1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
|
/* Copyright (c) 2002-2018 Dovecot authors, see the included COPYING file */
#include "lib.h"
#include "buffer.h"
#include "hex-binary.h"
#include "qp-decoder.h"
/* quoted-printable lines can be max 76 characters. if we've seen more than
that much whitespace, it means there really shouldn't be anything else left
in the line except trailing whitespace. */
#define QP_MAX_WHITESPACE_LEN 76
#define QP_IS_TRAILING_WHITESPACE(c) \
((c) == ' ' || (c) == '\t')
enum qp_state {
STATE_TEXT = 0,
STATE_WHITESPACE,
STATE_EQUALS,
STATE_EQUALS_WHITESPACE,
STATE_HEX2,
STATE_CR,
STATE_SOFTCR
};
struct qp_decoder {
buffer_t *dest;
buffer_t *whitespace;
enum qp_state state;
char hexchar;
};
struct qp_decoder *qp_decoder_init(buffer_t *dest)
{
struct qp_decoder *qp;
qp = i_new(struct qp_decoder, 1);
qp->dest = dest;
qp->whitespace = buffer_create_dynamic(default_pool, 80);
return qp;
}
void qp_decoder_deinit(struct qp_decoder **_qp)
{
struct qp_decoder *qp = *_qp;
buffer_free(&qp->whitespace);
i_free(qp);
}
static size_t
qp_decoder_more_text(struct qp_decoder *qp, const unsigned char *src,
size_t src_size)
{
size_t i, start = 0, ret = src_size;
for (i = 0; i < src_size; i++) {
if (src[i] > '=') {
/* fast path */
continue;
}
switch (src[i]) {
case '=':
qp->state = STATE_EQUALS;
break;
case '\r':
qp->state = STATE_CR;
break;
case '\n':
/* LF without preceding CR */
buffer_append(qp->dest, src+start, i-start);
buffer_append(qp->dest, "\r\n", 2);
start = i+1;
continue;
case ' ':
case '\t':
i_assert(qp->whitespace->used == 0);
qp->state = STATE_WHITESPACE;
buffer_append_c(qp->whitespace, src[i]);
break;
default:
continue;
}
ret = i+1;
break;
}
buffer_append(qp->dest, src+start, i-start);
return ret;
}
static void qp_decoder_invalid(struct qp_decoder *qp, const char **error_r)
{
switch (qp->state) {
case STATE_EQUALS:
buffer_append_c(qp->dest, '=');
*error_r = "'=' not followed by two hex digits";
break;
case STATE_HEX2:
buffer_append_c(qp->dest, '=');
buffer_append_c(qp->dest, qp->hexchar);
*error_r = "'=<hex>' not followed by a hex digit";
break;
case STATE_EQUALS_WHITESPACE:
buffer_append_c(qp->dest, '=');
buffer_append_buf(qp->dest, qp->whitespace, 0, SIZE_MAX);
buffer_set_used_size(qp->whitespace, 0);
*error_r = "'=<whitespace>' not followed by newline";
break;
case STATE_CR:
buffer_append_buf(qp->dest, qp->whitespace, 0, SIZE_MAX);
buffer_set_used_size(qp->whitespace, 0);
buffer_append_c(qp->dest, '\r');
*error_r = "CR not followed by LF";
break;
case STATE_SOFTCR:
buffer_append_c(qp->dest, '=');
buffer_append_buf(qp->dest, qp->whitespace, 0, SIZE_MAX);
buffer_set_used_size(qp->whitespace, 0);
buffer_append_c(qp->dest, '\r');
*error_r = "CR not followed by LF";
break;
case STATE_TEXT:
case STATE_WHITESPACE:
i_unreached();
}
qp->state = STATE_TEXT;
i_assert(*error_r != NULL);
}
int qp_decoder_more(struct qp_decoder *qp, const unsigned char *src,
size_t src_size, size_t *invalid_src_pos_r,
const char **error_r)
{
const char *error;
size_t i;
*invalid_src_pos_r = SIZE_MAX;
*error_r = NULL;
for (i = 0; i < src_size; ) {
switch (qp->state) {
case STATE_TEXT:
i += qp_decoder_more_text(qp, src+i, src_size-i);
/* don't increment i any more than we already did,
so continue instead of break */
continue;
case STATE_WHITESPACE:
if (QP_IS_TRAILING_WHITESPACE(src[i])) {
/* more whitespace */
if (qp->whitespace->used <= QP_MAX_WHITESPACE_LEN)
buffer_append_c(qp->whitespace, src[i]);
} else if (src[i] == '\r') {
qp->state = STATE_CR;
} else if (src[i] == '\n') {
/* drop the trailing whitespace */
buffer_append(qp->dest, "\r\n", 2);
buffer_set_used_size(qp->whitespace, 0);
} else {
/* this wasn't trailing whitespace.
put it back. */
buffer_append_buf(qp->dest, qp->whitespace,
0, SIZE_MAX);
if (qp->whitespace->used > QP_MAX_WHITESPACE_LEN) {
/* we already truncated some of the
whitespace away, because the line
is too long */
if (*invalid_src_pos_r == SIZE_MAX) {
*invalid_src_pos_r = i;
*error_r = "Too much whitespace";
}
}
buffer_set_used_size(qp->whitespace, 0);
qp->state = STATE_TEXT;
continue; /* don't increment i */
}
break;
case STATE_EQUALS:
if ((src[i] >= '0' && src[i] <= '9') ||
(src[i] >= 'A' && src[i] <= 'F') ||
/* lowercase hex isn't strictly valid, but allow */
(src[i] >= 'a' && src[i] <= 'f')) {
qp->hexchar = src[i];
qp->state = STATE_HEX2;
} else if (QP_IS_TRAILING_WHITESPACE(src[i])) {
i_assert(qp->whitespace->used == 0);
buffer_append_c(qp->whitespace, src[i]);
qp->state = STATE_EQUALS_WHITESPACE;
} else if (src[i] == '\r')
qp->state = STATE_SOFTCR;
else if (src[i] == '\n') {
qp->state = STATE_TEXT;
} else {
/* invalid input */
qp_decoder_invalid(qp, &error);
if (*invalid_src_pos_r == SIZE_MAX) {
*invalid_src_pos_r = i;
*error_r = error;
}
continue; /* don't increment i */
}
break;
case STATE_HEX2:
if ((src[i] >= '0' && src[i] <= '9') ||
(src[i] >= 'A' && src[i] <= 'F') ||
(src[i] >= 'a' && src[i] <= 'f')) {
char data[3];
data[0] = qp->hexchar;
data[1] = src[i];
data[2] = '\0';
if (hex_to_binary(data, qp->dest) < 0)
i_unreached();
qp->state = STATE_TEXT;
} else {
/* invalid input */
qp_decoder_invalid(qp, &error);
if (*invalid_src_pos_r == SIZE_MAX) {
*invalid_src_pos_r = i;
*error_r = error;
}
continue; /* don't increment i */
}
break;
case STATE_EQUALS_WHITESPACE:
if (QP_IS_TRAILING_WHITESPACE(src[i])) {
if (qp->whitespace->used <= QP_MAX_WHITESPACE_LEN)
buffer_append_c(qp->whitespace, src[i]);
else {
/* if this isn't going to get truncated
anyway, it's going to be an error */
}
} else if (src[i] == '\r')
qp->state = STATE_SOFTCR;
else if (src[i] == '\n') {
buffer_set_used_size(qp->whitespace, 0);
qp->state = STATE_TEXT;
} else {
/* =<whitespace> not followed by [CR]LF
is invalid. */
qp_decoder_invalid(qp, &error);
if (*invalid_src_pos_r == SIZE_MAX) {
*invalid_src_pos_r = i;
*error_r = error;
}
continue; /* don't increment i */
}
break;
case STATE_CR:
case STATE_SOFTCR:
if (src[i] == '\n') {
buffer_set_used_size(qp->whitespace, 0);
if (qp->state != STATE_SOFTCR)
buffer_append(qp->dest, "\r\n", 2);
qp->state = STATE_TEXT;
} else {
qp_decoder_invalid(qp, &error);
if (*invalid_src_pos_r == SIZE_MAX) {
*invalid_src_pos_r = i;
*error_r = error;
}
continue; /* don't increment i */
}
break;
}
i++;
}
i_assert((*invalid_src_pos_r == SIZE_MAX) == (*error_r == NULL));
return *invalid_src_pos_r == SIZE_MAX ? 0 : -1;
}
int qp_decoder_finish(struct qp_decoder *qp, const char **error_r)
{
int ret;
if (qp->state == STATE_TEXT || qp->state == STATE_WHITESPACE) {
ret = 0;
*error_r = NULL;
} else {
qp_decoder_invalid(qp, error_r);
ret = -1;
}
qp->state = STATE_TEXT;
buffer_set_used_size(qp->whitespace, 0);
return ret;
}
|
