threatstack-agent-ruby 0.2.1
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +7 -0
- data/Gemfile +3 -0
- data/LICENSE +6 -0
- data/ext/libinjection/extconf.rb +4 -0
- data/ext/libinjection/libinjection.h +65 -0
- data/ext/libinjection/libinjection.i +13 -0
- data/ext/libinjection/libinjection_html5.c +850 -0
- data/ext/libinjection/libinjection_html5.h +54 -0
- data/ext/libinjection/libinjection_sqli.c +2325 -0
- data/ext/libinjection/libinjection_sqli.h +298 -0
- data/ext/libinjection/libinjection_sqli_data.h +9654 -0
- data/ext/libinjection/libinjection_wrap.c +2393 -0
- data/ext/libinjection/libinjection_xss.c +532 -0
- data/ext/libinjection/libinjection_xss.h +21 -0
- data/lib/constants.rb +110 -0
- data/lib/control.rb +61 -0
- data/lib/events/event_accumulator.rb +36 -0
- data/lib/events/models/attack_event.rb +58 -0
- data/lib/events/models/base_event.rb +41 -0
- data/lib/events/models/dependency_event.rb +93 -0
- data/lib/events/models/environment_event.rb +93 -0
- data/lib/events/models/instrumentation_event.rb +46 -0
- data/lib/exceptions/request_blocked_error.rb +11 -0
- data/lib/instrumentation/common.rb +172 -0
- data/lib/instrumentation/instrumenter.rb +144 -0
- data/lib/instrumentation/kernel.rb +45 -0
- data/lib/instrumentation/rails.rb +61 -0
- data/lib/jobs/delayed_job.rb +26 -0
- data/lib/jobs/event_submitter.rb +101 -0
- data/lib/jobs/job_queue.rb +38 -0
- data/lib/jobs/recurrent_job.rb +61 -0
- data/lib/threatstack-agent-ruby.rb +7 -0
- data/lib/utils/aws_utils.rb +46 -0
- data/lib/utils/formatter.rb +47 -0
- data/lib/utils/logger.rb +43 -0
- data/threatstack-agent-ruby.gemspec +35 -0
- metadata +221 -0
checksums.yaml
ADDED
@@ -0,0 +1,7 @@
|
|
1
|
+
---
|
2
|
+
SHA256:
|
3
|
+
metadata.gz: d56612a7470c1f394b66c6473479883be0ddc44810bebd77206acd42da1b31a6
|
4
|
+
data.tar.gz: 0d1bbad69a0ef6bb2ab53b662a8f004d06d353b2b02746ffb5616731eaef1c90
|
5
|
+
SHA512:
|
6
|
+
metadata.gz: 0c98e82db4c39fbfd7d930049b9ffe84c6efeda17e5f6b97ea971e240ffaec39c45683630af92ba95626a6571e0d9ecc02d01431c1c70284959b240576bb3e03
|
7
|
+
data.tar.gz: 767fb9d9a56bb8b4efdd9d7d9a28a373e86851550ba8c9e4401d4b2803d6bda1964ba0f47b566c25c12b54f8db5df49c48249bc12734f29f6dcb57a7e706879c
|
data/Gemfile
ADDED
data/LICENSE
ADDED
@@ -0,0 +1,6 @@
|
|
1
|
+
Threat Stack AppSec for Ruby is free-to-use, proprietary software.
|
2
|
+
Please refer to our terms for more information: https://www.threatstack.com/terms
|
3
|
+
|
4
|
+
The Threat Stack Ruby AppSec agent also uses code from the following open source projects under the following licenses:
|
5
|
+
|
6
|
+
libinjection http://opensource.org/licenses/BSD-3-Clause
|
@@ -0,0 +1,65 @@
|
|
1
|
+
/**
|
2
|
+
* Copyright 2012-2016 Nick Galbreath
|
3
|
+
* nickg@client9.com
|
4
|
+
* BSD License -- see COPYING.txt for details
|
5
|
+
*
|
6
|
+
* https://libinjection.client9.com/
|
7
|
+
*
|
8
|
+
*/
|
9
|
+
|
10
|
+
#ifndef LIBINJECTION_H
|
11
|
+
#define LIBINJECTION_H
|
12
|
+
|
13
|
+
#ifdef __cplusplus
|
14
|
+
# define LIBINJECTION_BEGIN_DECLS extern "C" {
|
15
|
+
# define LIBINJECTION_END_DECLS }
|
16
|
+
#else
|
17
|
+
# define LIBINJECTION_BEGIN_DECLS
|
18
|
+
# define LIBINJECTION_END_DECLS
|
19
|
+
#endif
|
20
|
+
|
21
|
+
LIBINJECTION_BEGIN_DECLS
|
22
|
+
|
23
|
+
/*
|
24
|
+
* Pull in size_t
|
25
|
+
*/
|
26
|
+
#include <string.h>
|
27
|
+
|
28
|
+
/*
|
29
|
+
* Version info.
|
30
|
+
*
|
31
|
+
* This is moved into a function to allow SWIG and other auto-generated
|
32
|
+
* binding to not be modified during minor release changes. We change
|
33
|
+
* change the version number in the c source file, and not regenerated
|
34
|
+
* the binding
|
35
|
+
*
|
36
|
+
* See python's normalized version
|
37
|
+
* http://www.python.org/dev/peps/pep-0386/#normalizedversion
|
38
|
+
*/
|
39
|
+
const char* libinjection_version(void);
|
40
|
+
|
41
|
+
/**
|
42
|
+
* Simple API for SQLi detection - returns a SQLi fingerprint or NULL
|
43
|
+
* is benign input
|
44
|
+
*
|
45
|
+
* \param[in] s input string, may contain nulls, does not need to be null-terminated
|
46
|
+
* \param[in] slen input string length
|
47
|
+
* \param[out] fingerprint buffer of 8+ characters. c-string,
|
48
|
+
* \return 1 if SQLi, 0 if benign. fingerprint will be set or set to empty string.
|
49
|
+
*/
|
50
|
+
int libinjection_sqli(const char* s, size_t slen, char fingerprint[]);
|
51
|
+
|
52
|
+
/** ALPHA version of xss detector.
|
53
|
+
*
|
54
|
+
* NOT DONE.
|
55
|
+
*
|
56
|
+
* \param[in] s input string, may contain nulls, does not need to be null-terminated
|
57
|
+
* \param[in] slen input string length
|
58
|
+
* \return 1 if XSS found, 0 if benign
|
59
|
+
*
|
60
|
+
*/
|
61
|
+
int libinjection_xss(const char* s, size_t slen);
|
62
|
+
|
63
|
+
LIBINJECTION_END_DECLS
|
64
|
+
|
65
|
+
#endif /* LIBINJECTION_H */
|
@@ -0,0 +1,13 @@
|
|
1
|
+
/* libinjection.i SWIG interface file */
|
2
|
+
|
3
|
+
%module libinjection
|
4
|
+
%{
|
5
|
+
#include "libinjection.h"
|
6
|
+
#include "libinjection_sqli.h"
|
7
|
+
#include "libinjection_xss.h"
|
8
|
+
%}
|
9
|
+
|
10
|
+
%include "libinjection.h"
|
11
|
+
|
12
|
+
// int libinjection_sqli(char *, size_t, char *OUTPUT);
|
13
|
+
// int libinjection_xss(char *, size_t);
|
@@ -0,0 +1,850 @@
|
|
1
|
+
#include "libinjection_html5.h"
|
2
|
+
|
3
|
+
#include <string.h>
|
4
|
+
#include <assert.h>
|
5
|
+
|
6
|
+
#ifdef DEBUG
|
7
|
+
#include <stdio.h>
|
8
|
+
#define TRACE() printf("%s:%d\n", __FUNCTION__, __LINE__)
|
9
|
+
#else
|
10
|
+
#define TRACE()
|
11
|
+
#endif
|
12
|
+
|
13
|
+
|
14
|
+
#define CHAR_EOF -1
|
15
|
+
#define CHAR_NULL 0
|
16
|
+
#define CHAR_BANG 33
|
17
|
+
#define CHAR_DOUBLE 34
|
18
|
+
#define CHAR_PERCENT 37
|
19
|
+
#define CHAR_SINGLE 39
|
20
|
+
#define CHAR_DASH 45
|
21
|
+
#define CHAR_SLASH 47
|
22
|
+
#define CHAR_LT 60
|
23
|
+
#define CHAR_EQUALS 61
|
24
|
+
#define CHAR_GT 62
|
25
|
+
#define CHAR_QUESTION 63
|
26
|
+
#define CHAR_RIGHTB 93
|
27
|
+
#define CHAR_TICK 96
|
28
|
+
|
29
|
+
/* prototypes */
|
30
|
+
|
31
|
+
static int h5_skip_white(h5_state_t* hs);
|
32
|
+
static int h5_is_white(char c);
|
33
|
+
static int h5_state_eof(h5_state_t* hs);
|
34
|
+
static int h5_state_data(h5_state_t* hs);
|
35
|
+
static int h5_state_tag_open(h5_state_t* hs);
|
36
|
+
static int h5_state_tag_name(h5_state_t* hs);
|
37
|
+
static int h5_state_tag_name_close(h5_state_t* hs);
|
38
|
+
static int h5_state_end_tag_open(h5_state_t* hs);
|
39
|
+
static int h5_state_self_closing_start_tag(h5_state_t* hs);
|
40
|
+
static int h5_state_attribute_name(h5_state_t* hs);
|
41
|
+
static int h5_state_after_attribute_name(h5_state_t* hs);
|
42
|
+
static int h5_state_before_attribute_name(h5_state_t* hs);
|
43
|
+
static int h5_state_before_attribute_value(h5_state_t* hs);
|
44
|
+
static int h5_state_attribute_value_double_quote(h5_state_t* hs);
|
45
|
+
static int h5_state_attribute_value_single_quote(h5_state_t* hs);
|
46
|
+
static int h5_state_attribute_value_back_quote(h5_state_t* hs);
|
47
|
+
static int h5_state_attribute_value_no_quote(h5_state_t* hs);
|
48
|
+
static int h5_state_after_attribute_value_quoted_state(h5_state_t* hs);
|
49
|
+
static int h5_state_comment(h5_state_t* hs);
|
50
|
+
static int h5_state_cdata(h5_state_t* hs);
|
51
|
+
|
52
|
+
|
53
|
+
/* 12.2.4.44 */
|
54
|
+
static int h5_state_bogus_comment(h5_state_t* hs);
|
55
|
+
static int h5_state_bogus_comment2(h5_state_t* hs);
|
56
|
+
|
57
|
+
/* 12.2.4.45 */
|
58
|
+
static int h5_state_markup_declaration_open(h5_state_t* hs);
|
59
|
+
|
60
|
+
/* 8.2.4.52 */
|
61
|
+
static int h5_state_doctype(h5_state_t* hs);
|
62
|
+
|
63
|
+
/**
|
64
|
+
* public function
|
65
|
+
*/
|
66
|
+
void libinjection_h5_init(h5_state_t* hs, const char* s, size_t len, enum html5_flags flags)
|
67
|
+
{
|
68
|
+
memset(hs, 0, sizeof(h5_state_t));
|
69
|
+
hs->s = s;
|
70
|
+
hs->len = len;
|
71
|
+
|
72
|
+
switch (flags) {
|
73
|
+
case DATA_STATE:
|
74
|
+
hs->state = h5_state_data;
|
75
|
+
break;
|
76
|
+
case VALUE_NO_QUOTE:
|
77
|
+
hs->state = h5_state_before_attribute_name;
|
78
|
+
break;
|
79
|
+
case VALUE_SINGLE_QUOTE:
|
80
|
+
hs->state = h5_state_attribute_value_single_quote;
|
81
|
+
break;
|
82
|
+
case VALUE_DOUBLE_QUOTE:
|
83
|
+
hs->state = h5_state_attribute_value_double_quote;
|
84
|
+
break;
|
85
|
+
case VALUE_BACK_QUOTE:
|
86
|
+
hs->state = h5_state_attribute_value_back_quote;
|
87
|
+
break;
|
88
|
+
}
|
89
|
+
}
|
90
|
+
|
91
|
+
/**
|
92
|
+
* public function
|
93
|
+
*/
|
94
|
+
int libinjection_h5_next(h5_state_t* hs)
|
95
|
+
{
|
96
|
+
assert(hs->state != NULL);
|
97
|
+
return (*hs->state)(hs);
|
98
|
+
}
|
99
|
+
|
100
|
+
/**
|
101
|
+
* Everything below here is private
|
102
|
+
*
|
103
|
+
*/
|
104
|
+
|
105
|
+
|
106
|
+
static int h5_is_white(char ch)
|
107
|
+
{
|
108
|
+
/*
|
109
|
+
* \t = horizontal tab = 0x09
|
110
|
+
* \n = newline = 0x0A
|
111
|
+
* \v = vertical tab = 0x0B
|
112
|
+
* \f = form feed = 0x0C
|
113
|
+
* \r = cr = 0x0D
|
114
|
+
*/
|
115
|
+
return strchr(" \t\n\v\f\r", ch) != NULL;
|
116
|
+
}
|
117
|
+
|
118
|
+
static int h5_skip_white(h5_state_t* hs)
|
119
|
+
{
|
120
|
+
char ch;
|
121
|
+
while (hs->pos < hs->len) {
|
122
|
+
ch = hs->s[hs->pos];
|
123
|
+
switch (ch) {
|
124
|
+
case 0x00: /* IE only */
|
125
|
+
case 0x20:
|
126
|
+
case 0x09:
|
127
|
+
case 0x0A:
|
128
|
+
case 0x0B: /* IE only */
|
129
|
+
case 0x0C:
|
130
|
+
case 0x0D: /* IE only */
|
131
|
+
hs->pos += 1;
|
132
|
+
break;
|
133
|
+
default:
|
134
|
+
return ch;
|
135
|
+
}
|
136
|
+
}
|
137
|
+
return CHAR_EOF;
|
138
|
+
}
|
139
|
+
|
140
|
+
static int h5_state_eof(h5_state_t* hs)
|
141
|
+
{
|
142
|
+
/* eliminate unused function argument warning */
|
143
|
+
(void)hs;
|
144
|
+
return 0;
|
145
|
+
}
|
146
|
+
|
147
|
+
static int h5_state_data(h5_state_t* hs)
|
148
|
+
{
|
149
|
+
const char* idx;
|
150
|
+
|
151
|
+
TRACE();
|
152
|
+
assert(hs->len >= hs->pos);
|
153
|
+
idx = (const char*) memchr(hs->s + hs->pos, CHAR_LT, hs->len - hs->pos);
|
154
|
+
if (idx == NULL) {
|
155
|
+
hs->token_start = hs->s + hs->pos;
|
156
|
+
hs->token_len = hs->len - hs->pos;
|
157
|
+
hs->token_type = DATA_TEXT;
|
158
|
+
hs->state = h5_state_eof;
|
159
|
+
if (hs->token_len == 0) {
|
160
|
+
return 0;
|
161
|
+
}
|
162
|
+
} else {
|
163
|
+
hs->token_start = hs->s + hs->pos;
|
164
|
+
hs->token_type = DATA_TEXT;
|
165
|
+
hs->token_len = (size_t)(idx - hs->s) - hs->pos;
|
166
|
+
hs->pos = (size_t)(idx - hs->s) + 1;
|
167
|
+
hs->state = h5_state_tag_open;
|
168
|
+
if (hs->token_len == 0) {
|
169
|
+
return h5_state_tag_open(hs);
|
170
|
+
}
|
171
|
+
}
|
172
|
+
return 1;
|
173
|
+
}
|
174
|
+
|
175
|
+
/**
|
176
|
+
* 12 2.4.8
|
177
|
+
*/
|
178
|
+
static int h5_state_tag_open(h5_state_t* hs)
|
179
|
+
{
|
180
|
+
char ch;
|
181
|
+
|
182
|
+
TRACE();
|
183
|
+
if (hs->pos >= hs->len) {
|
184
|
+
return 0;
|
185
|
+
}
|
186
|
+
ch = hs->s[hs->pos];
|
187
|
+
if (ch == CHAR_BANG) {
|
188
|
+
hs->pos += 1;
|
189
|
+
return h5_state_markup_declaration_open(hs);
|
190
|
+
} else if (ch == CHAR_SLASH) {
|
191
|
+
hs->pos += 1;
|
192
|
+
hs->is_close = 1;
|
193
|
+
return h5_state_end_tag_open(hs);
|
194
|
+
} else if (ch == CHAR_QUESTION) {
|
195
|
+
hs->pos += 1;
|
196
|
+
return h5_state_bogus_comment(hs);
|
197
|
+
} else if (ch == CHAR_PERCENT) {
|
198
|
+
/* this is not in spec.. alternative comment format used
|
199
|
+
by IE <= 9 and Safari < 4.0.3 */
|
200
|
+
hs->pos += 1;
|
201
|
+
return h5_state_bogus_comment2(hs);
|
202
|
+
} else if ((ch >= 'a' && ch <= 'z') || (ch >= 'A' && ch <= 'Z')) {
|
203
|
+
return h5_state_tag_name(hs);
|
204
|
+
} else if (ch == CHAR_NULL) {
|
205
|
+
/* IE-ism NULL characters are ignored */
|
206
|
+
return h5_state_tag_name(hs);
|
207
|
+
} else {
|
208
|
+
/* user input mistake in configuring state */
|
209
|
+
if (hs->pos == 0) {
|
210
|
+
return h5_state_data(hs);
|
211
|
+
}
|
212
|
+
hs->token_start = hs->s + hs->pos - 1;
|
213
|
+
hs->token_len = 1;
|
214
|
+
hs->token_type = DATA_TEXT;
|
215
|
+
hs->state = h5_state_data;
|
216
|
+
return 1;
|
217
|
+
}
|
218
|
+
}
|
219
|
+
/**
|
220
|
+
* 12.2.4.9
|
221
|
+
*/
|
222
|
+
static int h5_state_end_tag_open(h5_state_t* hs)
|
223
|
+
{
|
224
|
+
char ch;
|
225
|
+
|
226
|
+
TRACE();
|
227
|
+
|
228
|
+
if (hs->pos >= hs->len) {
|
229
|
+
return 0;
|
230
|
+
}
|
231
|
+
ch = hs->s[hs->pos];
|
232
|
+
if (ch == CHAR_GT) {
|
233
|
+
return h5_state_data(hs);
|
234
|
+
} else if ((ch >= 'a' && ch <= 'z') || (ch >= 'A' && ch <= 'Z')) {
|
235
|
+
return h5_state_tag_name(hs);
|
236
|
+
}
|
237
|
+
|
238
|
+
hs->is_close = 0;
|
239
|
+
return h5_state_bogus_comment(hs);
|
240
|
+
}
|
241
|
+
/*
|
242
|
+
*
|
243
|
+
*/
|
244
|
+
static int h5_state_tag_name_close(h5_state_t* hs)
|
245
|
+
{
|
246
|
+
TRACE();
|
247
|
+
hs->is_close = 0;
|
248
|
+
hs->token_start = hs->s + hs->pos;
|
249
|
+
hs->token_len = 1;
|
250
|
+
hs->token_type = TAG_NAME_CLOSE;
|
251
|
+
hs->pos += 1;
|
252
|
+
if (hs->pos < hs->len) {
|
253
|
+
hs->state = h5_state_data;
|
254
|
+
} else {
|
255
|
+
hs->state = h5_state_eof;
|
256
|
+
}
|
257
|
+
|
258
|
+
return 1;
|
259
|
+
}
|
260
|
+
|
261
|
+
/**
|
262
|
+
* 12.2.4.10
|
263
|
+
*/
|
264
|
+
static int h5_state_tag_name(h5_state_t* hs)
|
265
|
+
{
|
266
|
+
char ch;
|
267
|
+
size_t pos;
|
268
|
+
|
269
|
+
TRACE();
|
270
|
+
pos = hs->pos;
|
271
|
+
while (pos < hs->len) {
|
272
|
+
ch = hs->s[pos];
|
273
|
+
if (ch == 0) {
|
274
|
+
/* special non-standard case */
|
275
|
+
/* allow nulls in tag name */
|
276
|
+
/* some old browsers apparently allow and ignore them */
|
277
|
+
pos += 1;
|
278
|
+
} else if (h5_is_white(ch)) {
|
279
|
+
hs->token_start = hs->s + hs->pos;
|
280
|
+
hs->token_len = pos - hs->pos;
|
281
|
+
hs->token_type = TAG_NAME_OPEN;
|
282
|
+
hs->pos = pos + 1;
|
283
|
+
hs->state = h5_state_before_attribute_name;
|
284
|
+
return 1;
|
285
|
+
} else if (ch == CHAR_SLASH) {
|
286
|
+
hs->token_start = hs->s + hs->pos;
|
287
|
+
hs->token_len = pos - hs->pos;
|
288
|
+
hs->token_type = TAG_NAME_OPEN;
|
289
|
+
hs->pos = pos + 1;
|
290
|
+
hs->state = h5_state_self_closing_start_tag;
|
291
|
+
return 1;
|
292
|
+
} else if (ch == CHAR_GT) {
|
293
|
+
hs->token_start = hs->s + hs->pos;
|
294
|
+
hs->token_len = pos - hs->pos;
|
295
|
+
if (hs->is_close) {
|
296
|
+
hs->pos = pos + 1;
|
297
|
+
hs->is_close = 0;
|
298
|
+
hs->token_type = TAG_CLOSE;
|
299
|
+
hs->state = h5_state_data;
|
300
|
+
} else {
|
301
|
+
hs->pos = pos;
|
302
|
+
hs->token_type = TAG_NAME_OPEN;
|
303
|
+
hs->state = h5_state_tag_name_close;
|
304
|
+
}
|
305
|
+
return 1;
|
306
|
+
} else {
|
307
|
+
pos += 1;
|
308
|
+
}
|
309
|
+
}
|
310
|
+
|
311
|
+
hs->token_start = hs->s + hs->pos;
|
312
|
+
hs->token_len = hs->len - hs->pos;
|
313
|
+
hs->token_type = TAG_NAME_OPEN;
|
314
|
+
hs->state = h5_state_eof;
|
315
|
+
return 1;
|
316
|
+
}
|
317
|
+
|
318
|
+
/**
|
319
|
+
* 12.2.4.34
|
320
|
+
*/
|
321
|
+
static int h5_state_before_attribute_name(h5_state_t* hs)
|
322
|
+
{
|
323
|
+
int ch;
|
324
|
+
|
325
|
+
TRACE();
|
326
|
+
ch = h5_skip_white(hs);
|
327
|
+
switch (ch) {
|
328
|
+
case CHAR_EOF: {
|
329
|
+
return 0;
|
330
|
+
}
|
331
|
+
case CHAR_SLASH: {
|
332
|
+
hs->pos += 1;
|
333
|
+
return h5_state_self_closing_start_tag(hs);
|
334
|
+
}
|
335
|
+
case CHAR_GT: {
|
336
|
+
hs->state = h5_state_data;
|
337
|
+
hs->token_start = hs->s + hs->pos;
|
338
|
+
hs->token_len = 1;
|
339
|
+
hs->token_type = TAG_NAME_CLOSE;
|
340
|
+
hs->pos += 1;
|
341
|
+
return 1;
|
342
|
+
}
|
343
|
+
default: {
|
344
|
+
return h5_state_attribute_name(hs);
|
345
|
+
}
|
346
|
+
}
|
347
|
+
}
|
348
|
+
|
349
|
+
static int h5_state_attribute_name(h5_state_t* hs)
|
350
|
+
{
|
351
|
+
char ch;
|
352
|
+
size_t pos;
|
353
|
+
|
354
|
+
TRACE();
|
355
|
+
pos = hs->pos + 1;
|
356
|
+
while (pos < hs->len) {
|
357
|
+
ch = hs->s[pos];
|
358
|
+
if (h5_is_white(ch)) {
|
359
|
+
hs->token_start = hs->s + hs->pos;
|
360
|
+
hs->token_len = pos - hs->pos;
|
361
|
+
hs->token_type = ATTR_NAME;
|
362
|
+
hs->state = h5_state_after_attribute_name;
|
363
|
+
hs->pos = pos + 1;
|
364
|
+
return 1;
|
365
|
+
} else if (ch == CHAR_SLASH) {
|
366
|
+
hs->token_start = hs->s + hs->pos;
|
367
|
+
hs->token_len = pos - hs->pos;
|
368
|
+
hs->token_type = ATTR_NAME;
|
369
|
+
hs->state = h5_state_self_closing_start_tag;
|
370
|
+
hs->pos = pos + 1;
|
371
|
+
return 1;
|
372
|
+
} else if (ch == CHAR_EQUALS) {
|
373
|
+
hs->token_start = hs->s + hs->pos;
|
374
|
+
hs->token_len = pos - hs->pos;
|
375
|
+
hs->token_type = ATTR_NAME;
|
376
|
+
hs->state = h5_state_before_attribute_value;
|
377
|
+
hs->pos = pos + 1;
|
378
|
+
return 1;
|
379
|
+
} else if (ch == CHAR_GT) {
|
380
|
+
hs->token_start = hs->s + hs->pos;
|
381
|
+
hs->token_len = pos - hs->pos;
|
382
|
+
hs->token_type = ATTR_NAME;
|
383
|
+
hs->state = h5_state_tag_name_close;
|
384
|
+
hs->pos = pos;
|
385
|
+
return 1;
|
386
|
+
} else {
|
387
|
+
pos += 1;
|
388
|
+
}
|
389
|
+
}
|
390
|
+
/* EOF */
|
391
|
+
hs->token_start = hs->s + hs->pos;
|
392
|
+
hs->token_len = hs->len - hs->pos;
|
393
|
+
hs->token_type = ATTR_NAME;
|
394
|
+
hs->state = h5_state_eof;
|
395
|
+
hs->pos = hs->len;
|
396
|
+
return 1;
|
397
|
+
}
|
398
|
+
|
399
|
+
/**
|
400
|
+
* 12.2.4.36
|
401
|
+
*/
|
402
|
+
static int h5_state_after_attribute_name(h5_state_t* hs)
|
403
|
+
{
|
404
|
+
int c;
|
405
|
+
|
406
|
+
TRACE();
|
407
|
+
c = h5_skip_white(hs);
|
408
|
+
switch (c) {
|
409
|
+
case CHAR_EOF: {
|
410
|
+
return 0;
|
411
|
+
}
|
412
|
+
case CHAR_SLASH: {
|
413
|
+
hs->pos += 1;
|
414
|
+
return h5_state_self_closing_start_tag(hs);
|
415
|
+
}
|
416
|
+
case CHAR_EQUALS: {
|
417
|
+
hs->pos += 1;
|
418
|
+
return h5_state_before_attribute_value(hs);
|
419
|
+
}
|
420
|
+
case CHAR_GT: {
|
421
|
+
return h5_state_tag_name_close(hs);
|
422
|
+
}
|
423
|
+
default: {
|
424
|
+
return h5_state_attribute_name(hs);
|
425
|
+
}
|
426
|
+
}
|
427
|
+
}
|
428
|
+
|
429
|
+
/**
|
430
|
+
* 12.2.4.37
|
431
|
+
*/
|
432
|
+
static int h5_state_before_attribute_value(h5_state_t* hs)
|
433
|
+
{
|
434
|
+
int c;
|
435
|
+
TRACE();
|
436
|
+
|
437
|
+
c = h5_skip_white(hs);
|
438
|
+
|
439
|
+
if (c == CHAR_EOF) {
|
440
|
+
hs->state = h5_state_eof;
|
441
|
+
return 0;
|
442
|
+
}
|
443
|
+
|
444
|
+
if (c == CHAR_DOUBLE) {
|
445
|
+
return h5_state_attribute_value_double_quote(hs);
|
446
|
+
} else if (c == CHAR_SINGLE) {
|
447
|
+
return h5_state_attribute_value_single_quote(hs);
|
448
|
+
} else if (c == CHAR_TICK) {
|
449
|
+
/* NON STANDARD IE */
|
450
|
+
return h5_state_attribute_value_back_quote(hs);
|
451
|
+
} else {
|
452
|
+
return h5_state_attribute_value_no_quote(hs);
|
453
|
+
}
|
454
|
+
}
|
455
|
+
|
456
|
+
|
457
|
+
static int h5_state_attribute_value_quote(h5_state_t* hs, char qchar)
|
458
|
+
{
|
459
|
+
const char* idx;
|
460
|
+
|
461
|
+
TRACE();
|
462
|
+
|
463
|
+
/* skip initial quote in normal case.
|
464
|
+
* don't do this "if (pos == 0)" since it means we have started
|
465
|
+
* in a non-data state. given an input of '><foo
|
466
|
+
* we want to make 0-length attribute name
|
467
|
+
*/
|
468
|
+
if (hs->pos > 0) {
|
469
|
+
hs->pos += 1;
|
470
|
+
}
|
471
|
+
|
472
|
+
|
473
|
+
idx = (const char*) memchr(hs->s + hs->pos, qchar, hs->len - hs->pos);
|
474
|
+
if (idx == NULL) {
|
475
|
+
hs->token_start = hs->s + hs->pos;
|
476
|
+
hs->token_len = hs->len - hs->pos;
|
477
|
+
hs->token_type = ATTR_VALUE;
|
478
|
+
hs->state = h5_state_eof;
|
479
|
+
} else {
|
480
|
+
hs->token_start = hs->s + hs->pos;
|
481
|
+
hs->token_len = (size_t)(idx - hs->s) - hs->pos;
|
482
|
+
hs->token_type = ATTR_VALUE;
|
483
|
+
hs->state = h5_state_after_attribute_value_quoted_state;
|
484
|
+
hs->pos += hs->token_len + 1;
|
485
|
+
}
|
486
|
+
return 1;
|
487
|
+
}
|
488
|
+
|
489
|
+
static
|
490
|
+
int h5_state_attribute_value_double_quote(h5_state_t* hs)
|
491
|
+
{
|
492
|
+
TRACE();
|
493
|
+
return h5_state_attribute_value_quote(hs, CHAR_DOUBLE);
|
494
|
+
}
|
495
|
+
|
496
|
+
static
|
497
|
+
int h5_state_attribute_value_single_quote(h5_state_t* hs)
|
498
|
+
{
|
499
|
+
TRACE();
|
500
|
+
return h5_state_attribute_value_quote(hs, CHAR_SINGLE);
|
501
|
+
}
|
502
|
+
|
503
|
+
static
|
504
|
+
int h5_state_attribute_value_back_quote(h5_state_t* hs)
|
505
|
+
{
|
506
|
+
TRACE();
|
507
|
+
return h5_state_attribute_value_quote(hs, CHAR_TICK);
|
508
|
+
}
|
509
|
+
|
510
|
+
static int h5_state_attribute_value_no_quote(h5_state_t* hs)
|
511
|
+
{
|
512
|
+
char ch;
|
513
|
+
size_t pos;
|
514
|
+
|
515
|
+
TRACE();
|
516
|
+
pos = hs->pos;
|
517
|
+
while (pos < hs->len) {
|
518
|
+
ch = hs->s[pos];
|
519
|
+
if (h5_is_white(ch)) {
|
520
|
+
hs->token_type = ATTR_VALUE;
|
521
|
+
hs->token_start = hs->s + hs->pos;
|
522
|
+
hs->token_len = pos - hs->pos;
|
523
|
+
hs->pos = pos + 1;
|
524
|
+
hs->state = h5_state_before_attribute_name;
|
525
|
+
return 1;
|
526
|
+
} else if (ch == CHAR_GT) {
|
527
|
+
hs->token_type = ATTR_VALUE;
|
528
|
+
hs->token_start = hs->s + hs->pos;
|
529
|
+
hs->token_len = pos - hs->pos;
|
530
|
+
hs->pos = pos;
|
531
|
+
hs->state = h5_state_tag_name_close;
|
532
|
+
return 1;
|
533
|
+
}
|
534
|
+
pos += 1;
|
535
|
+
}
|
536
|
+
TRACE();
|
537
|
+
/* EOF */
|
538
|
+
hs->state = h5_state_eof;
|
539
|
+
hs->token_start = hs->s + hs->pos;
|
540
|
+
hs->token_len = hs->len - hs->pos;
|
541
|
+
hs->token_type = ATTR_VALUE;
|
542
|
+
return 1;
|
543
|
+
}
|
544
|
+
|
545
|
+
/**
|
546
|
+
* 12.2.4.41
|
547
|
+
*/
|
548
|
+
static int h5_state_after_attribute_value_quoted_state(h5_state_t* hs)
|
549
|
+
{
|
550
|
+
char ch;
|
551
|
+
|
552
|
+
TRACE();
|
553
|
+
if (hs->pos >= hs->len) {
|
554
|
+
return 0;
|
555
|
+
}
|
556
|
+
ch = hs->s[hs->pos];
|
557
|
+
if (h5_is_white(ch)) {
|
558
|
+
hs->pos += 1;
|
559
|
+
return h5_state_before_attribute_name(hs);
|
560
|
+
} else if (ch == CHAR_SLASH) {
|
561
|
+
hs->pos += 1;
|
562
|
+
return h5_state_self_closing_start_tag(hs);
|
563
|
+
} else if (ch == CHAR_GT) {
|
564
|
+
hs->token_start = hs->s + hs->pos;
|
565
|
+
hs->token_len = 1;
|
566
|
+
hs->token_type = TAG_NAME_CLOSE;
|
567
|
+
hs->pos += 1;
|
568
|
+
hs->state = h5_state_data;
|
569
|
+
return 1;
|
570
|
+
} else {
|
571
|
+
return h5_state_before_attribute_name(hs);
|
572
|
+
}
|
573
|
+
}
|
574
|
+
|
575
|
+
/**
|
576
|
+
* 12.2.4.43
|
577
|
+
*/
|
578
|
+
static int h5_state_self_closing_start_tag(h5_state_t* hs)
|
579
|
+
{
|
580
|
+
char ch;
|
581
|
+
|
582
|
+
TRACE();
|
583
|
+
if (hs->pos >= hs->len) {
|
584
|
+
return 0;
|
585
|
+
}
|
586
|
+
ch = hs->s[hs->pos];
|
587
|
+
if (ch == CHAR_GT) {
|
588
|
+
assert(hs->pos > 0);
|
589
|
+
hs->token_start = hs->s + hs->pos -1;
|
590
|
+
hs->token_len = 2;
|
591
|
+
hs->token_type = TAG_NAME_SELFCLOSE;
|
592
|
+
hs->state = h5_state_data;
|
593
|
+
hs->pos += 1;
|
594
|
+
return 1;
|
595
|
+
} else {
|
596
|
+
return h5_state_before_attribute_name(hs);
|
597
|
+
}
|
598
|
+
}
|
599
|
+
|
600
|
+
/**
|
601
|
+
* 12.2.4.44
|
602
|
+
*/
|
603
|
+
static int h5_state_bogus_comment(h5_state_t* hs)
|
604
|
+
{
|
605
|
+
const char* idx;
|
606
|
+
|
607
|
+
TRACE();
|
608
|
+
idx = (const char*) memchr(hs->s + hs->pos, CHAR_GT, hs->len - hs->pos);
|
609
|
+
if (idx == NULL) {
|
610
|
+
hs->token_start = hs->s + hs->pos;
|
611
|
+
hs->token_len = hs->len - hs->pos;
|
612
|
+
hs->pos = hs->len;
|
613
|
+
hs->state = h5_state_eof;
|
614
|
+
} else {
|
615
|
+
hs->token_start = hs->s + hs->pos;
|
616
|
+
hs->token_len = (size_t)(idx - hs->s) - hs->pos;
|
617
|
+
hs->pos = (size_t)(idx - hs->s) + 1;
|
618
|
+
hs->state = h5_state_data;
|
619
|
+
}
|
620
|
+
|
621
|
+
hs->token_type = TAG_COMMENT;
|
622
|
+
return 1;
|
623
|
+
}
|
624
|
+
|
625
|
+
/**
|
626
|
+
* 12.2.4.44 ALT
|
627
|
+
*/
|
628
|
+
static int h5_state_bogus_comment2(h5_state_t* hs)
|
629
|
+
{
|
630
|
+
const char* idx;
|
631
|
+
size_t pos;
|
632
|
+
|
633
|
+
TRACE();
|
634
|
+
pos = hs->pos;
|
635
|
+
while (1) {
|
636
|
+
idx = (const char*) memchr(hs->s + pos, CHAR_PERCENT, hs->len - pos);
|
637
|
+
if (idx == NULL || (idx + 1 >= hs->s + hs->len)) {
|
638
|
+
hs->token_start = hs->s + hs->pos;
|
639
|
+
hs->token_len = hs->len - hs->pos;
|
640
|
+
hs->pos = hs->len;
|
641
|
+
hs->token_type = TAG_COMMENT;
|
642
|
+
hs->state = h5_state_eof;
|
643
|
+
return 1;
|
644
|
+
}
|
645
|
+
|
646
|
+
if (*(idx +1) != CHAR_GT) {
|
647
|
+
pos = (size_t)(idx - hs->s) + 1;
|
648
|
+
continue;
|
649
|
+
}
|
650
|
+
|
651
|
+
/* ends in %> */
|
652
|
+
hs->token_start = hs->s + hs->pos;
|
653
|
+
hs->token_len = (size_t)(idx - hs->s) - hs->pos;
|
654
|
+
hs->pos = (size_t)(idx - hs->s) + 2;
|
655
|
+
hs->state = h5_state_data;
|
656
|
+
hs->token_type = TAG_COMMENT;
|
657
|
+
return 1;
|
658
|
+
}
|
659
|
+
}
|
660
|
+
|
661
|
+
/**
|
662
|
+
* 8.2.4.45
|
663
|
+
*/
|
664
|
+
static int h5_state_markup_declaration_open(h5_state_t* hs)
|
665
|
+
{
|
666
|
+
size_t remaining;
|
667
|
+
|
668
|
+
TRACE();
|
669
|
+
remaining = hs->len - hs->pos;
|
670
|
+
if (remaining >= 7 &&
|
671
|
+
/* case insensitive */
|
672
|
+
(hs->s[hs->pos + 0] == 'D' || hs->s[hs->pos + 0] == 'd') &&
|
673
|
+
(hs->s[hs->pos + 1] == 'O' || hs->s[hs->pos + 1] == 'o') &&
|
674
|
+
(hs->s[hs->pos + 2] == 'C' || hs->s[hs->pos + 2] == 'c') &&
|
675
|
+
(hs->s[hs->pos + 3] == 'T' || hs->s[hs->pos + 3] == 't') &&
|
676
|
+
(hs->s[hs->pos + 4] == 'Y' || hs->s[hs->pos + 4] == 'y') &&
|
677
|
+
(hs->s[hs->pos + 5] == 'P' || hs->s[hs->pos + 5] == 'p') &&
|
678
|
+
(hs->s[hs->pos + 6] == 'E' || hs->s[hs->pos + 6] == 'e')
|
679
|
+
) {
|
680
|
+
return h5_state_doctype(hs);
|
681
|
+
} else if (remaining >= 7 &&
|
682
|
+
/* upper case required */
|
683
|
+
hs->s[hs->pos + 0] == '[' &&
|
684
|
+
hs->s[hs->pos + 1] == 'C' &&
|
685
|
+
hs->s[hs->pos + 2] == 'D' &&
|
686
|
+
hs->s[hs->pos + 3] == 'A' &&
|
687
|
+
hs->s[hs->pos + 4] == 'T' &&
|
688
|
+
hs->s[hs->pos + 5] == 'A' &&
|
689
|
+
hs->s[hs->pos + 6] == '['
|
690
|
+
) {
|
691
|
+
hs->pos += 7;
|
692
|
+
return h5_state_cdata(hs);
|
693
|
+
} else if (remaining >= 2 &&
|
694
|
+
hs->s[hs->pos + 0] == '-' &&
|
695
|
+
hs->s[hs->pos + 1] == '-') {
|
696
|
+
hs->pos += 2;
|
697
|
+
return h5_state_comment(hs);
|
698
|
+
}
|
699
|
+
|
700
|
+
return h5_state_bogus_comment(hs);
|
701
|
+
}
|
702
|
+
|
703
|
+
/**
|
704
|
+
* 12.2.4.48
|
705
|
+
* 12.2.4.49
|
706
|
+
* 12.2.4.50
|
707
|
+
* 12.2.4.51
|
708
|
+
* state machine spec is confusing since it can only look
|
709
|
+
* at one character at a time but simply it's comments end by:
|
710
|
+
* 1) EOF
|
711
|
+
* 2) ending in -->
|
712
|
+
* 3) ending in -!>
|
713
|
+
*/
|
714
|
+
static int h5_state_comment(h5_state_t* hs)
|
715
|
+
{
|
716
|
+
char ch;
|
717
|
+
const char* idx;
|
718
|
+
size_t pos;
|
719
|
+
size_t offset;
|
720
|
+
const char* end = hs->s + hs->len;
|
721
|
+
|
722
|
+
TRACE();
|
723
|
+
pos = hs->pos;
|
724
|
+
while (1) {
|
725
|
+
|
726
|
+
idx = (const char*) memchr(hs->s + pos, CHAR_DASH, hs->len - pos);
|
727
|
+
|
728
|
+
/* did not find anything or has less than 3 chars left */
|
729
|
+
if (idx == NULL || idx > hs->s + hs->len - 3) {
|
730
|
+
hs->state = h5_state_eof;
|
731
|
+
hs->token_start = hs->s + hs->pos;
|
732
|
+
hs->token_len = hs->len - hs->pos;
|
733
|
+
hs->token_type = TAG_COMMENT;
|
734
|
+
return 1;
|
735
|
+
}
|
736
|
+
offset = 1;
|
737
|
+
|
738
|
+
/* skip all nulls */
|
739
|
+
while (idx + offset < end && *(idx + offset) == 0) {
|
740
|
+
offset += 1;
|
741
|
+
}
|
742
|
+
if (idx + offset == end) {
|
743
|
+
hs->state = h5_state_eof;
|
744
|
+
hs->token_start = hs->s + hs->pos;
|
745
|
+
hs->token_len = hs->len - hs->pos;
|
746
|
+
hs->token_type = TAG_COMMENT;
|
747
|
+
return 1;
|
748
|
+
}
|
749
|
+
|
750
|
+
ch = *(idx + offset);
|
751
|
+
if (ch != CHAR_DASH && ch != CHAR_BANG) {
|
752
|
+
pos = (size_t)(idx - hs->s) + 1;
|
753
|
+
continue;
|
754
|
+
}
|
755
|
+
|
756
|
+
/* need to test */
|
757
|
+
#if 0
|
758
|
+
/* skip all nulls */
|
759
|
+
while (idx + offset < end && *(idx + offset) == 0) {
|
760
|
+
offset += 1;
|
761
|
+
}
|
762
|
+
if (idx + offset == end) {
|
763
|
+
hs->state = h5_state_eof;
|
764
|
+
hs->token_start = hs->s + hs->pos;
|
765
|
+
hs->token_len = hs->len - hs->pos;
|
766
|
+
hs->token_type = TAG_COMMENT;
|
767
|
+
return 1;
|
768
|
+
}
|
769
|
+
#endif
|
770
|
+
|
771
|
+
offset += 1;
|
772
|
+
if (idx + offset == end) {
|
773
|
+
hs->state = h5_state_eof;
|
774
|
+
hs->token_start = hs->s + hs->pos;
|
775
|
+
hs->token_len = hs->len - hs->pos;
|
776
|
+
hs->token_type = TAG_COMMENT;
|
777
|
+
return 1;
|
778
|
+
}
|
779
|
+
|
780
|
+
|
781
|
+
ch = *(idx + offset);
|
782
|
+
if (ch != CHAR_GT) {
|
783
|
+
pos = (size_t)(idx - hs->s) + 1;
|
784
|
+
continue;
|
785
|
+
}
|
786
|
+
offset += 1;
|
787
|
+
|
788
|
+
/* ends in --> or -!> */
|
789
|
+
hs->token_start = hs->s + hs->pos;
|
790
|
+
hs->token_len = (size_t)(idx - hs->s) - hs->pos;
|
791
|
+
hs->pos = (size_t)(idx + offset - hs->s);
|
792
|
+
hs->state = h5_state_data;
|
793
|
+
hs->token_type = TAG_COMMENT;
|
794
|
+
return 1;
|
795
|
+
}
|
796
|
+
}
|
797
|
+
|
798
|
+
static int h5_state_cdata(h5_state_t* hs)
|
799
|
+
{
|
800
|
+
const char* idx;
|
801
|
+
size_t pos;
|
802
|
+
|
803
|
+
TRACE();
|
804
|
+
pos = hs->pos;
|
805
|
+
while (1) {
|
806
|
+
idx = (const char*) memchr(hs->s + pos, CHAR_RIGHTB, hs->len - pos);
|
807
|
+
|
808
|
+
/* did not find anything or has less than 3 chars left */
|
809
|
+
if (idx == NULL || idx > hs->s + hs->len - 3) {
|
810
|
+
hs->state = h5_state_eof;
|
811
|
+
hs->token_start = hs->s + hs->pos;
|
812
|
+
hs->token_len = hs->len - hs->pos;
|
813
|
+
hs->token_type = DATA_TEXT;
|
814
|
+
return 1;
|
815
|
+
} else if ( *(idx+1) == CHAR_RIGHTB && *(idx+2) == CHAR_GT) {
|
816
|
+
hs->state = h5_state_data;
|
817
|
+
hs->token_start = hs->s + hs->pos;
|
818
|
+
hs->token_len = (size_t)(idx - hs->s) - hs->pos;
|
819
|
+
hs->pos = (size_t)(idx - hs->s) + 3;
|
820
|
+
hs->token_type = DATA_TEXT;
|
821
|
+
return 1;
|
822
|
+
} else {
|
823
|
+
pos = (size_t)(idx - hs->s) + 1;
|
824
|
+
}
|
825
|
+
}
|
826
|
+
}
|
827
|
+
|
828
|
+
/**
|
829
|
+
* 8.2.4.52
|
830
|
+
* http://www.w3.org/html/wg/drafts/html/master/syntax.html#doctype-state
|
831
|
+
*/
|
832
|
+
static int h5_state_doctype(h5_state_t* hs)
|
833
|
+
{
|
834
|
+
const char* idx;
|
835
|
+
|
836
|
+
TRACE();
|
837
|
+
hs->token_start = hs->s + hs->pos;
|
838
|
+
hs->token_type = DOCTYPE;
|
839
|
+
|
840
|
+
idx = (const char*) memchr(hs->s + hs->pos, CHAR_GT, hs->len - hs->pos);
|
841
|
+
if (idx == NULL) {
|
842
|
+
hs->state = h5_state_eof;
|
843
|
+
hs->token_len = hs->len - hs->pos;
|
844
|
+
} else {
|
845
|
+
hs->state = h5_state_data;
|
846
|
+
hs->token_len = (size_t)(idx - hs->s) - hs->pos;
|
847
|
+
hs->pos = (size_t)(idx - hs->s) + 1;
|
848
|
+
}
|
849
|
+
return 1;
|
850
|
+
}
|