete(this);
|
1577
|
+
state = new_message();
|
1578
|
+
} else {
|
1579
|
+
/* Read body until EOF */
|
1580
|
+
state = State.body_identity_eof;
|
1581
|
+
}
|
1582
|
+
}
|
1583
|
+
}
|
1584
|
+
return true;
|
1585
|
+
} // headers_almost_fone
|
1586
|
+
|
1587
|
+
|
1588
|
+
final int min (int a, int b) {
|
1589
|
+
return a < b ? a : b;
|
1590
|
+
}
|
1591
|
+
|
1592
|
+
/* probably not the best place to hide this ... */
|
1593
|
+
public boolean HTTP_PARSER_STRICT;
|
1594
|
+
State new_message() {
|
1595
|
+
if (HTTP_PARSER_STRICT){
|
1596
|
+
return http_should_keep_alive() ? start_state() : State.dead;
|
1597
|
+
} else {
|
1598
|
+
return start_state();
|
1599
|
+
}
|
1600
|
+
|
1601
|
+
}
|
1602
|
+
|
1603
|
+
State start_state() {
|
1604
|
+
return type == ParserType.HTTP_REQUEST ? State.start_req : State.start_res;
|
1605
|
+
}
|
1606
|
+
|
1607
|
+
|
1608
|
+
boolean parsing_header(State state) {
|
1609
|
+
|
1610
|
+
switch (state) {
|
1611
|
+
case chunk_size_start :
|
1612
|
+
case chunk_size :
|
1613
|
+
case chunk_size_almost_done :
|
1614
|
+
case chunk_parameters :
|
1615
|
+
case chunk_data :
|
1616
|
+
case chunk_data_almost_done :
|
1617
|
+
case chunk_data_done :
|
1618
|
+
case body_identity :
|
1619
|
+
case body_identity_eof :
|
1620
|
+
return false;
|
1621
|
+
|
1622
|
+
}
|
1623
|
+
return (0==(flags & F_TRAILING));
|
1624
|
+
}
|
1625
|
+
|
1626
|
+
/* "Dial C for Constants" */
|
1627
|
+
static class C {
|
1628
|
+
static final int HTTP_MAX_HEADER_SIZE = 80 * 1024;
|
1629
|
+
|
1630
|
+
static final int F_CHUNKED = 1 << 0;
|
1631
|
+
static final int F_CONNECTION_KEEP_ALIVE = 1 << 1;
|
1632
|
+
static final int F_CONNECTION_CLOSE = 1 << 2;
|
1633
|
+
static final int F_TRAILING = 1 << 3;
|
1634
|
+
static final int F_UPGRADE = 1 << 4;
|
1635
|
+
static final int F_SKIPBODY = 1 << 5;
|
1636
|
+
|
1637
|
+
static final byte [] UPCASE = {
|
1638
|
+
0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
|
1639
|
+
0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
|
1640
|
+
0x20,0x00,0x00,0x00,0x00,0x00,0x00,0x00, 0x00,0x00,0x00,0x00,0x00,0x2d,0x00,0x2f,
|
1641
|
+
0x30,0x31,0x32,0x33,0x34,0x35,0x36,0x37, 0x38,0x39,0x00,0x00,0x00,0x00,0x00,0x00,
|
1642
|
+
0x00,0x41,0x42,0x43,0x44,0x45,0x46,0x47, 0x48,0x49,0x4a,0x4b,0x4c,0x4d,0x4e,0x4f,
|
1643
|
+
0x50,0x51,0x52,0x53,0x54,0x55,0x56,0x57, 0x58,0x59,0x5a,0x00,0x00,0x00,0x00,0x5f,
|
1644
|
+
0x00,0x41,0x42,0x43,0x44,0x45,0x46,0x47, 0x48,0x49,0x4a,0x4b,0x4c,0x4d,0x4e,0x4f,
|
1645
|
+
0x50,0x51,0x52,0x53,0x54,0x55,0x56,0x57, 0x58,0x59,0x5a,0x00,0x00,0x00,0x00,0x00,
|
1646
|
+
0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
|
1647
|
+
0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
|
1648
|
+
0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
|
1649
|
+
0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
|
1650
|
+
0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
|
1651
|
+
0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
|
1652
|
+
0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
|
1653
|
+
0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
|
1654
|
+
};
|
1655
|
+
static final byte [] CONNECTION = {
|
1656
|
+
0x43, 0x4f, 0x4e, 0x4e, 0x45, 0x43, 0x54, 0x49, 0x4f, 0x4e,
|
1657
|
+
};
|
1658
|
+
static final byte [] PROXY_CONNECTION = {
|
1659
|
+
0x50, 0x52, 0x4f, 0x58, 0x59, 0x2d, 0x43, 0x4f, 0x4e, 0x4e, 0x45, 0x43, 0x54, 0x49, 0x4f, 0x4e,
|
1660
|
+
};
|
1661
|
+
static final byte [] CONTENT_LENGTH = {
|
1662
|
+
0x43, 0x4f, 0x4e, 0x54, 0x45, 0x4e, 0x54, 0x2d, 0x4c, 0x45, 0x4e, 0x47, 0x54, 0x48,
|
1663
|
+
};
|
1664
|
+
static final byte [] TRANSFER_ENCODING = {
|
1665
|
+
0x54, 0x52, 0x41, 0x4e, 0x53, 0x46, 0x45, 0x52, 0x2d, 0x45, 0x4e, 0x43, 0x4f, 0x44, 0x49, 0x4e, 0x47,
|
1666
|
+
};
|
1667
|
+
static final byte [] UPGRADE = {
|
1668
|
+
0x55, 0x50, 0x47, 0x52, 0x41, 0x44, 0x45,
|
1669
|
+
};
|
1670
|
+
static final byte [] CHUNKED = {
|
1671
|
+
0x43, 0x48, 0x55, 0x4e, 0x4b, 0x45, 0x44,
|
1672
|
+
};
|
1673
|
+
static final byte [] KEEP_ALIVE = {
|
1674
|
+
0x4b, 0x45, 0x45, 0x50, 0x2d, 0x41, 0x4c, 0x49, 0x56, 0x45,
|
1675
|
+
};
|
1676
|
+
static final byte [] CLOSE = {
|
1677
|
+
0x43, 0x4c, 0x4f, 0x53, 0x45,
|
1678
|
+
};
|
1679
|
+
/*
|
1680
|
+
* ' ', '_', '-' and all alpha-numeric ascii characters are accepted by acceptable_header.
|
1681
|
+
* The 'A'-'Z' are upper-cased.
|
1682
|
+
*/
|
1683
|
+
|
1684
|
+
static final char [] acceptable_header = {
|
1685
|
+
/* 0 nul 1 soh 2 stx 3 etx 4 eot 5 enq 6 ack 7 bel */
|
1686
|
+
0, 0, 0, 0, 0, 0, 0, 0,
|
1687
|
+
/* 8 bs 9 ht 10 nl 11 vt 12 np 13 cr 14 so 15 si */
|
1688
|
+
0, 0, 0, 0, 0, 0, 0, 0,
|
1689
|
+
/* 16 dle 17 dc1 18 dc2 19 dc3 20 dc4 21 nak 22 syn 23 etb */
|
1690
|
+
0, 0, 0, 0, 0, 0, 0, 0,
|
1691
|
+
/* 24 can 25 em 26 sub 27 esc 28 fs 29 gs 30 rs 31 us */
|
1692
|
+
0, 0, 0, 0, 0, 0, 0, 0,
|
1693
|
+
/* 32 sp 33 ! 34 " 35 # 36 $ 37 % 38 & 39 ' */
|
1694
|
+
' ', 0, 0, 0, 0, 0, 0, 0,
|
1695
|
+
/* 40 ( 41 ) 42 * 43 + 44 , 45 - 46 . 47 / */
|
1696
|
+
0, 0, 0, 0, 0, '-', 0, 0,
|
1697
|
+
/* 48 0 49 1 50 2 51 3 52 4 53 5 54 6 55 7 */
|
1698
|
+
'0', '1', '2', '3', '4', '5', '6', '7',
|
1699
|
+
/* 56 8 57 9 58 : 59 ; 60 < 61 = 62 > 63 ? */
|
1700
|
+
'8', '9', 0, 0, 0, 0, 0, 0,
|
1701
|
+
/* 64 @ 65 A 66 B 67 C 68 D 69 E 70 F 71 G */
|
1702
|
+
0, 'A', 'B', 'C', 'D', 'E', 'F', 'G',
|
1703
|
+
/* 72 H 73 I 74 J 75 K 76 L 77 M 78 N 79 O */
|
1704
|
+
'H', 'I', 'J', 'K', 'L', 'M', 'N', 'O',
|
1705
|
+
/* 80 P 81 Q 82 R 83 S 84 T 85 U 86 V 87 W */
|
1706
|
+
'P', 'Q', 'R', 'S', 'T', 'U', 'V', 'W',
|
1707
|
+
/* 88 X 89 Y 90 Z 91 [ 92 \ 93 ] 94 ^ 95 _ */
|
1708
|
+
'X', 'Y', 'Z', 0, 0, 0, 0, '_',
|
1709
|
+
/* 96 ` 97 a 98 b 99 c 100 d 101 e 102 f 103 g */
|
1710
|
+
0, 'A', 'B', 'C', 'D', 'E', 'F', 'G',
|
1711
|
+
/* 104 h 105 i 106 j 107 k 108 l 109 m 110 n 111 o */
|
1712
|
+
'H', 'I', 'J', 'K', 'L', 'M', 'N', 'O',
|
1713
|
+
/* 112 p 113 q 114 r 115 s 116 t 117 u 118 v 119 w */
|
1714
|
+
'P', 'Q', 'R', 'S', 'T', 'U', 'V', 'W',
|
1715
|
+
/* 120 x 121 y 122 z 123 { 124 | 125 } 126 ~ 127 del */
|
1716
|
+
'X', 'Y', 'Z', 0, 0, 0, 0, 0 };
|
1717
|
+
|
1718
|
+
static final byte [] UNHEX =
|
1719
|
+
{ -1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1
|
1720
|
+
,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1
|
1721
|
+
,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1
|
1722
|
+
, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9,-1,-1,-1,-1,-1,-1
|
1723
|
+
,-1,10,11,12,13,14,15,-1,-1,-1,-1,-1,-1,-1,-1,-1
|
1724
|
+
,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1
|
1725
|
+
,-1,10,11,12,13,14,15,-1,-1,-1,-1,-1,-1,-1,-1,-1
|
1726
|
+
,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1,-1
|
1727
|
+
};
|
1728
|
+
|
1729
|
+
static final boolean [] normal_url_char = {
|
1730
|
+
/* 0 nul 1 soh 2 stx 3 etx 4 eot 5 enq 6 ack 7 bel */
|
1731
|
+
false, false, false, false, false, false, false, false,
|
1732
|
+
/* 8 bs 9 ht 10 nl 11 vt 12 np 13 cr 14 so 15 si */
|
1733
|
+
false, false, false, false, false, false, false, false,
|
1734
|
+
/* 16 dle 17 dc1 18 dc2 19 dc3 20 dc4 21 nak 22 syn 23 etb */
|
1735
|
+
false, false, false, false, false, false, false, false,
|
1736
|
+
/* 24 can 25 em 26 sub 27 esc 28 fs 29 gs 30 rs 31 us */
|
1737
|
+
false, false, false, false, false, false, false, false,
|
1738
|
+
/* 32 sp 33 ! 34 " 35 # 36 $ 37 % 38 & 39 ' */
|
1739
|
+
false, true, true, false, true, true, true, true,
|
1740
|
+
/* 40 ( 41 ) 42 * 43 + 44 , 45 - 46 . 47 / */
|
1741
|
+
true, true, true, true, true, true, true, true,
|
1742
|
+
/* 48 0 49 1 50 2 51 3 52 4 53 5 54 6 55 7 */
|
1743
|
+
true, true, true, true, true, true, true, true,
|
1744
|
+
/* 56 8 57 9 58 : 59 ; 60 < 61 = 62 > 63 ? */
|
1745
|
+
true, true, true, true, true, true, true, false,
|
1746
|
+
/* 64 @ 65 A 66 B 67 C 68 D 69 E 70 F 71 G */
|
1747
|
+
true, true, true, true, true, true, true, true,
|
1748
|
+
/* 72 H 73 I 74 J 75 K 76 L 77 M 78 N 79 O */
|
1749
|
+
true, true, true, true, true, true, true, true,
|
1750
|
+
/* 80 P 81 Q 82 R 83 S 84 T 85 U 86 V 87 W */
|
1751
|
+
true, true, true, true, true, true, true, true,
|
1752
|
+
/* 88 X 89 Y 90 Z 91 [ 92 \ 93 ] 94 ^ 95 _ */
|
1753
|
+
true, true, true, true, true, true, true, true,
|
1754
|
+
/* 96 ` 97 a 98 b 99 c 100 d 101 e 102 f 103 g */
|
1755
|
+
true, true, true, true, true, true, true, true,
|
1756
|
+
/* 104 h 105 i 106 j 107 k 108 l 109 m 110 n 111 o */
|
1757
|
+
true, true, true, true, true, true, true, true,
|
1758
|
+
/* 112 p 113 q 114 r 115 s 116 t 117 u 118 v 119 w */
|
1759
|
+
true, true, true, true, true, true, true, true,
|
1760
|
+
/* 120 x 121 y 122 z 123 { 124 | 125 } 126 ~ 127 del */
|
1761
|
+
true, true, true, true, true, true, true, false };
|
1762
|
+
|
1763
|
+
public static final byte A = 0x41;
|
1764
|
+
public static final byte B = 0x42;
|
1765
|
+
public static final byte C = 0x43;
|
1766
|
+
public static final byte D = 0x44;
|
1767
|
+
public static final byte E = 0x45;
|
1768
|
+
public static final byte F = 0x46;
|
1769
|
+
public static final byte G = 0x47;
|
1770
|
+
public static final byte H = 0x48;
|
1771
|
+
public static final byte I = 0x49;
|
1772
|
+
public static final byte J = 0x4a;
|
1773
|
+
public static final byte K = 0x4b;
|
1774
|
+
public static final byte L = 0x4c;
|
1775
|
+
public static final byte M = 0x4d;
|
1776
|
+
public static final byte N = 0x4e;
|
1777
|
+
public static final byte O = 0x4f;
|
1778
|
+
public static final byte P = 0x50;
|
1779
|
+
public static final byte Q = 0x51;
|
1780
|
+
public static final byte R = 0x52;
|
1781
|
+
public static final byte S = 0x53;
|
1782
|
+
public static final byte T = 0x54;
|
1783
|
+
public static final byte U = 0x55;
|
1784
|
+
public static final byte V = 0x56;
|
1785
|
+
public static final byte W = 0x57;
|
1786
|
+
public static final byte X = 0x58;
|
1787
|
+
public static final byte Y = 0x59;
|
1788
|
+
public static final byte Z = 0x5a;
|
1789
|
+
public static final byte CR = 0x0d;
|
1790
|
+
public static final byte LF = 0x0a;
|
1791
|
+
public static final byte DOT = 0x2e;
|
1792
|
+
public static final byte SPACE = 0x20;
|
1793
|
+
public static final byte SEMI = 0x3b;
|
1794
|
+
public static final byte COLON = 0x3a;
|
1795
|
+
public static final byte HASH = 0x23;
|
1796
|
+
public static final byte QMARK = 0x3f;
|
1797
|
+
public static final byte SLASH = 0x2f;
|
1798
|
+
public static final byte DASH = 0x2d;
|
1799
|
+
public static final byte NULL = 0x00;
|
1800
|
+
}
|
1801
|
+
|
1802
|
+
enum State {
|
1803
|
+
|
1804
|
+
dead
|
1805
|
+
|
1806
|
+
, start_res_or_res
|
1807
|
+
, res_or_resp_H
|
1808
|
+
, start_res
|
1809
|
+
, res_H
|
1810
|
+
, res_HT
|
1811
|
+
, res_HTT
|
1812
|
+
, res_HTTP
|
1813
|
+
, res_first_http_major
|
1814
|
+
, res_http_major
|
1815
|
+
, res_first_http_minor
|
1816
|
+
, res_http_minor
|
1817
|
+
, res_first_status_code
|
1818
|
+
, res_status_code
|
1819
|
+
, res_status
|
1820
|
+
, res_line_almost_done
|
1821
|
+
|
1822
|
+
, start_req
|
1823
|
+
|
1824
|
+
, req_method
|
1825
|
+
, req_spaces_before_url
|
1826
|
+
, req_schema
|
1827
|
+
, req_schema_slash
|
1828
|
+
, req_schema_slash_slash
|
1829
|
+
, req_host
|
1830
|
+
, req_port
|
1831
|
+
, req_path
|
1832
|
+
, req_query_string_start
|
1833
|
+
, req_query_string
|
1834
|
+
, req_fragment_start
|
1835
|
+
, req_fragment
|
1836
|
+
, req_http_start
|
1837
|
+
, req_http_H
|
1838
|
+
, req_http_HT
|
1839
|
+
, req_http_HTT
|
1840
|
+
, req_http_HTTP
|
1841
|
+
, req_first_http_major
|
1842
|
+
, req_http_major
|
1843
|
+
, req_first_http_minor
|
1844
|
+
, req_http_minor
|
1845
|
+
, req_line_almost_done
|
1846
|
+
|
1847
|
+
, header_field_start
|
1848
|
+
, header_field
|
1849
|
+
, header_value_start
|
1850
|
+
, header_value
|
1851
|
+
|
1852
|
+
, header_almost_done
|
1853
|
+
|
1854
|
+
, headers_almost_done
|
1855
|
+
|
1856
|
+
, chunk_size_start
|
1857
|
+
, chunk_size
|
1858
|
+
, chunk_size_almost_done
|
1859
|
+
, chunk_parameters
|
1860
|
+
, chunk_data
|
1861
|
+
, chunk_data_almost_done
|
1862
|
+
, chunk_data_done
|
1863
|
+
|
1864
|
+
, body_identity
|
1865
|
+
, body_identity_eof;
|
1866
|
+
|
1867
|
+
|
1868
|
+
}
|
1869
|
+
enum HState {
|
1870
|
+
general
|
1871
|
+
, C
|
1872
|
+
, CO
|
1873
|
+
, CON
|
1874
|
+
|
1875
|
+
, matching_connection
|
1876
|
+
, matching_proxy_connection
|
1877
|
+
, matching_content_length
|
1878
|
+
, matching_transfer_encoding
|
1879
|
+
, matching_upgrade
|
1880
|
+
|
1881
|
+
, connection
|
1882
|
+
, content_length
|
1883
|
+
, transfer_encoding
|
1884
|
+
, upgrade
|
1885
|
+
|
1886
|
+
, matching_transfer_encoding_chunked
|
1887
|
+
, matching_connection_keep_alive
|
1888
|
+
, matching_connection_close
|
1889
|
+
|
1890
|
+
, transfer_encoding_chunked
|
1891
|
+
, connection_keep_alive
|
1892
|
+
, connection_close
|
1893
|
+
}
|
1894
|
+
}
|