zscan 2.0.4 → 2.0.5
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +5 -5
- data/ext/extconf.rb +4 -1
- data/ext/pack/internal-26.h +2141 -0
- data/ext/pack/internal.h +319 -70
- data/ext/pack/pack-26.c +2010 -0
- data/ext/pack/pack.c +923 -15
- data/lib/zscan.rb +1 -1
- data/zscan.gemspec +1 -1
- metadata +5 -3
data/ext/pack/pack-26.c
ADDED
@@ -0,0 +1,2010 @@
|
|
1
|
+
/**********************************************************************
|
2
|
+
|
3
|
+
pack.c -
|
4
|
+
|
5
|
+
$Author: naruse $
|
6
|
+
created at: Thu Feb 10 15:17:05 JST 1994
|
7
|
+
|
8
|
+
Copyright (C) 1993-2007 Yukihiro Matsumoto
|
9
|
+
|
10
|
+
**********************************************************************/
|
11
|
+
|
12
|
+
#include "ruby/encoding.h"
|
13
|
+
#include "internal.h"
|
14
|
+
#include <sys/types.h>
|
15
|
+
#include <ctype.h>
|
16
|
+
#include <errno.h>
|
17
|
+
|
18
|
+
/*
|
19
|
+
* It is intentional that the condition for natstr is HAVE_TRUE_LONG_LONG
|
20
|
+
* instead of HAVE_LONG_LONG or LONG_LONG.
|
21
|
+
* This means q! and Q! means always the standard long long type and
|
22
|
+
* causes ArgumentError for platforms which has no long long type,
|
23
|
+
* even if the platform has an implementation specific 64bit type.
|
24
|
+
* This behavior is consistent with the document of pack/unpack.
|
25
|
+
*/
|
26
|
+
#ifdef HAVE_TRUE_LONG_LONG
|
27
|
+
static const char natstr[] = "sSiIlLqQjJ";
|
28
|
+
#else
|
29
|
+
static const char natstr[] = "sSiIlLjJ";
|
30
|
+
#endif
|
31
|
+
static const char endstr[] = "sSiIlLqQjJ";
|
32
|
+
|
33
|
+
#ifdef HAVE_TRUE_LONG_LONG
|
34
|
+
/* It is intentional to use long long instead of LONG_LONG. */
|
35
|
+
# define NATINT_LEN_Q NATINT_LEN(long long, 8)
|
36
|
+
#else
|
37
|
+
# define NATINT_LEN_Q 8
|
38
|
+
#endif
|
39
|
+
|
40
|
+
#if SIZEOF_SHORT != 2 || SIZEOF_LONG != 4 || (defined(HAVE_TRUE_LONG_LONG) && SIZEOF_LONG_LONG != 8)
|
41
|
+
# define NATINT_PACK
|
42
|
+
#endif
|
43
|
+
|
44
|
+
#ifdef DYNAMIC_ENDIAN
|
45
|
+
/* for universal binary of NEXTSTEP and MacOS X */
|
46
|
+
/* useless since autoconf 2.63? */
|
47
|
+
static int
|
48
|
+
is_bigendian(void)
|
49
|
+
{
|
50
|
+
static int init = 0;
|
51
|
+
static int endian_value;
|
52
|
+
char *p;
|
53
|
+
|
54
|
+
if (init) return endian_value;
|
55
|
+
init = 1;
|
56
|
+
p = (char*)&init;
|
57
|
+
return endian_value = p[0]?0:1;
|
58
|
+
}
|
59
|
+
# define BIGENDIAN_P() (is_bigendian())
|
60
|
+
#elif defined(WORDS_BIGENDIAN)
|
61
|
+
# define BIGENDIAN_P() 1
|
62
|
+
#else
|
63
|
+
# define BIGENDIAN_P() 0
|
64
|
+
#endif
|
65
|
+
|
66
|
+
#ifdef NATINT_PACK
|
67
|
+
# define NATINT_LEN(type,len) (natint?(int)sizeof(type):(int)(len))
|
68
|
+
#else
|
69
|
+
# define NATINT_LEN(type,len) ((int)sizeof(type))
|
70
|
+
#endif
|
71
|
+
|
72
|
+
typedef union {
|
73
|
+
float f;
|
74
|
+
uint32_t u;
|
75
|
+
char buf[4];
|
76
|
+
} FLOAT_SWAPPER;
|
77
|
+
typedef union {
|
78
|
+
double d;
|
79
|
+
uint64_t u;
|
80
|
+
char buf[8];
|
81
|
+
} DOUBLE_SWAPPER;
|
82
|
+
#define swapf(x) swap32(x)
|
83
|
+
#define swapd(x) swap64(x)
|
84
|
+
|
85
|
+
#define rb_ntohf(x) (BIGENDIAN_P()?(x):swapf(x))
|
86
|
+
#define rb_ntohd(x) (BIGENDIAN_P()?(x):swapd(x))
|
87
|
+
#define rb_htonf(x) (BIGENDIAN_P()?(x):swapf(x))
|
88
|
+
#define rb_htond(x) (BIGENDIAN_P()?(x):swapd(x))
|
89
|
+
#define rb_htovf(x) (BIGENDIAN_P()?swapf(x):(x))
|
90
|
+
#define rb_htovd(x) (BIGENDIAN_P()?swapd(x):(x))
|
91
|
+
#define rb_vtohf(x) (BIGENDIAN_P()?swapf(x):(x))
|
92
|
+
#define rb_vtohd(x) (BIGENDIAN_P()?swapd(x):(x))
|
93
|
+
|
94
|
+
#define FLOAT_CONVWITH(x) FLOAT_SWAPPER x;
|
95
|
+
#define HTONF(x) ((x).u = rb_htonf((x).u))
|
96
|
+
#define HTOVF(x) ((x).u = rb_htovf((x).u))
|
97
|
+
#define NTOHF(x) ((x).u = rb_ntohf((x).u))
|
98
|
+
#define VTOHF(x) ((x).u = rb_vtohf((x).u))
|
99
|
+
|
100
|
+
#define DOUBLE_CONVWITH(x) DOUBLE_SWAPPER x;
|
101
|
+
#define HTOND(x) ((x).u = rb_htond((x).u))
|
102
|
+
#define HTOVD(x) ((x).u = rb_htovd((x).u))
|
103
|
+
#define NTOHD(x) ((x).u = rb_ntohd((x).u))
|
104
|
+
#define VTOHD(x) ((x).u = rb_vtohd((x).u))
|
105
|
+
|
106
|
+
#define MAX_INTEGER_PACK_SIZE 8
|
107
|
+
|
108
|
+
static const char toofew[] = "too few arguments";
|
109
|
+
|
110
|
+
static void encodes(VALUE,const char*,long,int,int);
|
111
|
+
static void qpencode(VALUE,VALUE,long);
|
112
|
+
|
113
|
+
static unsigned long utf8_to_uv(const char*,long*);
|
114
|
+
|
115
|
+
static ID id_associated;
|
116
|
+
|
117
|
+
static void
|
118
|
+
str_associate(VALUE str, VALUE add)
|
119
|
+
{
|
120
|
+
/* assert(NIL_P(rb_attr_get(str, id_associated))); */
|
121
|
+
rb_ivar_set(str, id_associated, add);
|
122
|
+
}
|
123
|
+
|
124
|
+
static VALUE
|
125
|
+
str_associated(VALUE str)
|
126
|
+
{
|
127
|
+
return rb_ivar_lookup(str, id_associated, Qfalse);
|
128
|
+
}
|
129
|
+
|
130
|
+
/*
|
131
|
+
* call-seq:
|
132
|
+
* arr.pack( aTemplateString ) -> aBinaryString
|
133
|
+
* arr.pack( aTemplateString, buffer: aBufferString ) -> aBufferString
|
134
|
+
*
|
135
|
+
* Packs the contents of <i>arr</i> into a binary sequence according to
|
136
|
+
* the directives in <i>aTemplateString</i> (see the table below)
|
137
|
+
* Directives ``A,'' ``a,'' and ``Z'' may be followed by a count,
|
138
|
+
* which gives the width of the resulting field. The remaining
|
139
|
+
* directives also may take a count, indicating the number of array
|
140
|
+
* elements to convert. If the count is an asterisk
|
141
|
+
* (``<code>*</code>''), all remaining array elements will be
|
142
|
+
* converted. Any of the directives ``<code>sSiIlL</code>'' may be
|
143
|
+
* followed by an underscore (``<code>_</code>'') or
|
144
|
+
* exclamation mark (``<code>!</code>'') to use the underlying
|
145
|
+
* platform's native size for the specified type; otherwise, they use a
|
146
|
+
* platform-independent size. Spaces are ignored in the template
|
147
|
+
* string. See also <code>String#unpack</code>.
|
148
|
+
*
|
149
|
+
* a = [ "a", "b", "c" ]
|
150
|
+
* n = [ 65, 66, 67 ]
|
151
|
+
* a.pack("A3A3A3") #=> "a b c "
|
152
|
+
* a.pack("a3a3a3") #=> "a\000\000b\000\000c\000\000"
|
153
|
+
* n.pack("ccc") #=> "ABC"
|
154
|
+
*
|
155
|
+
* If <i>aBufferString</i> is specified and its capacity is enough,
|
156
|
+
* +pack+ uses it as the buffer and returns it.
|
157
|
+
* When the offset is specified by the beginning of <i>aTemplateString</i>,
|
158
|
+
* the result is filled after the offset.
|
159
|
+
* If original contents of <i>aBufferString</i> exists and it's longer than
|
160
|
+
* the offset, the rest of <i>offsetOfBuffer</i> are overwritten by the result.
|
161
|
+
* If it's shorter, the gap is filled with ``<code>\0</code>''.
|
162
|
+
*
|
163
|
+
* Note that ``buffer:'' option does not guarantee not to allocate memory
|
164
|
+
* in +pack+. If the capacity of <i>aBufferString</i> is not enough,
|
165
|
+
* +pack+ allocates memory.
|
166
|
+
*
|
167
|
+
* Directives for +pack+.
|
168
|
+
*
|
169
|
+
* Integer | Array |
|
170
|
+
* Directive | Element | Meaning
|
171
|
+
* ----------------------------------------------------------------------------
|
172
|
+
* C | Integer | 8-bit unsigned (unsigned char)
|
173
|
+
* S | Integer | 16-bit unsigned, native endian (uint16_t)
|
174
|
+
* L | Integer | 32-bit unsigned, native endian (uint32_t)
|
175
|
+
* Q | Integer | 64-bit unsigned, native endian (uint64_t)
|
176
|
+
* J | Integer | pointer width unsigned, native endian (uintptr_t)
|
177
|
+
* | | (J is available since Ruby 2.3.)
|
178
|
+
* | |
|
179
|
+
* c | Integer | 8-bit signed (signed char)
|
180
|
+
* s | Integer | 16-bit signed, native endian (int16_t)
|
181
|
+
* l | Integer | 32-bit signed, native endian (int32_t)
|
182
|
+
* q | Integer | 64-bit signed, native endian (int64_t)
|
183
|
+
* j | Integer | pointer width signed, native endian (intptr_t)
|
184
|
+
* | | (j is available since Ruby 2.3.)
|
185
|
+
* | |
|
186
|
+
* S_ S! | Integer | unsigned short, native endian
|
187
|
+
* I I_ I! | Integer | unsigned int, native endian
|
188
|
+
* L_ L! | Integer | unsigned long, native endian
|
189
|
+
* Q_ Q! | Integer | unsigned long long, native endian (ArgumentError
|
190
|
+
* | | if the platform has no long long type.)
|
191
|
+
* | | (Q_ and Q! is available since Ruby 2.1.)
|
192
|
+
* J! | Integer | uintptr_t, native endian (same with J)
|
193
|
+
* | | (J! is available since Ruby 2.3.)
|
194
|
+
* | |
|
195
|
+
* s_ s! | Integer | signed short, native endian
|
196
|
+
* i i_ i! | Integer | signed int, native endian
|
197
|
+
* l_ l! | Integer | signed long, native endian
|
198
|
+
* q_ q! | Integer | signed long long, native endian (ArgumentError
|
199
|
+
* | | if the platform has no long long type.)
|
200
|
+
* | | (q_ and q! is available since Ruby 2.1.)
|
201
|
+
* j! | Integer | intptr_t, native endian (same with j)
|
202
|
+
* | | (j! is available since Ruby 2.3.)
|
203
|
+
* | |
|
204
|
+
* S> s> S!> s!> | Integer | same as the directives without ">" except
|
205
|
+
* L> l> L!> l!> | | big endian
|
206
|
+
* I!> i!> | | (available since Ruby 1.9.3)
|
207
|
+
* Q> q> Q!> q!> | | "S>" is same as "n"
|
208
|
+
* J> j> J!> j!> | | "L>" is same as "N"
|
209
|
+
* | |
|
210
|
+
* S< s< S!< s!< | Integer | same as the directives without "<" except
|
211
|
+
* L< l< L!< l!< | | little endian
|
212
|
+
* I!< i!< | | (available since Ruby 1.9.3)
|
213
|
+
* Q< q< Q!< q!< | | "S<" is same as "v"
|
214
|
+
* J< j< J!< j!< | | "L<" is same as "V"
|
215
|
+
* | |
|
216
|
+
* n | Integer | 16-bit unsigned, network (big-endian) byte order
|
217
|
+
* N | Integer | 32-bit unsigned, network (big-endian) byte order
|
218
|
+
* v | Integer | 16-bit unsigned, VAX (little-endian) byte order
|
219
|
+
* V | Integer | 32-bit unsigned, VAX (little-endian) byte order
|
220
|
+
* | |
|
221
|
+
* U | Integer | UTF-8 character
|
222
|
+
* w | Integer | BER-compressed integer
|
223
|
+
*
|
224
|
+
* Float | Array |
|
225
|
+
* Directive | Element | Meaning
|
226
|
+
* ---------------------------------------------------------------------------
|
227
|
+
* D d | Float | double-precision, native format
|
228
|
+
* F f | Float | single-precision, native format
|
229
|
+
* E | Float | double-precision, little-endian byte order
|
230
|
+
* e | Float | single-precision, little-endian byte order
|
231
|
+
* G | Float | double-precision, network (big-endian) byte order
|
232
|
+
* g | Float | single-precision, network (big-endian) byte order
|
233
|
+
*
|
234
|
+
* String | Array |
|
235
|
+
* Directive | Element | Meaning
|
236
|
+
* ---------------------------------------------------------------------------
|
237
|
+
* A | String | arbitrary binary string (space padded, count is width)
|
238
|
+
* a | String | arbitrary binary string (null padded, count is width)
|
239
|
+
* Z | String | same as ``a'', except that null is added with *
|
240
|
+
* B | String | bit string (MSB first)
|
241
|
+
* b | String | bit string (LSB first)
|
242
|
+
* H | String | hex string (high nibble first)
|
243
|
+
* h | String | hex string (low nibble first)
|
244
|
+
* u | String | UU-encoded string
|
245
|
+
* M | String | quoted printable, MIME encoding (see also RFC2045)
|
246
|
+
* | | (text mode but input must use LF and output LF)
|
247
|
+
* m | String | base64 encoded string (see RFC 2045, count is width)
|
248
|
+
* | | (if count is 0, no line feed are added, see RFC 4648)
|
249
|
+
* P | String | pointer to a structure (fixed-length string)
|
250
|
+
* p | String | pointer to a null-terminated string
|
251
|
+
*
|
252
|
+
* Misc. | Array |
|
253
|
+
* Directive | Element | Meaning
|
254
|
+
* ---------------------------------------------------------------------------
|
255
|
+
* @ | --- | moves to absolute position
|
256
|
+
* X | --- | back up a byte
|
257
|
+
* x | --- | null byte
|
258
|
+
*/
|
259
|
+
|
260
|
+
static VALUE
|
261
|
+
pack_pack(int argc, VALUE *argv, VALUE ary)
|
262
|
+
{
|
263
|
+
static const char nul10[] = "\0\0\0\0\0\0\0\0\0\0";
|
264
|
+
static const char spc10[] = " ";
|
265
|
+
const char *p, *pend;
|
266
|
+
VALUE fmt, opt = Qnil, res, from, associates = 0, buffer = 0;
|
267
|
+
char type;
|
268
|
+
long len, idx, plen;
|
269
|
+
const char *ptr;
|
270
|
+
int enc_info = 1; /* 0 - BINARY, 1 - US-ASCII, 2 - UTF-8 */
|
271
|
+
#ifdef NATINT_PACK
|
272
|
+
int natint; /* native integer */
|
273
|
+
#endif
|
274
|
+
int integer_size, bigendian_p;
|
275
|
+
|
276
|
+
rb_scan_args(argc, argv, "10:", &fmt, &opt);
|
277
|
+
|
278
|
+
StringValue(fmt);
|
279
|
+
p = RSTRING_PTR(fmt);
|
280
|
+
pend = p + RSTRING_LEN(fmt);
|
281
|
+
if (!NIL_P(opt)) {
|
282
|
+
static ID keyword_ids[1];
|
283
|
+
if (!keyword_ids[0])
|
284
|
+
CONST_ID(keyword_ids[0], "buffer");
|
285
|
+
|
286
|
+
rb_get_kwargs(opt, keyword_ids, 0, 1, &buffer);
|
287
|
+
|
288
|
+
if (buffer != Qundef && !RB_TYPE_P(buffer, T_STRING))
|
289
|
+
rb_raise(rb_eTypeError, "buffer must be String, not %s", rb_obj_classname(buffer));
|
290
|
+
}
|
291
|
+
if (buffer)
|
292
|
+
res = buffer;
|
293
|
+
else
|
294
|
+
res = rb_str_buf_new(0);
|
295
|
+
|
296
|
+
idx = 0;
|
297
|
+
|
298
|
+
#define TOO_FEW (rb_raise(rb_eArgError, toofew), 0)
|
299
|
+
#define MORE_ITEM (idx < RARRAY_LEN(ary))
|
300
|
+
#define THISFROM (MORE_ITEM ? RARRAY_AREF(ary, idx) : TOO_FEW)
|
301
|
+
#define NEXTFROM (MORE_ITEM ? RARRAY_AREF(ary, idx++) : TOO_FEW)
|
302
|
+
|
303
|
+
while (p < pend) {
|
304
|
+
int explicit_endian = 0;
|
305
|
+
if (RSTRING_PTR(fmt) + RSTRING_LEN(fmt) != pend) {
|
306
|
+
rb_raise(rb_eRuntimeError, "format string modified");
|
307
|
+
}
|
308
|
+
type = *p++; /* get data type */
|
309
|
+
#ifdef NATINT_PACK
|
310
|
+
natint = 0;
|
311
|
+
#endif
|
312
|
+
|
313
|
+
if (ISSPACE(type)) continue;
|
314
|
+
if (type == '#') {
|
315
|
+
while ((p < pend) && (*p != '\n')) {
|
316
|
+
p++;
|
317
|
+
}
|
318
|
+
continue;
|
319
|
+
}
|
320
|
+
|
321
|
+
{
|
322
|
+
modifiers:
|
323
|
+
switch (*p) {
|
324
|
+
case '_':
|
325
|
+
case '!':
|
326
|
+
if (strchr(natstr, type)) {
|
327
|
+
#ifdef NATINT_PACK
|
328
|
+
natint = 1;
|
329
|
+
#endif
|
330
|
+
p++;
|
331
|
+
}
|
332
|
+
else {
|
333
|
+
rb_raise(rb_eArgError, "'%c' allowed only after types %s", *p, natstr);
|
334
|
+
}
|
335
|
+
goto modifiers;
|
336
|
+
|
337
|
+
case '<':
|
338
|
+
case '>':
|
339
|
+
if (!strchr(endstr, type)) {
|
340
|
+
rb_raise(rb_eArgError, "'%c' allowed only after types %s", *p, endstr);
|
341
|
+
}
|
342
|
+
if (explicit_endian) {
|
343
|
+
rb_raise(rb_eRangeError, "Can't use both '<' and '>'");
|
344
|
+
}
|
345
|
+
explicit_endian = *p++;
|
346
|
+
goto modifiers;
|
347
|
+
}
|
348
|
+
}
|
349
|
+
|
350
|
+
if (*p == '*') { /* set data length */
|
351
|
+
len = strchr("@Xxu", type) ? 0
|
352
|
+
: strchr("PMm", type) ? 1
|
353
|
+
: RARRAY_LEN(ary) - idx;
|
354
|
+
p++;
|
355
|
+
}
|
356
|
+
else if (ISDIGIT(*p)) {
|
357
|
+
errno = 0;
|
358
|
+
len = STRTOUL(p, (char**)&p, 10);
|
359
|
+
if (errno) {
|
360
|
+
rb_raise(rb_eRangeError, "pack length too big");
|
361
|
+
}
|
362
|
+
}
|
363
|
+
else {
|
364
|
+
len = 1;
|
365
|
+
}
|
366
|
+
|
367
|
+
switch (type) {
|
368
|
+
case 'U':
|
369
|
+
/* if encoding is US-ASCII, upgrade to UTF-8 */
|
370
|
+
if (enc_info == 1) enc_info = 2;
|
371
|
+
break;
|
372
|
+
case 'm': case 'M': case 'u':
|
373
|
+
/* keep US-ASCII (do nothing) */
|
374
|
+
break;
|
375
|
+
default:
|
376
|
+
/* fall back to BINARY */
|
377
|
+
enc_info = 0;
|
378
|
+
break;
|
379
|
+
}
|
380
|
+
switch (type) {
|
381
|
+
case 'A': case 'a': case 'Z':
|
382
|
+
case 'B': case 'b':
|
383
|
+
case 'H': case 'h':
|
384
|
+
from = NEXTFROM;
|
385
|
+
if (NIL_P(from)) {
|
386
|
+
ptr = "";
|
387
|
+
plen = 0;
|
388
|
+
}
|
389
|
+
else {
|
390
|
+
StringValue(from);
|
391
|
+
ptr = RSTRING_PTR(from);
|
392
|
+
plen = RSTRING_LEN(from);
|
393
|
+
OBJ_INFECT(res, from);
|
394
|
+
}
|
395
|
+
|
396
|
+
if (p[-1] == '*')
|
397
|
+
len = plen;
|
398
|
+
|
399
|
+
switch (type) {
|
400
|
+
case 'a': /* arbitrary binary string (null padded) */
|
401
|
+
case 'A': /* arbitrary binary string (ASCII space padded) */
|
402
|
+
case 'Z': /* null terminated string */
|
403
|
+
if (plen >= len) {
|
404
|
+
rb_str_buf_cat(res, ptr, len);
|
405
|
+
if (p[-1] == '*' && type == 'Z')
|
406
|
+
rb_str_buf_cat(res, nul10, 1);
|
407
|
+
}
|
408
|
+
else {
|
409
|
+
rb_str_buf_cat(res, ptr, plen);
|
410
|
+
len -= plen;
|
411
|
+
while (len >= 10) {
|
412
|
+
rb_str_buf_cat(res, (type == 'A')?spc10:nul10, 10);
|
413
|
+
len -= 10;
|
414
|
+
}
|
415
|
+
rb_str_buf_cat(res, (type == 'A')?spc10:nul10, len);
|
416
|
+
}
|
417
|
+
break;
|
418
|
+
|
419
|
+
#define castchar(from) (char)((from) & 0xff)
|
420
|
+
|
421
|
+
case 'b': /* bit string (ascending) */
|
422
|
+
{
|
423
|
+
int byte = 0;
|
424
|
+
long i, j = 0;
|
425
|
+
|
426
|
+
if (len > plen) {
|
427
|
+
j = (len - plen + 1)/2;
|
428
|
+
len = plen;
|
429
|
+
}
|
430
|
+
for (i=0; i++ < len; ptr++) {
|
431
|
+
if (*ptr & 1)
|
432
|
+
byte |= 128;
|
433
|
+
if (i & 7)
|
434
|
+
byte >>= 1;
|
435
|
+
else {
|
436
|
+
char c = castchar(byte);
|
437
|
+
rb_str_buf_cat(res, &c, 1);
|
438
|
+
byte = 0;
|
439
|
+
}
|
440
|
+
}
|
441
|
+
if (len & 7) {
|
442
|
+
char c;
|
443
|
+
byte >>= 7 - (len & 7);
|
444
|
+
c = castchar(byte);
|
445
|
+
rb_str_buf_cat(res, &c, 1);
|
446
|
+
}
|
447
|
+
len = j;
|
448
|
+
goto grow;
|
449
|
+
}
|
450
|
+
break;
|
451
|
+
|
452
|
+
case 'B': /* bit string (descending) */
|
453
|
+
{
|
454
|
+
int byte = 0;
|
455
|
+
long i, j = 0;
|
456
|
+
|
457
|
+
if (len > plen) {
|
458
|
+
j = (len - plen + 1)/2;
|
459
|
+
len = plen;
|
460
|
+
}
|
461
|
+
for (i=0; i++ < len; ptr++) {
|
462
|
+
byte |= *ptr & 1;
|
463
|
+
if (i & 7)
|
464
|
+
byte <<= 1;
|
465
|
+
else {
|
466
|
+
char c = castchar(byte);
|
467
|
+
rb_str_buf_cat(res, &c, 1);
|
468
|
+
byte = 0;
|
469
|
+
}
|
470
|
+
}
|
471
|
+
if (len & 7) {
|
472
|
+
char c;
|
473
|
+
byte <<= 7 - (len & 7);
|
474
|
+
c = castchar(byte);
|
475
|
+
rb_str_buf_cat(res, &c, 1);
|
476
|
+
}
|
477
|
+
len = j;
|
478
|
+
goto grow;
|
479
|
+
}
|
480
|
+
break;
|
481
|
+
|
482
|
+
case 'h': /* hex string (low nibble first) */
|
483
|
+
{
|
484
|
+
int byte = 0;
|
485
|
+
long i, j = 0;
|
486
|
+
|
487
|
+
if (len > plen) {
|
488
|
+
j = (len + 1) / 2 - (plen + 1) / 2;
|
489
|
+
len = plen;
|
490
|
+
}
|
491
|
+
for (i=0; i++ < len; ptr++) {
|
492
|
+
if (ISALPHA(*ptr))
|
493
|
+
byte |= (((*ptr & 15) + 9) & 15) << 4;
|
494
|
+
else
|
495
|
+
byte |= (*ptr & 15) << 4;
|
496
|
+
if (i & 1)
|
497
|
+
byte >>= 4;
|
498
|
+
else {
|
499
|
+
char c = castchar(byte);
|
500
|
+
rb_str_buf_cat(res, &c, 1);
|
501
|
+
byte = 0;
|
502
|
+
}
|
503
|
+
}
|
504
|
+
if (len & 1) {
|
505
|
+
char c = castchar(byte);
|
506
|
+
rb_str_buf_cat(res, &c, 1);
|
507
|
+
}
|
508
|
+
len = j;
|
509
|
+
goto grow;
|
510
|
+
}
|
511
|
+
break;
|
512
|
+
|
513
|
+
case 'H': /* hex string (high nibble first) */
|
514
|
+
{
|
515
|
+
int byte = 0;
|
516
|
+
long i, j = 0;
|
517
|
+
|
518
|
+
if (len > plen) {
|
519
|
+
j = (len + 1) / 2 - (plen + 1) / 2;
|
520
|
+
len = plen;
|
521
|
+
}
|
522
|
+
for (i=0; i++ < len; ptr++) {
|
523
|
+
if (ISALPHA(*ptr))
|
524
|
+
byte |= ((*ptr & 15) + 9) & 15;
|
525
|
+
else
|
526
|
+
byte |= *ptr & 15;
|
527
|
+
if (i & 1)
|
528
|
+
byte <<= 4;
|
529
|
+
else {
|
530
|
+
char c = castchar(byte);
|
531
|
+
rb_str_buf_cat(res, &c, 1);
|
532
|
+
byte = 0;
|
533
|
+
}
|
534
|
+
}
|
535
|
+
if (len & 1) {
|
536
|
+
char c = castchar(byte);
|
537
|
+
rb_str_buf_cat(res, &c, 1);
|
538
|
+
}
|
539
|
+
len = j;
|
540
|
+
goto grow;
|
541
|
+
}
|
542
|
+
break;
|
543
|
+
}
|
544
|
+
break;
|
545
|
+
|
546
|
+
case 'c': /* signed char */
|
547
|
+
case 'C': /* unsigned char */
|
548
|
+
integer_size = 1;
|
549
|
+
bigendian_p = BIGENDIAN_P(); /* not effective */
|
550
|
+
goto pack_integer;
|
551
|
+
|
552
|
+
case 's': /* s for int16_t, s! for signed short */
|
553
|
+
integer_size = NATINT_LEN(short, 2);
|
554
|
+
bigendian_p = BIGENDIAN_P();
|
555
|
+
goto pack_integer;
|
556
|
+
|
557
|
+
case 'S': /* S for uint16_t, S! for unsigned short */
|
558
|
+
integer_size = NATINT_LEN(short, 2);
|
559
|
+
bigendian_p = BIGENDIAN_P();
|
560
|
+
goto pack_integer;
|
561
|
+
|
562
|
+
case 'i': /* i and i! for signed int */
|
563
|
+
integer_size = (int)sizeof(int);
|
564
|
+
bigendian_p = BIGENDIAN_P();
|
565
|
+
goto pack_integer;
|
566
|
+
|
567
|
+
case 'I': /* I and I! for unsigned int */
|
568
|
+
integer_size = (int)sizeof(int);
|
569
|
+
bigendian_p = BIGENDIAN_P();
|
570
|
+
goto pack_integer;
|
571
|
+
|
572
|
+
case 'l': /* l for int32_t, l! for signed long */
|
573
|
+
integer_size = NATINT_LEN(long, 4);
|
574
|
+
bigendian_p = BIGENDIAN_P();
|
575
|
+
goto pack_integer;
|
576
|
+
|
577
|
+
case 'L': /* L for uint32_t, L! for unsigned long */
|
578
|
+
integer_size = NATINT_LEN(long, 4);
|
579
|
+
bigendian_p = BIGENDIAN_P();
|
580
|
+
goto pack_integer;
|
581
|
+
|
582
|
+
case 'q': /* q for int64_t, q! for signed long long */
|
583
|
+
integer_size = NATINT_LEN_Q;
|
584
|
+
bigendian_p = BIGENDIAN_P();
|
585
|
+
goto pack_integer;
|
586
|
+
|
587
|
+
case 'Q': /* Q for uint64_t, Q! for unsigned long long */
|
588
|
+
integer_size = NATINT_LEN_Q;
|
589
|
+
bigendian_p = BIGENDIAN_P();
|
590
|
+
goto pack_integer;
|
591
|
+
|
592
|
+
case 'j': /* j for intptr_t */
|
593
|
+
integer_size = sizeof(intptr_t);
|
594
|
+
bigendian_p = BIGENDIAN_P();
|
595
|
+
goto pack_integer;
|
596
|
+
|
597
|
+
case 'J': /* J for uintptr_t */
|
598
|
+
integer_size = sizeof(uintptr_t);
|
599
|
+
bigendian_p = BIGENDIAN_P();
|
600
|
+
goto pack_integer;
|
601
|
+
|
602
|
+
case 'n': /* 16 bit (2 bytes) integer (network byte-order) */
|
603
|
+
integer_size = 2;
|
604
|
+
bigendian_p = 1;
|
605
|
+
goto pack_integer;
|
606
|
+
|
607
|
+
case 'N': /* 32 bit (4 bytes) integer (network byte-order) */
|
608
|
+
integer_size = 4;
|
609
|
+
bigendian_p = 1;
|
610
|
+
goto pack_integer;
|
611
|
+
|
612
|
+
case 'v': /* 16 bit (2 bytes) integer (VAX byte-order) */
|
613
|
+
integer_size = 2;
|
614
|
+
bigendian_p = 0;
|
615
|
+
goto pack_integer;
|
616
|
+
|
617
|
+
case 'V': /* 32 bit (4 bytes) integer (VAX byte-order) */
|
618
|
+
integer_size = 4;
|
619
|
+
bigendian_p = 0;
|
620
|
+
goto pack_integer;
|
621
|
+
|
622
|
+
pack_integer:
|
623
|
+
if (explicit_endian) {
|
624
|
+
bigendian_p = explicit_endian == '>';
|
625
|
+
}
|
626
|
+
if (integer_size > MAX_INTEGER_PACK_SIZE)
|
627
|
+
rb_bug("unexpected intger size for pack: %d", integer_size);
|
628
|
+
while (len-- > 0) {
|
629
|
+
char intbuf[MAX_INTEGER_PACK_SIZE];
|
630
|
+
|
631
|
+
from = NEXTFROM;
|
632
|
+
rb_integer_pack(from, intbuf, integer_size, 1, 0,
|
633
|
+
INTEGER_PACK_2COMP |
|
634
|
+
(bigendian_p ? INTEGER_PACK_BIG_ENDIAN : INTEGER_PACK_LITTLE_ENDIAN));
|
635
|
+
rb_str_buf_cat(res, intbuf, integer_size);
|
636
|
+
}
|
637
|
+
break;
|
638
|
+
|
639
|
+
case 'f': /* single precision float in native format */
|
640
|
+
case 'F': /* ditto */
|
641
|
+
while (len-- > 0) {
|
642
|
+
float f;
|
643
|
+
|
644
|
+
from = NEXTFROM;
|
645
|
+
f = (float)RFLOAT_VALUE(rb_to_float(from));
|
646
|
+
rb_str_buf_cat(res, (char*)&f, sizeof(float));
|
647
|
+
}
|
648
|
+
break;
|
649
|
+
|
650
|
+
case 'e': /* single precision float in VAX byte-order */
|
651
|
+
while (len-- > 0) {
|
652
|
+
FLOAT_CONVWITH(tmp);
|
653
|
+
|
654
|
+
from = NEXTFROM;
|
655
|
+
tmp.f = (float)RFLOAT_VALUE(rb_to_float(from));
|
656
|
+
HTOVF(tmp);
|
657
|
+
rb_str_buf_cat(res, tmp.buf, sizeof(float));
|
658
|
+
}
|
659
|
+
break;
|
660
|
+
|
661
|
+
case 'E': /* double precision float in VAX byte-order */
|
662
|
+
while (len-- > 0) {
|
663
|
+
DOUBLE_CONVWITH(tmp);
|
664
|
+
from = NEXTFROM;
|
665
|
+
tmp.d = RFLOAT_VALUE(rb_to_float(from));
|
666
|
+
HTOVD(tmp);
|
667
|
+
rb_str_buf_cat(res, tmp.buf, sizeof(double));
|
668
|
+
}
|
669
|
+
break;
|
670
|
+
|
671
|
+
case 'd': /* double precision float in native format */
|
672
|
+
case 'D': /* ditto */
|
673
|
+
while (len-- > 0) {
|
674
|
+
double d;
|
675
|
+
|
676
|
+
from = NEXTFROM;
|
677
|
+
d = RFLOAT_VALUE(rb_to_float(from));
|
678
|
+
rb_str_buf_cat(res, (char*)&d, sizeof(double));
|
679
|
+
}
|
680
|
+
break;
|
681
|
+
|
682
|
+
case 'g': /* single precision float in network byte-order */
|
683
|
+
while (len-- > 0) {
|
684
|
+
FLOAT_CONVWITH(tmp);
|
685
|
+
from = NEXTFROM;
|
686
|
+
tmp.f = (float)RFLOAT_VALUE(rb_to_float(from));
|
687
|
+
HTONF(tmp);
|
688
|
+
rb_str_buf_cat(res, tmp.buf, sizeof(float));
|
689
|
+
}
|
690
|
+
break;
|
691
|
+
|
692
|
+
case 'G': /* double precision float in network byte-order */
|
693
|
+
while (len-- > 0) {
|
694
|
+
DOUBLE_CONVWITH(tmp);
|
695
|
+
|
696
|
+
from = NEXTFROM;
|
697
|
+
tmp.d = RFLOAT_VALUE(rb_to_float(from));
|
698
|
+
HTOND(tmp);
|
699
|
+
rb_str_buf_cat(res, tmp.buf, sizeof(double));
|
700
|
+
}
|
701
|
+
break;
|
702
|
+
|
703
|
+
case 'x': /* null byte */
|
704
|
+
grow:
|
705
|
+
while (len >= 10) {
|
706
|
+
rb_str_buf_cat(res, nul10, 10);
|
707
|
+
len -= 10;
|
708
|
+
}
|
709
|
+
rb_str_buf_cat(res, nul10, len);
|
710
|
+
break;
|
711
|
+
|
712
|
+
case 'X': /* back up byte */
|
713
|
+
shrink:
|
714
|
+
plen = RSTRING_LEN(res);
|
715
|
+
if (plen < len)
|
716
|
+
rb_raise(rb_eArgError, "X outside of string");
|
717
|
+
rb_str_set_len(res, plen - len);
|
718
|
+
break;
|
719
|
+
|
720
|
+
case '@': /* null fill to absolute position */
|
721
|
+
len -= RSTRING_LEN(res);
|
722
|
+
if (len > 0) goto grow;
|
723
|
+
len = -len;
|
724
|
+
if (len > 0) goto shrink;
|
725
|
+
break;
|
726
|
+
|
727
|
+
case '%':
|
728
|
+
rb_raise(rb_eArgError, "%% is not supported");
|
729
|
+
break;
|
730
|
+
|
731
|
+
case 'U': /* Unicode character */
|
732
|
+
while (len-- > 0) {
|
733
|
+
SIGNED_VALUE l;
|
734
|
+
char buf[8];
|
735
|
+
int le;
|
736
|
+
|
737
|
+
from = NEXTFROM;
|
738
|
+
from = rb_to_int(from);
|
739
|
+
l = NUM2LONG(from);
|
740
|
+
if (l < 0) {
|
741
|
+
rb_raise(rb_eRangeError, "pack(U): value out of range");
|
742
|
+
}
|
743
|
+
le = rb_uv_to_utf8(buf, l);
|
744
|
+
rb_str_buf_cat(res, (char*)buf, le);
|
745
|
+
}
|
746
|
+
break;
|
747
|
+
|
748
|
+
case 'u': /* uuencoded string */
|
749
|
+
case 'm': /* base64 encoded string */
|
750
|
+
from = NEXTFROM;
|
751
|
+
StringValue(from);
|
752
|
+
ptr = RSTRING_PTR(from);
|
753
|
+
plen = RSTRING_LEN(from);
|
754
|
+
|
755
|
+
if (len == 0 && type == 'm') {
|
756
|
+
encodes(res, ptr, plen, type, 0);
|
757
|
+
ptr += plen;
|
758
|
+
break;
|
759
|
+
}
|
760
|
+
if (len <= 2)
|
761
|
+
len = 45;
|
762
|
+
else if (len > 63 && type == 'u')
|
763
|
+
len = 63;
|
764
|
+
else
|
765
|
+
len = len / 3 * 3;
|
766
|
+
while (plen > 0) {
|
767
|
+
long todo;
|
768
|
+
|
769
|
+
if (plen > len)
|
770
|
+
todo = len;
|
771
|
+
else
|
772
|
+
todo = plen;
|
773
|
+
encodes(res, ptr, todo, type, 1);
|
774
|
+
plen -= todo;
|
775
|
+
ptr += todo;
|
776
|
+
}
|
777
|
+
break;
|
778
|
+
|
779
|
+
case 'M': /* quoted-printable encoded string */
|
780
|
+
from = rb_obj_as_string(NEXTFROM);
|
781
|
+
if (len <= 1)
|
782
|
+
len = 72;
|
783
|
+
qpencode(res, from, len);
|
784
|
+
break;
|
785
|
+
|
786
|
+
case 'P': /* pointer to packed byte string */
|
787
|
+
from = THISFROM;
|
788
|
+
if (!NIL_P(from)) {
|
789
|
+
StringValue(from);
|
790
|
+
if (RSTRING_LEN(from) < len) {
|
791
|
+
rb_raise(rb_eArgError, "too short buffer for P(%ld for %ld)",
|
792
|
+
RSTRING_LEN(from), len);
|
793
|
+
}
|
794
|
+
}
|
795
|
+
len = 1;
|
796
|
+
/* FALL THROUGH */
|
797
|
+
case 'p': /* pointer to string */
|
798
|
+
while (len-- > 0) {
|
799
|
+
char *t;
|
800
|
+
from = NEXTFROM;
|
801
|
+
if (NIL_P(from)) {
|
802
|
+
t = 0;
|
803
|
+
}
|
804
|
+
else {
|
805
|
+
t = StringValuePtr(from);
|
806
|
+
rb_obj_taint(from);
|
807
|
+
}
|
808
|
+
if (!associates) {
|
809
|
+
associates = rb_ary_new();
|
810
|
+
}
|
811
|
+
rb_ary_push(associates, from);
|
812
|
+
rb_str_buf_cat(res, (char*)&t, sizeof(char*));
|
813
|
+
}
|
814
|
+
break;
|
815
|
+
|
816
|
+
case 'w': /* BER compressed integer */
|
817
|
+
while (len-- > 0) {
|
818
|
+
VALUE buf = rb_str_new(0, 0);
|
819
|
+
size_t numbytes;
|
820
|
+
int sign;
|
821
|
+
char *cp;
|
822
|
+
|
823
|
+
from = NEXTFROM;
|
824
|
+
from = rb_to_int(from);
|
825
|
+
numbytes = rb_absint_numwords(from, 7, NULL);
|
826
|
+
if (numbytes == 0)
|
827
|
+
numbytes = 1;
|
828
|
+
buf = rb_str_new(NULL, numbytes);
|
829
|
+
|
830
|
+
sign = rb_integer_pack(from, RSTRING_PTR(buf), RSTRING_LEN(buf), 1, 1, INTEGER_PACK_BIG_ENDIAN);
|
831
|
+
|
832
|
+
if (sign < 0)
|
833
|
+
rb_raise(rb_eArgError, "can't compress negative numbers");
|
834
|
+
if (sign == 2)
|
835
|
+
rb_bug("buffer size problem?");
|
836
|
+
|
837
|
+
cp = RSTRING_PTR(buf);
|
838
|
+
while (1 < numbytes) {
|
839
|
+
*cp |= 0x80;
|
840
|
+
cp++;
|
841
|
+
numbytes--;
|
842
|
+
}
|
843
|
+
|
844
|
+
rb_str_buf_cat(res, RSTRING_PTR(buf), RSTRING_LEN(buf));
|
845
|
+
}
|
846
|
+
break;
|
847
|
+
|
848
|
+
default: {
|
849
|
+
char unknown[5];
|
850
|
+
if (ISPRINT(type)) {
|
851
|
+
unknown[0] = type;
|
852
|
+
unknown[1] = '\0';
|
853
|
+
}
|
854
|
+
else {
|
855
|
+
snprintf(unknown, sizeof(unknown), "\\x%.2x", type & 0xff);
|
856
|
+
}
|
857
|
+
rb_warning("unknown pack directive '%s' in '% "PRIsVALUE"'",
|
858
|
+
unknown, fmt);
|
859
|
+
break;
|
860
|
+
}
|
861
|
+
}
|
862
|
+
}
|
863
|
+
|
864
|
+
if (associates) {
|
865
|
+
str_associate(res, associates);
|
866
|
+
}
|
867
|
+
OBJ_INFECT(res, fmt);
|
868
|
+
switch (enc_info) {
|
869
|
+
case 1:
|
870
|
+
ENCODING_CODERANGE_SET(res, rb_usascii_encindex(), ENC_CODERANGE_7BIT);
|
871
|
+
break;
|
872
|
+
case 2:
|
873
|
+
rb_enc_set_index(res, rb_utf8_encindex());
|
874
|
+
break;
|
875
|
+
default:
|
876
|
+
/* do nothing, keep ASCII-8BIT */
|
877
|
+
break;
|
878
|
+
}
|
879
|
+
return res;
|
880
|
+
}
|
881
|
+
|
882
|
+
static const char uu_table[] =
|
883
|
+
"`!\"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\\]^_";
|
884
|
+
static const char b64_table[] =
|
885
|
+
"ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/";
|
886
|
+
|
887
|
+
static void
|
888
|
+
encodes(VALUE str, const char *s0, long len, int type, int tail_lf)
|
889
|
+
{
|
890
|
+
enum {buff_size = 4096, encoded_unit = 4, input_unit = 3};
|
891
|
+
char buff[buff_size + 1]; /* +1 for tail_lf */
|
892
|
+
long i = 0;
|
893
|
+
const char *const trans = type == 'u' ? uu_table : b64_table;
|
894
|
+
char padding;
|
895
|
+
const unsigned char *s = (const unsigned char *)s0;
|
896
|
+
|
897
|
+
if (type == 'u') {
|
898
|
+
buff[i++] = (char)len + ' ';
|
899
|
+
padding = '`';
|
900
|
+
}
|
901
|
+
else {
|
902
|
+
padding = '=';
|
903
|
+
}
|
904
|
+
while (len >= input_unit) {
|
905
|
+
while (len >= input_unit && buff_size-i >= encoded_unit) {
|
906
|
+
buff[i++] = trans[077 & (*s >> 2)];
|
907
|
+
buff[i++] = trans[077 & (((*s << 4) & 060) | ((s[1] >> 4) & 017))];
|
908
|
+
buff[i++] = trans[077 & (((s[1] << 2) & 074) | ((s[2] >> 6) & 03))];
|
909
|
+
buff[i++] = trans[077 & s[2]];
|
910
|
+
s += input_unit;
|
911
|
+
len -= input_unit;
|
912
|
+
}
|
913
|
+
if (buff_size-i < encoded_unit) {
|
914
|
+
rb_str_buf_cat(str, buff, i);
|
915
|
+
i = 0;
|
916
|
+
}
|
917
|
+
}
|
918
|
+
|
919
|
+
if (len == 2) {
|
920
|
+
buff[i++] = trans[077 & (*s >> 2)];
|
921
|
+
buff[i++] = trans[077 & (((*s << 4) & 060) | ((s[1] >> 4) & 017))];
|
922
|
+
buff[i++] = trans[077 & (((s[1] << 2) & 074) | (('\0' >> 6) & 03))];
|
923
|
+
buff[i++] = padding;
|
924
|
+
}
|
925
|
+
else if (len == 1) {
|
926
|
+
buff[i++] = trans[077 & (*s >> 2)];
|
927
|
+
buff[i++] = trans[077 & (((*s << 4) & 060) | (('\0' >> 4) & 017))];
|
928
|
+
buff[i++] = padding;
|
929
|
+
buff[i++] = padding;
|
930
|
+
}
|
931
|
+
if (tail_lf) buff[i++] = '\n';
|
932
|
+
rb_str_buf_cat(str, buff, i);
|
933
|
+
if ((size_t)i > sizeof(buff)) rb_bug("encodes() buffer overrun");
|
934
|
+
}
|
935
|
+
|
936
|
+
static const char hex_table[] = "0123456789ABCDEF";
|
937
|
+
|
938
|
+
static void
|
939
|
+
qpencode(VALUE str, VALUE from, long len)
|
940
|
+
{
|
941
|
+
char buff[1024];
|
942
|
+
long i = 0, n = 0, prev = EOF;
|
943
|
+
unsigned char *s = (unsigned char*)RSTRING_PTR(from);
|
944
|
+
unsigned char *send = s + RSTRING_LEN(from);
|
945
|
+
|
946
|
+
while (s < send) {
|
947
|
+
if ((*s > 126) ||
|
948
|
+
(*s < 32 && *s != '\n' && *s != '\t') ||
|
949
|
+
(*s == '=')) {
|
950
|
+
buff[i++] = '=';
|
951
|
+
buff[i++] = hex_table[*s >> 4];
|
952
|
+
buff[i++] = hex_table[*s & 0x0f];
|
953
|
+
n += 3;
|
954
|
+
prev = EOF;
|
955
|
+
}
|
956
|
+
else if (*s == '\n') {
|
957
|
+
if (prev == ' ' || prev == '\t') {
|
958
|
+
buff[i++] = '=';
|
959
|
+
buff[i++] = *s;
|
960
|
+
}
|
961
|
+
buff[i++] = *s;
|
962
|
+
n = 0;
|
963
|
+
prev = *s;
|
964
|
+
}
|
965
|
+
else {
|
966
|
+
buff[i++] = *s;
|
967
|
+
n++;
|
968
|
+
prev = *s;
|
969
|
+
}
|
970
|
+
if (n > len) {
|
971
|
+
buff[i++] = '=';
|
972
|
+
buff[i++] = '\n';
|
973
|
+
n = 0;
|
974
|
+
prev = '\n';
|
975
|
+
}
|
976
|
+
if (i > 1024 - 5) {
|
977
|
+
rb_str_buf_cat(str, buff, i);
|
978
|
+
i = 0;
|
979
|
+
}
|
980
|
+
s++;
|
981
|
+
}
|
982
|
+
if (n > 0) {
|
983
|
+
buff[i++] = '=';
|
984
|
+
buff[i++] = '\n';
|
985
|
+
}
|
986
|
+
if (i > 0) {
|
987
|
+
rb_str_buf_cat(str, buff, i);
|
988
|
+
}
|
989
|
+
}
|
990
|
+
|
991
|
+
static inline int
|
992
|
+
hex2num(char c)
|
993
|
+
{
|
994
|
+
int n;
|
995
|
+
n = ruby_digit36_to_number_table[(unsigned char)c];
|
996
|
+
if (16 <= n)
|
997
|
+
n = -1;
|
998
|
+
return n;
|
999
|
+
}
|
1000
|
+
|
1001
|
+
#define PACK_LENGTH_ADJUST_SIZE(sz) do { \
|
1002
|
+
tmp_len = 0; \
|
1003
|
+
if (len > (long)((send-s)/(sz))) { \
|
1004
|
+
if (!star) { \
|
1005
|
+
tmp_len = len-(send-s)/(sz); \
|
1006
|
+
} \
|
1007
|
+
len = (send-s)/(sz); \
|
1008
|
+
} \
|
1009
|
+
} while (0)
|
1010
|
+
|
1011
|
+
#define PACK_ITEM_ADJUST() do { \
|
1012
|
+
if (tmp_len > 0 && mode == UNPACK_ARRAY) \
|
1013
|
+
rb_ary_store(ary, RARRAY_LEN(ary)+tmp_len-1, Qnil); \
|
1014
|
+
} while (0)
|
1015
|
+
|
1016
|
+
/* Workaround for Oracle Developer Studio (Oracle Solaris Studio)
|
1017
|
+
* 12.4/12.5/12.6 C compiler optimization bug
|
1018
|
+
* with "-xO4" optimization option.
|
1019
|
+
*/
|
1020
|
+
#if defined(__SUNPRO_C) && 0x5130 <= __SUNPRO_C && __SUNPRO_C <= 0x5150
|
1021
|
+
# define AVOID_CC_BUG volatile
|
1022
|
+
#else
|
1023
|
+
# define AVOID_CC_BUG
|
1024
|
+
#endif
|
1025
|
+
|
1026
|
+
static VALUE
|
1027
|
+
infected_str_new(const char *ptr, long len, VALUE str)
|
1028
|
+
{
|
1029
|
+
VALUE s = rb_str_new(ptr, len);
|
1030
|
+
|
1031
|
+
OBJ_INFECT(s, str);
|
1032
|
+
return s;
|
1033
|
+
}
|
1034
|
+
|
1035
|
+
/* unpack mode */
|
1036
|
+
#define UNPACK_ARRAY 0
|
1037
|
+
#define UNPACK_BLOCK 1
|
1038
|
+
#define UNPACK_1 2
|
1039
|
+
|
1040
|
+
static VALUE
|
1041
|
+
pack_unpack_internal(VALUE str, VALUE fmt, int mode)
|
1042
|
+
{
|
1043
|
+
#define hexdigits ruby_hexdigits
|
1044
|
+
char *s, *send;
|
1045
|
+
char *p, *pend;
|
1046
|
+
VALUE ary;
|
1047
|
+
char type;
|
1048
|
+
long len;
|
1049
|
+
AVOID_CC_BUG long tmp_len;
|
1050
|
+
int star;
|
1051
|
+
#ifdef NATINT_PACK
|
1052
|
+
int natint; /* native integer */
|
1053
|
+
#endif
|
1054
|
+
int signed_p, integer_size, bigendian_p;
|
1055
|
+
#define UNPACK_PUSH(item) do {\
|
1056
|
+
VALUE item_val = (item);\
|
1057
|
+
if ((mode) == UNPACK_BLOCK) {\
|
1058
|
+
rb_yield(item_val);\
|
1059
|
+
}\
|
1060
|
+
else if ((mode) == UNPACK_ARRAY) {\
|
1061
|
+
rb_ary_push(ary, item_val);\
|
1062
|
+
}\
|
1063
|
+
else /* if ((mode) == UNPACK_1) { */ {\
|
1064
|
+
return item_val; \
|
1065
|
+
}\
|
1066
|
+
} while (0)
|
1067
|
+
|
1068
|
+
StringValue(str);
|
1069
|
+
StringValue(fmt);
|
1070
|
+
s = RSTRING_PTR(str);
|
1071
|
+
send = s + RSTRING_LEN(str);
|
1072
|
+
p = RSTRING_PTR(fmt);
|
1073
|
+
pend = p + RSTRING_LEN(fmt);
|
1074
|
+
|
1075
|
+
ary = mode == UNPACK_ARRAY ? rb_ary_new() : Qnil;
|
1076
|
+
while (p < pend) {
|
1077
|
+
int explicit_endian = 0;
|
1078
|
+
type = *p++;
|
1079
|
+
#ifdef NATINT_PACK
|
1080
|
+
natint = 0;
|
1081
|
+
#endif
|
1082
|
+
|
1083
|
+
if (ISSPACE(type)) continue;
|
1084
|
+
if (type == '#') {
|
1085
|
+
while ((p < pend) && (*p != '\n')) {
|
1086
|
+
p++;
|
1087
|
+
}
|
1088
|
+
continue;
|
1089
|
+
}
|
1090
|
+
|
1091
|
+
star = 0;
|
1092
|
+
{
|
1093
|
+
modifiers:
|
1094
|
+
switch (*p) {
|
1095
|
+
case '_':
|
1096
|
+
case '!':
|
1097
|
+
|
1098
|
+
if (strchr(natstr, type)) {
|
1099
|
+
#ifdef NATINT_PACK
|
1100
|
+
natint = 1;
|
1101
|
+
#endif
|
1102
|
+
p++;
|
1103
|
+
}
|
1104
|
+
else {
|
1105
|
+
rb_raise(rb_eArgError, "'%c' allowed only after types %s", *p, natstr);
|
1106
|
+
}
|
1107
|
+
goto modifiers;
|
1108
|
+
|
1109
|
+
case '<':
|
1110
|
+
case '>':
|
1111
|
+
if (!strchr(endstr, type)) {
|
1112
|
+
rb_raise(rb_eArgError, "'%c' allowed only after types %s", *p, endstr);
|
1113
|
+
}
|
1114
|
+
if (explicit_endian) {
|
1115
|
+
rb_raise(rb_eRangeError, "Can't use both '<' and '>'");
|
1116
|
+
}
|
1117
|
+
explicit_endian = *p++;
|
1118
|
+
goto modifiers;
|
1119
|
+
}
|
1120
|
+
}
|
1121
|
+
|
1122
|
+
if (p >= pend)
|
1123
|
+
len = 1;
|
1124
|
+
else if (*p == '*') {
|
1125
|
+
star = 1;
|
1126
|
+
len = send - s;
|
1127
|
+
p++;
|
1128
|
+
}
|
1129
|
+
else if (ISDIGIT(*p)) {
|
1130
|
+
errno = 0;
|
1131
|
+
len = STRTOUL(p, (char**)&p, 10);
|
1132
|
+
if (len < 0 || errno) {
|
1133
|
+
rb_raise(rb_eRangeError, "pack length too big");
|
1134
|
+
}
|
1135
|
+
}
|
1136
|
+
else {
|
1137
|
+
len = (type != '@');
|
1138
|
+
}
|
1139
|
+
|
1140
|
+
switch (type) {
|
1141
|
+
case '%':
|
1142
|
+
rb_raise(rb_eArgError, "%% is not supported");
|
1143
|
+
break;
|
1144
|
+
|
1145
|
+
case 'A':
|
1146
|
+
if (len > send - s) len = send - s;
|
1147
|
+
{
|
1148
|
+
long end = len;
|
1149
|
+
char *t = s + len - 1;
|
1150
|
+
|
1151
|
+
while (t >= s) {
|
1152
|
+
if (*t != ' ' && *t != '\0') break;
|
1153
|
+
t--; len--;
|
1154
|
+
}
|
1155
|
+
UNPACK_PUSH(infected_str_new(s, len, str));
|
1156
|
+
s += end;
|
1157
|
+
}
|
1158
|
+
break;
|
1159
|
+
|
1160
|
+
case 'Z':
|
1161
|
+
{
|
1162
|
+
char *t = s;
|
1163
|
+
|
1164
|
+
if (len > send-s) len = send-s;
|
1165
|
+
while (t < s+len && *t) t++;
|
1166
|
+
UNPACK_PUSH(infected_str_new(s, t-s, str));
|
1167
|
+
if (t < send) t++;
|
1168
|
+
s = star ? t : s+len;
|
1169
|
+
}
|
1170
|
+
break;
|
1171
|
+
|
1172
|
+
case 'a':
|
1173
|
+
if (len > send - s) len = send - s;
|
1174
|
+
UNPACK_PUSH(infected_str_new(s, len, str));
|
1175
|
+
s += len;
|
1176
|
+
break;
|
1177
|
+
|
1178
|
+
case 'b':
|
1179
|
+
{
|
1180
|
+
VALUE bitstr;
|
1181
|
+
char *t;
|
1182
|
+
int bits;
|
1183
|
+
long i;
|
1184
|
+
|
1185
|
+
if (p[-1] == '*' || len > (send - s) * 8)
|
1186
|
+
len = (send - s) * 8;
|
1187
|
+
bits = 0;
|
1188
|
+
bitstr = rb_usascii_str_new(0, len);
|
1189
|
+
t = RSTRING_PTR(bitstr);
|
1190
|
+
for (i=0; i<len; i++) {
|
1191
|
+
if (i & 7) bits >>= 1;
|
1192
|
+
else bits = (unsigned char)*s++;
|
1193
|
+
*t++ = (bits & 1) ? '1' : '0';
|
1194
|
+
}
|
1195
|
+
UNPACK_PUSH(bitstr);
|
1196
|
+
}
|
1197
|
+
break;
|
1198
|
+
|
1199
|
+
case 'B':
|
1200
|
+
{
|
1201
|
+
VALUE bitstr;
|
1202
|
+
char *t;
|
1203
|
+
int bits;
|
1204
|
+
long i;
|
1205
|
+
|
1206
|
+
if (p[-1] == '*' || len > (send - s) * 8)
|
1207
|
+
len = (send - s) * 8;
|
1208
|
+
bits = 0;
|
1209
|
+
bitstr = rb_usascii_str_new(0, len);
|
1210
|
+
t = RSTRING_PTR(bitstr);
|
1211
|
+
for (i=0; i<len; i++) {
|
1212
|
+
if (i & 7) bits <<= 1;
|
1213
|
+
else bits = (unsigned char)*s++;
|
1214
|
+
*t++ = (bits & 128) ? '1' : '0';
|
1215
|
+
}
|
1216
|
+
UNPACK_PUSH(bitstr);
|
1217
|
+
}
|
1218
|
+
break;
|
1219
|
+
|
1220
|
+
case 'h':
|
1221
|
+
{
|
1222
|
+
VALUE bitstr;
|
1223
|
+
char *t;
|
1224
|
+
int bits;
|
1225
|
+
long i;
|
1226
|
+
|
1227
|
+
if (p[-1] == '*' || len > (send - s) * 2)
|
1228
|
+
len = (send - s) * 2;
|
1229
|
+
bits = 0;
|
1230
|
+
bitstr = rb_usascii_str_new(0, len);
|
1231
|
+
t = RSTRING_PTR(bitstr);
|
1232
|
+
for (i=0; i<len; i++) {
|
1233
|
+
if (i & 1)
|
1234
|
+
bits >>= 4;
|
1235
|
+
else
|
1236
|
+
bits = (unsigned char)*s++;
|
1237
|
+
*t++ = hexdigits[bits & 15];
|
1238
|
+
}
|
1239
|
+
UNPACK_PUSH(bitstr);
|
1240
|
+
}
|
1241
|
+
break;
|
1242
|
+
|
1243
|
+
case 'H':
|
1244
|
+
{
|
1245
|
+
VALUE bitstr;
|
1246
|
+
char *t;
|
1247
|
+
int bits;
|
1248
|
+
long i;
|
1249
|
+
|
1250
|
+
if (p[-1] == '*' || len > (send - s) * 2)
|
1251
|
+
len = (send - s) * 2;
|
1252
|
+
bits = 0;
|
1253
|
+
bitstr = rb_usascii_str_new(0, len);
|
1254
|
+
t = RSTRING_PTR(bitstr);
|
1255
|
+
for (i=0; i<len; i++) {
|
1256
|
+
if (i & 1)
|
1257
|
+
bits <<= 4;
|
1258
|
+
else
|
1259
|
+
bits = (unsigned char)*s++;
|
1260
|
+
*t++ = hexdigits[(bits >> 4) & 15];
|
1261
|
+
}
|
1262
|
+
UNPACK_PUSH(bitstr);
|
1263
|
+
}
|
1264
|
+
break;
|
1265
|
+
|
1266
|
+
case 'c':
|
1267
|
+
signed_p = 1;
|
1268
|
+
integer_size = 1;
|
1269
|
+
bigendian_p = BIGENDIAN_P(); /* not effective */
|
1270
|
+
goto unpack_integer;
|
1271
|
+
|
1272
|
+
case 'C':
|
1273
|
+
signed_p = 0;
|
1274
|
+
integer_size = 1;
|
1275
|
+
bigendian_p = BIGENDIAN_P(); /* not effective */
|
1276
|
+
goto unpack_integer;
|
1277
|
+
|
1278
|
+
case 's':
|
1279
|
+
signed_p = 1;
|
1280
|
+
integer_size = NATINT_LEN(short, 2);
|
1281
|
+
bigendian_p = BIGENDIAN_P();
|
1282
|
+
goto unpack_integer;
|
1283
|
+
|
1284
|
+
case 'S':
|
1285
|
+
signed_p = 0;
|
1286
|
+
integer_size = NATINT_LEN(short, 2);
|
1287
|
+
bigendian_p = BIGENDIAN_P();
|
1288
|
+
goto unpack_integer;
|
1289
|
+
|
1290
|
+
case 'i':
|
1291
|
+
signed_p = 1;
|
1292
|
+
integer_size = (int)sizeof(int);
|
1293
|
+
bigendian_p = BIGENDIAN_P();
|
1294
|
+
goto unpack_integer;
|
1295
|
+
|
1296
|
+
case 'I':
|
1297
|
+
signed_p = 0;
|
1298
|
+
integer_size = (int)sizeof(int);
|
1299
|
+
bigendian_p = BIGENDIAN_P();
|
1300
|
+
goto unpack_integer;
|
1301
|
+
|
1302
|
+
case 'l':
|
1303
|
+
signed_p = 1;
|
1304
|
+
integer_size = NATINT_LEN(long, 4);
|
1305
|
+
bigendian_p = BIGENDIAN_P();
|
1306
|
+
goto unpack_integer;
|
1307
|
+
|
1308
|
+
case 'L':
|
1309
|
+
signed_p = 0;
|
1310
|
+
integer_size = NATINT_LEN(long, 4);
|
1311
|
+
bigendian_p = BIGENDIAN_P();
|
1312
|
+
goto unpack_integer;
|
1313
|
+
|
1314
|
+
case 'q':
|
1315
|
+
signed_p = 1;
|
1316
|
+
integer_size = NATINT_LEN_Q;
|
1317
|
+
bigendian_p = BIGENDIAN_P();
|
1318
|
+
goto unpack_integer;
|
1319
|
+
|
1320
|
+
case 'Q':
|
1321
|
+
signed_p = 0;
|
1322
|
+
integer_size = NATINT_LEN_Q;
|
1323
|
+
bigendian_p = BIGENDIAN_P();
|
1324
|
+
goto unpack_integer;
|
1325
|
+
|
1326
|
+
case 'j':
|
1327
|
+
signed_p = 1;
|
1328
|
+
integer_size = sizeof(intptr_t);
|
1329
|
+
bigendian_p = BIGENDIAN_P();
|
1330
|
+
goto unpack_integer;
|
1331
|
+
|
1332
|
+
case 'J':
|
1333
|
+
signed_p = 0;
|
1334
|
+
integer_size = sizeof(uintptr_t);
|
1335
|
+
bigendian_p = BIGENDIAN_P();
|
1336
|
+
goto unpack_integer;
|
1337
|
+
|
1338
|
+
case 'n':
|
1339
|
+
signed_p = 0;
|
1340
|
+
integer_size = 2;
|
1341
|
+
bigendian_p = 1;
|
1342
|
+
goto unpack_integer;
|
1343
|
+
|
1344
|
+
case 'N':
|
1345
|
+
signed_p = 0;
|
1346
|
+
integer_size = 4;
|
1347
|
+
bigendian_p = 1;
|
1348
|
+
goto unpack_integer;
|
1349
|
+
|
1350
|
+
case 'v':
|
1351
|
+
signed_p = 0;
|
1352
|
+
integer_size = 2;
|
1353
|
+
bigendian_p = 0;
|
1354
|
+
goto unpack_integer;
|
1355
|
+
|
1356
|
+
case 'V':
|
1357
|
+
signed_p = 0;
|
1358
|
+
integer_size = 4;
|
1359
|
+
bigendian_p = 0;
|
1360
|
+
goto unpack_integer;
|
1361
|
+
|
1362
|
+
unpack_integer:
|
1363
|
+
if (explicit_endian) {
|
1364
|
+
bigendian_p = explicit_endian == '>';
|
1365
|
+
}
|
1366
|
+
PACK_LENGTH_ADJUST_SIZE(integer_size);
|
1367
|
+
while (len-- > 0) {
|
1368
|
+
int flags = bigendian_p ? INTEGER_PACK_BIG_ENDIAN : INTEGER_PACK_LITTLE_ENDIAN;
|
1369
|
+
VALUE val;
|
1370
|
+
if (signed_p)
|
1371
|
+
flags |= INTEGER_PACK_2COMP;
|
1372
|
+
val = rb_integer_unpack(s, integer_size, 1, 0, flags);
|
1373
|
+
UNPACK_PUSH(val);
|
1374
|
+
s += integer_size;
|
1375
|
+
}
|
1376
|
+
PACK_ITEM_ADJUST();
|
1377
|
+
break;
|
1378
|
+
|
1379
|
+
case 'f':
|
1380
|
+
case 'F':
|
1381
|
+
PACK_LENGTH_ADJUST_SIZE(sizeof(float));
|
1382
|
+
while (len-- > 0) {
|
1383
|
+
float tmp;
|
1384
|
+
memcpy(&tmp, s, sizeof(float));
|
1385
|
+
s += sizeof(float);
|
1386
|
+
UNPACK_PUSH(DBL2NUM((double)tmp));
|
1387
|
+
}
|
1388
|
+
PACK_ITEM_ADJUST();
|
1389
|
+
break;
|
1390
|
+
|
1391
|
+
case 'e':
|
1392
|
+
PACK_LENGTH_ADJUST_SIZE(sizeof(float));
|
1393
|
+
while (len-- > 0) {
|
1394
|
+
FLOAT_CONVWITH(tmp);
|
1395
|
+
memcpy(tmp.buf, s, sizeof(float));
|
1396
|
+
s += sizeof(float);
|
1397
|
+
VTOHF(tmp);
|
1398
|
+
UNPACK_PUSH(DBL2NUM(tmp.f));
|
1399
|
+
}
|
1400
|
+
PACK_ITEM_ADJUST();
|
1401
|
+
break;
|
1402
|
+
|
1403
|
+
case 'E':
|
1404
|
+
PACK_LENGTH_ADJUST_SIZE(sizeof(double));
|
1405
|
+
while (len-- > 0) {
|
1406
|
+
DOUBLE_CONVWITH(tmp);
|
1407
|
+
memcpy(tmp.buf, s, sizeof(double));
|
1408
|
+
s += sizeof(double);
|
1409
|
+
VTOHD(tmp);
|
1410
|
+
UNPACK_PUSH(DBL2NUM(tmp.d));
|
1411
|
+
}
|
1412
|
+
PACK_ITEM_ADJUST();
|
1413
|
+
break;
|
1414
|
+
|
1415
|
+
case 'D':
|
1416
|
+
case 'd':
|
1417
|
+
PACK_LENGTH_ADJUST_SIZE(sizeof(double));
|
1418
|
+
while (len-- > 0) {
|
1419
|
+
double tmp;
|
1420
|
+
memcpy(&tmp, s, sizeof(double));
|
1421
|
+
s += sizeof(double);
|
1422
|
+
UNPACK_PUSH(DBL2NUM(tmp));
|
1423
|
+
}
|
1424
|
+
PACK_ITEM_ADJUST();
|
1425
|
+
break;
|
1426
|
+
|
1427
|
+
case 'g':
|
1428
|
+
PACK_LENGTH_ADJUST_SIZE(sizeof(float));
|
1429
|
+
while (len-- > 0) {
|
1430
|
+
FLOAT_CONVWITH(tmp);
|
1431
|
+
memcpy(tmp.buf, s, sizeof(float));
|
1432
|
+
s += sizeof(float);
|
1433
|
+
NTOHF(tmp);
|
1434
|
+
UNPACK_PUSH(DBL2NUM(tmp.f));
|
1435
|
+
}
|
1436
|
+
PACK_ITEM_ADJUST();
|
1437
|
+
break;
|
1438
|
+
|
1439
|
+
case 'G':
|
1440
|
+
PACK_LENGTH_ADJUST_SIZE(sizeof(double));
|
1441
|
+
while (len-- > 0) {
|
1442
|
+
DOUBLE_CONVWITH(tmp);
|
1443
|
+
memcpy(tmp.buf, s, sizeof(double));
|
1444
|
+
s += sizeof(double);
|
1445
|
+
NTOHD(tmp);
|
1446
|
+
UNPACK_PUSH(DBL2NUM(tmp.d));
|
1447
|
+
}
|
1448
|
+
PACK_ITEM_ADJUST();
|
1449
|
+
break;
|
1450
|
+
|
1451
|
+
case 'U':
|
1452
|
+
if (len > send - s) len = send - s;
|
1453
|
+
while (len > 0 && s < send) {
|
1454
|
+
long alen = send - s;
|
1455
|
+
unsigned long l;
|
1456
|
+
|
1457
|
+
l = utf8_to_uv(s, &alen);
|
1458
|
+
s += alen; len--;
|
1459
|
+
UNPACK_PUSH(ULONG2NUM(l));
|
1460
|
+
}
|
1461
|
+
break;
|
1462
|
+
|
1463
|
+
case 'u':
|
1464
|
+
{
|
1465
|
+
VALUE buf = infected_str_new(0, (send - s)*3/4, str);
|
1466
|
+
char *ptr = RSTRING_PTR(buf);
|
1467
|
+
long total = 0;
|
1468
|
+
|
1469
|
+
while (s < send && (unsigned char)*s > ' ' && (unsigned char)*s < 'a') {
|
1470
|
+
long a,b,c,d;
|
1471
|
+
char hunk[3];
|
1472
|
+
|
1473
|
+
len = ((unsigned char)*s++ - ' ') & 077;
|
1474
|
+
|
1475
|
+
total += len;
|
1476
|
+
if (total > RSTRING_LEN(buf)) {
|
1477
|
+
len -= total - RSTRING_LEN(buf);
|
1478
|
+
total = RSTRING_LEN(buf);
|
1479
|
+
}
|
1480
|
+
|
1481
|
+
while (len > 0) {
|
1482
|
+
long mlen = len > 3 ? 3 : len;
|
1483
|
+
|
1484
|
+
if (s < send && (unsigned char)*s >= ' ' && (unsigned char)*s < 'a')
|
1485
|
+
a = ((unsigned char)*s++ - ' ') & 077;
|
1486
|
+
else
|
1487
|
+
a = 0;
|
1488
|
+
if (s < send && (unsigned char)*s >= ' ' && (unsigned char)*s < 'a')
|
1489
|
+
b = ((unsigned char)*s++ - ' ') & 077;
|
1490
|
+
else
|
1491
|
+
b = 0;
|
1492
|
+
if (s < send && (unsigned char)*s >= ' ' && (unsigned char)*s < 'a')
|
1493
|
+
c = ((unsigned char)*s++ - ' ') & 077;
|
1494
|
+
else
|
1495
|
+
c = 0;
|
1496
|
+
if (s < send && (unsigned char)*s >= ' ' && (unsigned char)*s < 'a')
|
1497
|
+
d = ((unsigned char)*s++ - ' ') & 077;
|
1498
|
+
else
|
1499
|
+
d = 0;
|
1500
|
+
hunk[0] = (char)(a << 2 | b >> 4);
|
1501
|
+
hunk[1] = (char)(b << 4 | c >> 2);
|
1502
|
+
hunk[2] = (char)(c << 6 | d);
|
1503
|
+
memcpy(ptr, hunk, mlen);
|
1504
|
+
ptr += mlen;
|
1505
|
+
len -= mlen;
|
1506
|
+
}
|
1507
|
+
if (s < send && (unsigned char)*s != '\r' && *s != '\n')
|
1508
|
+
s++; /* possible checksum byte */
|
1509
|
+
if (s < send && *s == '\r') s++;
|
1510
|
+
if (s < send && *s == '\n') s++;
|
1511
|
+
}
|
1512
|
+
|
1513
|
+
rb_str_set_len(buf, total);
|
1514
|
+
UNPACK_PUSH(buf);
|
1515
|
+
}
|
1516
|
+
break;
|
1517
|
+
|
1518
|
+
case 'm':
|
1519
|
+
{
|
1520
|
+
VALUE buf = infected_str_new(0, (send - s + 3)*3/4, str); /* +3 is for skipping paddings */
|
1521
|
+
char *ptr = RSTRING_PTR(buf);
|
1522
|
+
int a = -1,b = -1,c = 0,d = 0;
|
1523
|
+
static signed char b64_xtable[256];
|
1524
|
+
|
1525
|
+
if (b64_xtable['/'] <= 0) {
|
1526
|
+
int i;
|
1527
|
+
|
1528
|
+
for (i = 0; i < 256; i++) {
|
1529
|
+
b64_xtable[i] = -1;
|
1530
|
+
}
|
1531
|
+
for (i = 0; i < 64; i++) {
|
1532
|
+
b64_xtable[(unsigned char)b64_table[i]] = (char)i;
|
1533
|
+
}
|
1534
|
+
}
|
1535
|
+
if (len == 0) {
|
1536
|
+
while (s < send) {
|
1537
|
+
a = b = c = d = -1;
|
1538
|
+
a = b64_xtable[(unsigned char)*s++];
|
1539
|
+
if (s >= send || a == -1) rb_raise(rb_eArgError, "invalid base64");
|
1540
|
+
b = b64_xtable[(unsigned char)*s++];
|
1541
|
+
if (s >= send || b == -1) rb_raise(rb_eArgError, "invalid base64");
|
1542
|
+
if (*s == '=') {
|
1543
|
+
if (s + 2 == send && *(s + 1) == '=') break;
|
1544
|
+
rb_raise(rb_eArgError, "invalid base64");
|
1545
|
+
}
|
1546
|
+
c = b64_xtable[(unsigned char)*s++];
|
1547
|
+
if (s >= send || c == -1) rb_raise(rb_eArgError, "invalid base64");
|
1548
|
+
if (s + 1 == send && *s == '=') break;
|
1549
|
+
d = b64_xtable[(unsigned char)*s++];
|
1550
|
+
if (d == -1) rb_raise(rb_eArgError, "invalid base64");
|
1551
|
+
*ptr++ = castchar(a << 2 | b >> 4);
|
1552
|
+
*ptr++ = castchar(b << 4 | c >> 2);
|
1553
|
+
*ptr++ = castchar(c << 6 | d);
|
1554
|
+
}
|
1555
|
+
if (c == -1) {
|
1556
|
+
*ptr++ = castchar(a << 2 | b >> 4);
|
1557
|
+
if (b & 0xf) rb_raise(rb_eArgError, "invalid base64");
|
1558
|
+
}
|
1559
|
+
else if (d == -1) {
|
1560
|
+
*ptr++ = castchar(a << 2 | b >> 4);
|
1561
|
+
*ptr++ = castchar(b << 4 | c >> 2);
|
1562
|
+
if (c & 0x3) rb_raise(rb_eArgError, "invalid base64");
|
1563
|
+
}
|
1564
|
+
}
|
1565
|
+
else {
|
1566
|
+
while (s < send) {
|
1567
|
+
a = b = c = d = -1;
|
1568
|
+
while ((a = b64_xtable[(unsigned char)*s]) == -1 && s < send) {s++;}
|
1569
|
+
if (s >= send) break;
|
1570
|
+
s++;
|
1571
|
+
while ((b = b64_xtable[(unsigned char)*s]) == -1 && s < send) {s++;}
|
1572
|
+
if (s >= send) break;
|
1573
|
+
s++;
|
1574
|
+
while ((c = b64_xtable[(unsigned char)*s]) == -1 && s < send) {if (*s == '=') break; s++;}
|
1575
|
+
if (*s == '=' || s >= send) break;
|
1576
|
+
s++;
|
1577
|
+
while ((d = b64_xtable[(unsigned char)*s]) == -1 && s < send) {if (*s == '=') break; s++;}
|
1578
|
+
if (*s == '=' || s >= send) break;
|
1579
|
+
s++;
|
1580
|
+
*ptr++ = castchar(a << 2 | b >> 4);
|
1581
|
+
*ptr++ = castchar(b << 4 | c >> 2);
|
1582
|
+
*ptr++ = castchar(c << 6 | d);
|
1583
|
+
a = -1;
|
1584
|
+
}
|
1585
|
+
if (a != -1 && b != -1) {
|
1586
|
+
if (c == -1)
|
1587
|
+
*ptr++ = castchar(a << 2 | b >> 4);
|
1588
|
+
else {
|
1589
|
+
*ptr++ = castchar(a << 2 | b >> 4);
|
1590
|
+
*ptr++ = castchar(b << 4 | c >> 2);
|
1591
|
+
}
|
1592
|
+
}
|
1593
|
+
}
|
1594
|
+
rb_str_set_len(buf, ptr - RSTRING_PTR(buf));
|
1595
|
+
UNPACK_PUSH(buf);
|
1596
|
+
}
|
1597
|
+
break;
|
1598
|
+
|
1599
|
+
case 'M':
|
1600
|
+
{
|
1601
|
+
VALUE buf = infected_str_new(0, send - s, str);
|
1602
|
+
char *ptr = RSTRING_PTR(buf), *ss = s;
|
1603
|
+
int csum = 0;
|
1604
|
+
int c1, c2;
|
1605
|
+
|
1606
|
+
while (s < send) {
|
1607
|
+
if (*s == '=') {
|
1608
|
+
if (++s == send) break;
|
1609
|
+
if (s+1 < send && *s == '\r' && *(s+1) == '\n')
|
1610
|
+
s++;
|
1611
|
+
if (*s != '\n') {
|
1612
|
+
if ((c1 = hex2num(*s)) == -1) break;
|
1613
|
+
if (++s == send) break;
|
1614
|
+
if ((c2 = hex2num(*s)) == -1) break;
|
1615
|
+
csum |= *ptr++ = castchar(c1 << 4 | c2);
|
1616
|
+
}
|
1617
|
+
}
|
1618
|
+
else {
|
1619
|
+
csum |= *ptr++ = *s;
|
1620
|
+
}
|
1621
|
+
s++;
|
1622
|
+
ss = s;
|
1623
|
+
}
|
1624
|
+
rb_str_set_len(buf, ptr - RSTRING_PTR(buf));
|
1625
|
+
rb_str_buf_cat(buf, ss, send-ss);
|
1626
|
+
csum = ISASCII(csum) ? ENC_CODERANGE_7BIT : ENC_CODERANGE_VALID;
|
1627
|
+
ENCODING_CODERANGE_SET(buf, rb_ascii8bit_encindex(), csum);
|
1628
|
+
UNPACK_PUSH(buf);
|
1629
|
+
}
|
1630
|
+
break;
|
1631
|
+
|
1632
|
+
case '@':
|
1633
|
+
if (len > RSTRING_LEN(str))
|
1634
|
+
rb_raise(rb_eArgError, "@ outside of string");
|
1635
|
+
s = RSTRING_PTR(str) + len;
|
1636
|
+
break;
|
1637
|
+
|
1638
|
+
case 'X':
|
1639
|
+
if (len > s - RSTRING_PTR(str))
|
1640
|
+
rb_raise(rb_eArgError, "X outside of string");
|
1641
|
+
s -= len;
|
1642
|
+
break;
|
1643
|
+
|
1644
|
+
case 'x':
|
1645
|
+
if (len > send - s)
|
1646
|
+
rb_raise(rb_eArgError, "x outside of string");
|
1647
|
+
s += len;
|
1648
|
+
break;
|
1649
|
+
|
1650
|
+
case 'P':
|
1651
|
+
if (sizeof(char *) <= (size_t)(send - s)) {
|
1652
|
+
VALUE tmp = Qnil;
|
1653
|
+
char *t;
|
1654
|
+
|
1655
|
+
memcpy(&t, s, sizeof(char *));
|
1656
|
+
s += sizeof(char *);
|
1657
|
+
|
1658
|
+
if (t) {
|
1659
|
+
VALUE a;
|
1660
|
+
const VALUE *p, *pend;
|
1661
|
+
|
1662
|
+
if (!(a = str_associated(str))) {
|
1663
|
+
rb_raise(rb_eArgError, "no associated pointer");
|
1664
|
+
}
|
1665
|
+
p = RARRAY_CONST_PTR(a);
|
1666
|
+
pend = p + RARRAY_LEN(a);
|
1667
|
+
while (p < pend) {
|
1668
|
+
if (RB_TYPE_P(*p, T_STRING) && RSTRING_PTR(*p) == t) {
|
1669
|
+
if (len < RSTRING_LEN(*p)) {
|
1670
|
+
tmp = rb_tainted_str_new(t, len);
|
1671
|
+
str_associate(tmp, a);
|
1672
|
+
}
|
1673
|
+
else {
|
1674
|
+
tmp = *p;
|
1675
|
+
}
|
1676
|
+
break;
|
1677
|
+
}
|
1678
|
+
p++;
|
1679
|
+
}
|
1680
|
+
if (p == pend) {
|
1681
|
+
rb_raise(rb_eArgError, "non associated pointer");
|
1682
|
+
}
|
1683
|
+
}
|
1684
|
+
UNPACK_PUSH(tmp);
|
1685
|
+
}
|
1686
|
+
break;
|
1687
|
+
|
1688
|
+
case 'p':
|
1689
|
+
if (len > (long)((send - s) / sizeof(char *)))
|
1690
|
+
len = (send - s) / sizeof(char *);
|
1691
|
+
while (len-- > 0) {
|
1692
|
+
if ((size_t)(send - s) < sizeof(char *))
|
1693
|
+
break;
|
1694
|
+
else {
|
1695
|
+
VALUE tmp = Qnil;
|
1696
|
+
char *t;
|
1697
|
+
|
1698
|
+
memcpy(&t, s, sizeof(char *));
|
1699
|
+
s += sizeof(char *);
|
1700
|
+
|
1701
|
+
if (t) {
|
1702
|
+
VALUE a;
|
1703
|
+
const VALUE *p, *pend;
|
1704
|
+
|
1705
|
+
if (!(a = str_associated(str))) {
|
1706
|
+
rb_raise(rb_eArgError, "no associated pointer");
|
1707
|
+
}
|
1708
|
+
p = RARRAY_CONST_PTR(a);
|
1709
|
+
pend = p + RARRAY_LEN(a);
|
1710
|
+
while (p < pend) {
|
1711
|
+
if (RB_TYPE_P(*p, T_STRING) && RSTRING_PTR(*p) == t) {
|
1712
|
+
tmp = *p;
|
1713
|
+
break;
|
1714
|
+
}
|
1715
|
+
p++;
|
1716
|
+
}
|
1717
|
+
if (p == pend) {
|
1718
|
+
rb_raise(rb_eArgError, "non associated pointer");
|
1719
|
+
}
|
1720
|
+
}
|
1721
|
+
UNPACK_PUSH(tmp);
|
1722
|
+
}
|
1723
|
+
}
|
1724
|
+
break;
|
1725
|
+
|
1726
|
+
case 'w':
|
1727
|
+
{
|
1728
|
+
char *s0 = s;
|
1729
|
+
while (len > 0 && s < send) {
|
1730
|
+
if (*s & 0x80) {
|
1731
|
+
s++;
|
1732
|
+
}
|
1733
|
+
else {
|
1734
|
+
s++;
|
1735
|
+
UNPACK_PUSH(rb_integer_unpack(s0, s-s0, 1, 1, INTEGER_PACK_BIG_ENDIAN));
|
1736
|
+
len--;
|
1737
|
+
s0 = s;
|
1738
|
+
}
|
1739
|
+
}
|
1740
|
+
}
|
1741
|
+
break;
|
1742
|
+
|
1743
|
+
default:
|
1744
|
+
rb_warning("unknown unpack directive '%c' in '%s'",
|
1745
|
+
type, RSTRING_PTR(fmt));
|
1746
|
+
break;
|
1747
|
+
}
|
1748
|
+
}
|
1749
|
+
|
1750
|
+
return ary;
|
1751
|
+
}
|
1752
|
+
|
1753
|
+
/*
|
1754
|
+
* call-seq:
|
1755
|
+
* str.unpack(format) -> anArray
|
1756
|
+
*
|
1757
|
+
* Decodes <i>str</i> (which may contain binary data) according to the
|
1758
|
+
* format string, returning an array of each value extracted. The
|
1759
|
+
* format string consists of a sequence of single-character directives,
|
1760
|
+
* summarized in the table at the end of this entry.
|
1761
|
+
* Each directive may be followed
|
1762
|
+
* by a number, indicating the number of times to repeat with this
|
1763
|
+
* directive. An asterisk (``<code>*</code>'') will use up all
|
1764
|
+
* remaining elements. The directives <code>sSiIlL</code> may each be
|
1765
|
+
* followed by an underscore (``<code>_</code>'') or
|
1766
|
+
* exclamation mark (``<code>!</code>'') to use the underlying
|
1767
|
+
* platform's native size for the specified type; otherwise, it uses a
|
1768
|
+
* platform-independent consistent size. Spaces are ignored in the
|
1769
|
+
* format string. See also <code>String#unpack1</code>, <code>Array#pack</code>.
|
1770
|
+
*
|
1771
|
+
* "abc \0\0abc \0\0".unpack('A6Z6') #=> ["abc", "abc "]
|
1772
|
+
* "abc \0\0".unpack('a3a3') #=> ["abc", " \000\000"]
|
1773
|
+
* "abc \0abc \0".unpack('Z*Z*') #=> ["abc ", "abc "]
|
1774
|
+
* "aa".unpack('b8B8') #=> ["10000110", "01100001"]
|
1775
|
+
* "aaa".unpack('h2H2c') #=> ["16", "61", 97]
|
1776
|
+
* "\xfe\xff\xfe\xff".unpack('sS') #=> [-2, 65534]
|
1777
|
+
* "now=20is".unpack('M*') #=> ["now is"]
|
1778
|
+
* "whole".unpack('xax2aX2aX1aX2a') #=> ["h", "e", "l", "l", "o"]
|
1779
|
+
*
|
1780
|
+
* This table summarizes the various formats and the Ruby classes
|
1781
|
+
* returned by each.
|
1782
|
+
*
|
1783
|
+
* Integer | |
|
1784
|
+
* Directive | Returns | Meaning
|
1785
|
+
* ------------------------------------------------------------------
|
1786
|
+
* C | Integer | 8-bit unsigned (unsigned char)
|
1787
|
+
* S | Integer | 16-bit unsigned, native endian (uint16_t)
|
1788
|
+
* L | Integer | 32-bit unsigned, native endian (uint32_t)
|
1789
|
+
* Q | Integer | 64-bit unsigned, native endian (uint64_t)
|
1790
|
+
* J | Integer | pointer width unsigned, native endian (uintptr_t)
|
1791
|
+
* | |
|
1792
|
+
* c | Integer | 8-bit signed (signed char)
|
1793
|
+
* s | Integer | 16-bit signed, native endian (int16_t)
|
1794
|
+
* l | Integer | 32-bit signed, native endian (int32_t)
|
1795
|
+
* q | Integer | 64-bit signed, native endian (int64_t)
|
1796
|
+
* j | Integer | pointer width signed, native endian (intptr_t)
|
1797
|
+
* | |
|
1798
|
+
* S_ S! | Integer | unsigned short, native endian
|
1799
|
+
* I I_ I! | Integer | unsigned int, native endian
|
1800
|
+
* L_ L! | Integer | unsigned long, native endian
|
1801
|
+
* Q_ Q! | Integer | unsigned long long, native endian (ArgumentError
|
1802
|
+
* | | if the platform has no long long type.)
|
1803
|
+
* J! | Integer | uintptr_t, native endian (same with J)
|
1804
|
+
* | |
|
1805
|
+
* s_ s! | Integer | signed short, native endian
|
1806
|
+
* i i_ i! | Integer | signed int, native endian
|
1807
|
+
* l_ l! | Integer | signed long, native endian
|
1808
|
+
* q_ q! | Integer | signed long long, native endian (ArgumentError
|
1809
|
+
* | | if the platform has no long long type.)
|
1810
|
+
* j! | Integer | intptr_t, native endian (same with j)
|
1811
|
+
* | |
|
1812
|
+
* S> s> S!> s!> | Integer | same as the directives without ">" except
|
1813
|
+
* L> l> L!> l!> | | big endian
|
1814
|
+
* I!> i!> | |
|
1815
|
+
* Q> q> Q!> q!> | | "S>" is same as "n"
|
1816
|
+
* J> j> J!> j!> | | "L>" is same as "N"
|
1817
|
+
* | |
|
1818
|
+
* S< s< S!< s!< | Integer | same as the directives without "<" except
|
1819
|
+
* L< l< L!< l!< | | little endian
|
1820
|
+
* I!< i!< | |
|
1821
|
+
* Q< q< Q!< q!< | | "S<" is same as "v"
|
1822
|
+
* J< j< J!< j!< | | "L<" is same as "V"
|
1823
|
+
* | |
|
1824
|
+
* n | Integer | 16-bit unsigned, network (big-endian) byte order
|
1825
|
+
* N | Integer | 32-bit unsigned, network (big-endian) byte order
|
1826
|
+
* v | Integer | 16-bit unsigned, VAX (little-endian) byte order
|
1827
|
+
* V | Integer | 32-bit unsigned, VAX (little-endian) byte order
|
1828
|
+
* | |
|
1829
|
+
* U | Integer | UTF-8 character
|
1830
|
+
* w | Integer | BER-compressed integer (see Array.pack)
|
1831
|
+
*
|
1832
|
+
* Float | |
|
1833
|
+
* Directive | Returns | Meaning
|
1834
|
+
* -----------------------------------------------------------------
|
1835
|
+
* D d | Float | double-precision, native format
|
1836
|
+
* F f | Float | single-precision, native format
|
1837
|
+
* E | Float | double-precision, little-endian byte order
|
1838
|
+
* e | Float | single-precision, little-endian byte order
|
1839
|
+
* G | Float | double-precision, network (big-endian) byte order
|
1840
|
+
* g | Float | single-precision, network (big-endian) byte order
|
1841
|
+
*
|
1842
|
+
* String | |
|
1843
|
+
* Directive | Returns | Meaning
|
1844
|
+
* -----------------------------------------------------------------
|
1845
|
+
* A | String | arbitrary binary string (remove trailing nulls and ASCII spaces)
|
1846
|
+
* a | String | arbitrary binary string
|
1847
|
+
* Z | String | null-terminated string
|
1848
|
+
* B | String | bit string (MSB first)
|
1849
|
+
* b | String | bit string (LSB first)
|
1850
|
+
* H | String | hex string (high nibble first)
|
1851
|
+
* h | String | hex string (low nibble first)
|
1852
|
+
* u | String | UU-encoded string
|
1853
|
+
* M | String | quoted-printable, MIME encoding (see RFC2045)
|
1854
|
+
* m | String | base64 encoded string (RFC 2045) (default)
|
1855
|
+
* | | base64 encoded string (RFC 4648) if followed by 0
|
1856
|
+
* P | String | pointer to a structure (fixed-length string)
|
1857
|
+
* p | String | pointer to a null-terminated string
|
1858
|
+
*
|
1859
|
+
* Misc. | |
|
1860
|
+
* Directive | Returns | Meaning
|
1861
|
+
* -----------------------------------------------------------------
|
1862
|
+
* @ | --- | skip to the offset given by the length argument
|
1863
|
+
* X | --- | skip backward one byte
|
1864
|
+
* x | --- | skip forward one byte
|
1865
|
+
*
|
1866
|
+
* HISTORY
|
1867
|
+
*
|
1868
|
+
* * J, J! j, and j! are available since Ruby 2.3.
|
1869
|
+
* * Q_, Q!, q_, and q! are available since Ruby 2.1.
|
1870
|
+
* * I!<, i!<, I!>, and i!> are available since Ruby 1.9.3.
|
1871
|
+
*/
|
1872
|
+
|
1873
|
+
static VALUE
|
1874
|
+
pack_unpack(VALUE str, VALUE fmt)
|
1875
|
+
{
|
1876
|
+
int mode = rb_block_given_p() ? UNPACK_BLOCK : UNPACK_ARRAY;
|
1877
|
+
return pack_unpack_internal(str, fmt, mode);
|
1878
|
+
}
|
1879
|
+
|
1880
|
+
/*
|
1881
|
+
* call-seq:
|
1882
|
+
* str.unpack1(format) -> obj
|
1883
|
+
*
|
1884
|
+
* Decodes <i>str</i> (which may contain binary data) according to the
|
1885
|
+
* format string, returning the first value extracted.
|
1886
|
+
* See also <code>String#unpack</code>, <code>Array#pack</code>.
|
1887
|
+
*/
|
1888
|
+
|
1889
|
+
static VALUE
|
1890
|
+
pack_unpack1(VALUE str, VALUE fmt)
|
1891
|
+
{
|
1892
|
+
return pack_unpack_internal(str, fmt, UNPACK_1);
|
1893
|
+
}
|
1894
|
+
|
1895
|
+
int
|
1896
|
+
rb_uv_to_utf8(char buf[6], unsigned long uv)
|
1897
|
+
{
|
1898
|
+
if (uv <= 0x7f) {
|
1899
|
+
buf[0] = (char)uv;
|
1900
|
+
return 1;
|
1901
|
+
}
|
1902
|
+
if (uv <= 0x7ff) {
|
1903
|
+
buf[0] = castchar(((uv>>6)&0xff)|0xc0);
|
1904
|
+
buf[1] = castchar((uv&0x3f)|0x80);
|
1905
|
+
return 2;
|
1906
|
+
}
|
1907
|
+
if (uv <= 0xffff) {
|
1908
|
+
buf[0] = castchar(((uv>>12)&0xff)|0xe0);
|
1909
|
+
buf[1] = castchar(((uv>>6)&0x3f)|0x80);
|
1910
|
+
buf[2] = castchar((uv&0x3f)|0x80);
|
1911
|
+
return 3;
|
1912
|
+
}
|
1913
|
+
if (uv <= 0x1fffff) {
|
1914
|
+
buf[0] = castchar(((uv>>18)&0xff)|0xf0);
|
1915
|
+
buf[1] = castchar(((uv>>12)&0x3f)|0x80);
|
1916
|
+
buf[2] = castchar(((uv>>6)&0x3f)|0x80);
|
1917
|
+
buf[3] = castchar((uv&0x3f)|0x80);
|
1918
|
+
return 4;
|
1919
|
+
}
|
1920
|
+
if (uv <= 0x3ffffff) {
|
1921
|
+
buf[0] = castchar(((uv>>24)&0xff)|0xf8);
|
1922
|
+
buf[1] = castchar(((uv>>18)&0x3f)|0x80);
|
1923
|
+
buf[2] = castchar(((uv>>12)&0x3f)|0x80);
|
1924
|
+
buf[3] = castchar(((uv>>6)&0x3f)|0x80);
|
1925
|
+
buf[4] = castchar((uv&0x3f)|0x80);
|
1926
|
+
return 5;
|
1927
|
+
}
|
1928
|
+
if (uv <= 0x7fffffff) {
|
1929
|
+
buf[0] = castchar(((uv>>30)&0xff)|0xfc);
|
1930
|
+
buf[1] = castchar(((uv>>24)&0x3f)|0x80);
|
1931
|
+
buf[2] = castchar(((uv>>18)&0x3f)|0x80);
|
1932
|
+
buf[3] = castchar(((uv>>12)&0x3f)|0x80);
|
1933
|
+
buf[4] = castchar(((uv>>6)&0x3f)|0x80);
|
1934
|
+
buf[5] = castchar((uv&0x3f)|0x80);
|
1935
|
+
return 6;
|
1936
|
+
}
|
1937
|
+
rb_raise(rb_eRangeError, "pack(U): value out of range");
|
1938
|
+
|
1939
|
+
UNREACHABLE;
|
1940
|
+
}
|
1941
|
+
|
1942
|
+
static const unsigned long utf8_limits[] = {
|
1943
|
+
0x0, /* 1 */
|
1944
|
+
0x80, /* 2 */
|
1945
|
+
0x800, /* 3 */
|
1946
|
+
0x10000, /* 4 */
|
1947
|
+
0x200000, /* 5 */
|
1948
|
+
0x4000000, /* 6 */
|
1949
|
+
0x80000000, /* 7 */
|
1950
|
+
};
|
1951
|
+
|
1952
|
+
static unsigned long
|
1953
|
+
utf8_to_uv(const char *p, long *lenp)
|
1954
|
+
{
|
1955
|
+
int c = *p++ & 0xff;
|
1956
|
+
unsigned long uv = c;
|
1957
|
+
long n;
|
1958
|
+
|
1959
|
+
if (!(uv & 0x80)) {
|
1960
|
+
*lenp = 1;
|
1961
|
+
return uv;
|
1962
|
+
}
|
1963
|
+
if (!(uv & 0x40)) {
|
1964
|
+
*lenp = 1;
|
1965
|
+
rb_raise(rb_eArgError, "malformed UTF-8 character");
|
1966
|
+
}
|
1967
|
+
|
1968
|
+
if (!(uv & 0x20)) { n = 2; uv &= 0x1f; }
|
1969
|
+
else if (!(uv & 0x10)) { n = 3; uv &= 0x0f; }
|
1970
|
+
else if (!(uv & 0x08)) { n = 4; uv &= 0x07; }
|
1971
|
+
else if (!(uv & 0x04)) { n = 5; uv &= 0x03; }
|
1972
|
+
else if (!(uv & 0x02)) { n = 6; uv &= 0x01; }
|
1973
|
+
else {
|
1974
|
+
*lenp = 1;
|
1975
|
+
rb_raise(rb_eArgError, "malformed UTF-8 character");
|
1976
|
+
}
|
1977
|
+
if (n > *lenp) {
|
1978
|
+
rb_raise(rb_eArgError, "malformed UTF-8 character (expected %ld bytes, given %ld bytes)",
|
1979
|
+
n, *lenp);
|
1980
|
+
}
|
1981
|
+
*lenp = n--;
|
1982
|
+
if (n != 0) {
|
1983
|
+
while (n--) {
|
1984
|
+
c = *p++ & 0xff;
|
1985
|
+
if ((c & 0xc0) != 0x80) {
|
1986
|
+
*lenp -= n + 1;
|
1987
|
+
rb_raise(rb_eArgError, "malformed UTF-8 character");
|
1988
|
+
}
|
1989
|
+
else {
|
1990
|
+
c &= 0x3f;
|
1991
|
+
uv = uv << 6 | c;
|
1992
|
+
}
|
1993
|
+
}
|
1994
|
+
}
|
1995
|
+
n = *lenp - 1;
|
1996
|
+
if (uv < utf8_limits[n]) {
|
1997
|
+
rb_raise(rb_eArgError, "redundant UTF-8 sequence");
|
1998
|
+
}
|
1999
|
+
return uv;
|
2000
|
+
}
|
2001
|
+
|
2002
|
+
void
|
2003
|
+
Init_pack(void)
|
2004
|
+
{
|
2005
|
+
rb_define_method(rb_cArray, "pack", pack_pack, -1);
|
2006
|
+
rb_define_method(rb_cString, "unpack", pack_unpack, 1);
|
2007
|
+
rb_define_method(rb_cString, "unpack1", pack_unpack1, 1);
|
2008
|
+
|
2009
|
+
id_associated = rb_make_internal_id();
|
2010
|
+
}
|