1 /*
2 * Copyright (c) 2010, 2017 Craig A. Berry
3 *
4 * Permission is hereby granted, free of charge, to any person obtaining a copy
5 * of this software and associated documentation files (the "Software"), to deal
6 * in the Software without restriction, including without limitation the rights
7 * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
8 * copies of the Software, and to permit persons to whom the Software is
9 * furnished to do so, subject to the following conditions:
10 * The above copyright notice and this permission notice shall be included in
11 * all copies or substantial portions of the Software.
12 *
13 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
14 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
15 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
16 * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
17 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
18 * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
19 * THE SOFTWARE.
20 */
21
22 /* vms_shorten_symbol
23 *
24 * This program provides shortening of long symbols (> 31 characters) using the
25 * same mechanism as the OpenVMS C compiler. The basic procedure is to compute
26 * an AUTODIN II checksum of the entire symbol, encode the checksum in base32,
27 * and glue together a shortened symbol from the first 23 characters of the
28 * original symbol plus the encoded checksum appended. The output format is
29 * the same used in the name mangler database, stored by default in
30 * [.CXX_REPOSITORY]CXX$DEMANGLER_DB.
31 *
32 * To obtain the same result as CC/NAMES=SHORTENED, run like so:
33 *
34 * $ mcr []vms_shorten_symbol "Please_forgive_this_absurdly_long_symbol_name"
35 * PLEASE_FORGIVE_THIS_ABS1ARO4QU$Please_forgive_this_absurdly_long_symbol_name
36 *
37 * To obtain the same result as CC/NAMES=(SHORTENED,AS_IS), pass a non-zero
38 * value as the second argument, like so:
39 *
40 * $ mcr []vms_shorten_symbol "Please_forgive_this_absurdly_long_symbol_name" 1
41 * Please_forgive_this_abs3rv8rnn$Please_forgive_this_absurdly_long_symbol_name
42 */
43
44 #include <ctype.h>
45 #include <stdio.h>
46 #include <stdlib.h>
47 #include <string.h>
48
49 #ifdef __VMS
50 #define UINT32 unsigned int
51 #else
52 #include <inttypes.h>
53 #define UINT32 uint32_t
54 #endif
55
56 extern UINT32 crc32(const char *input_string);
57 extern int u32_to_base32(UINT32 input, char *output);
58 extern int vms_shorten_symbol(const char *symbol, char *shortened, char as_is_flag);
59
60 /*
61 * This routine implements the AUTODIN II polynomial.
62 */
63
64 UINT32
crc32(const char * input_string)65 crc32(const char *input_string)
66 {
67
68 /*
69 * CRC code and data based partly on FreeBSD implementation, which
70 * notes:
71 *
72 * The crc32 functions and data was originally written by Spencer
73 * Garrett <[email protected]> and was cleaned from the PostgreSQL source
74 * tree via the files contrib/ltree/crc32.[ch]. No license was
75 * included, therefore it is assumed that this code is public
76 * domain. Attribution still noted.
77 *
78 * (I think they mean "gleaned" not "cleaned".)
79 */
80
81 static const UINT32 autodin_ii_table[256] = {
82 0x00000000, 0x77073096, 0xee0e612c, 0x990951ba, 0x076dc419, 0x706af48f,
83 0xe963a535, 0x9e6495a3, 0x0edb8832, 0x79dcb8a4, 0xe0d5e91e, 0x97d2d988,
84 0x09b64c2b, 0x7eb17cbd, 0xe7b82d07, 0x90bf1d91, 0x1db71064, 0x6ab020f2,
85 0xf3b97148, 0x84be41de, 0x1adad47d, 0x6ddde4eb, 0xf4d4b551, 0x83d385c7,
86 0x136c9856, 0x646ba8c0, 0xfd62f97a, 0x8a65c9ec, 0x14015c4f, 0x63066cd9,
87 0xfa0f3d63, 0x8d080df5, 0x3b6e20c8, 0x4c69105e, 0xd56041e4, 0xa2677172,
88 0x3c03e4d1, 0x4b04d447, 0xd20d85fd, 0xa50ab56b, 0x35b5a8fa, 0x42b2986c,
89 0xdbbbc9d6, 0xacbcf940, 0x32d86ce3, 0x45df5c75, 0xdcd60dcf, 0xabd13d59,
90 0x26d930ac, 0x51de003a, 0xc8d75180, 0xbfd06116, 0x21b4f4b5, 0x56b3c423,
91 0xcfba9599, 0xb8bda50f, 0x2802b89e, 0x5f058808, 0xc60cd9b2, 0xb10be924,
92 0x2f6f7c87, 0x58684c11, 0xc1611dab, 0xb6662d3d, 0x76dc4190, 0x01db7106,
93 0x98d220bc, 0xefd5102a, 0x71b18589, 0x06b6b51f, 0x9fbfe4a5, 0xe8b8d433,
94 0x7807c9a2, 0x0f00f934, 0x9609a88e, 0xe10e9818, 0x7f6a0dbb, 0x086d3d2d,
95 0x91646c97, 0xe6635c01, 0x6b6b51f4, 0x1c6c6162, 0x856530d8, 0xf262004e,
96 0x6c0695ed, 0x1b01a57b, 0x8208f4c1, 0xf50fc457, 0x65b0d9c6, 0x12b7e950,
97 0x8bbeb8ea, 0xfcb9887c, 0x62dd1ddf, 0x15da2d49, 0x8cd37cf3, 0xfbd44c65,
98 0x4db26158, 0x3ab551ce, 0xa3bc0074, 0xd4bb30e2, 0x4adfa541, 0x3dd895d7,
99 0xa4d1c46d, 0xd3d6f4fb, 0x4369e96a, 0x346ed9fc, 0xad678846, 0xda60b8d0,
100 0x44042d73, 0x33031de5, 0xaa0a4c5f, 0xdd0d7cc9, 0x5005713c, 0x270241aa,
101 0xbe0b1010, 0xc90c2086, 0x5768b525, 0x206f85b3, 0xb966d409, 0xce61e49f,
102 0x5edef90e, 0x29d9c998, 0xb0d09822, 0xc7d7a8b4, 0x59b33d17, 0x2eb40d81,
103 0xb7bd5c3b, 0xc0ba6cad, 0xedb88320, 0x9abfb3b6, 0x03b6e20c, 0x74b1d29a,
104 0xead54739, 0x9dd277af, 0x04db2615, 0x73dc1683, 0xe3630b12, 0x94643b84,
105 0x0d6d6a3e, 0x7a6a5aa8, 0xe40ecf0b, 0x9309ff9d, 0x0a00ae27, 0x7d079eb1,
106 0xf00f9344, 0x8708a3d2, 0x1e01f268, 0x6906c2fe, 0xf762575d, 0x806567cb,
107 0x196c3671, 0x6e6b06e7, 0xfed41b76, 0x89d32be0, 0x10da7a5a, 0x67dd4acc,
108 0xf9b9df6f, 0x8ebeeff9, 0x17b7be43, 0x60b08ed5, 0xd6d6a3e8, 0xa1d1937e,
109 0x38d8c2c4, 0x4fdff252, 0xd1bb67f1, 0xa6bc5767, 0x3fb506dd, 0x48b2364b,
110 0xd80d2bda, 0xaf0a1b4c, 0x36034af6, 0x41047a60, 0xdf60efc3, 0xa867df55,
111 0x316e8eef, 0x4669be79, 0xcb61b38c, 0xbc66831a, 0x256fd2a0, 0x5268e236,
112 0xcc0c7795, 0xbb0b4703, 0x220216b9, 0x5505262f, 0xc5ba3bbe, 0xb2bd0b28,
113 0x2bb45a92, 0x5cb36a04, 0xc2d7ffa7, 0xb5d0cf31, 0x2cd99e8b, 0x5bdeae1d,
114 0x9b64c2b0, 0xec63f226, 0x756aa39c, 0x026d930a, 0x9c0906a9, 0xeb0e363f,
115 0x72076785, 0x05005713, 0x95bf4a82, 0xe2b87a14, 0x7bb12bae, 0x0cb61b38,
116 0x92d28e9b, 0xe5d5be0d, 0x7cdcefb7, 0x0bdbdf21, 0x86d3d2d4, 0xf1d4e242,
117 0x68ddb3f8, 0x1fda836e, 0x81be16cd, 0xf6b9265b, 0x6fb077e1, 0x18b74777,
118 0x88085ae6, 0xff0f6a70, 0x66063bca, 0x11010b5c, 0x8f659eff, 0xf862ae69,
119 0x616bffd3, 0x166ccf45, 0xa00ae278, 0xd70dd2ee, 0x4e048354, 0x3903b3c2,
120 0xa7672661, 0xd06016f7, 0x4969474d, 0x3e6e77db, 0xaed16a4a, 0xd9d65adc,
121 0x40df0b66, 0x37d83bf0, 0xa9bcae53, 0xdebb9ec5, 0x47b2cf7f, 0x30b5ffe9,
122 0xbdbdf21c, 0xcabac28a, 0x53b39330, 0x24b4a3a6, 0xbad03605, 0xcdd70693,
123 0x54de5729, 0x23d967bf, 0xb3667a2e, 0xc4614ab8, 0x5d681b02, 0x2a6f2b94,
124 0xb40bbe37, 0xc30c8ea1, 0x5a05df1b, 0x2d02ef8d,
125 };
126
127 UINT32 crc = ~0U;
128 char *c;
129 for (c = (char *)input_string; *c; ++c)
130 crc = (crc >> 8) ^ autodin_ii_table[(crc ^ *c) & 0xff];
131 return ~crc;
132 }
133
134 /*
135 * This is the RFC2938 variant of base32, not RFC3548, Crockford's, or
136 * other newer variant. It produces an 8-byte encoded character string
137 * (plus trailing null) from a 32-bit integer input.
138 */
139
140 int
u32_to_base32(UINT32 input,char * output)141 u32_to_base32(UINT32 input, char *output)
142 {
143 static const char base32hex_table[32] = {
144 '0', '1', '2', '3', '4', '5', '6', '7', '8', '9',
145 'a', 'b', 'c', 'd', 'e', 'f', 'g', 'h', 'i', 'j',
146 'k', 'l', 'm', 'n', 'o', 'p', 'q', 'r', 's', 't',
147 'u', 'v'
148 };
149 int i;
150
151 /*
152 * Grab lowest 5 bits and look up conversion in table. Lather, rinse,
153 * repeat for a total of 7, 5-bit chunks to accommodate 32 bits of input.
154 */
155 for (i = 0; i < 7; i++) {
156 output[6 - i] = base32hex_table[input & 0x1f];
157 input >>= 5; /* position to look at next 5 */
158 }
159 output[7] = '$'; /* It's DEC, so use '$' not '=' to pad. */
160 output[8] = '\0';
161 return 0;
162 }
163
164 /*
165 * Take an input symbol name of arbitrary length and produce a symbol shortened
166 * to 31 characters. The shortened symbol consists of the first 23 characters
167 * of the original symbol plus the 8 characters of the encoded checksum. The
168 * third argument is a boolean indicating whether to emulate the compiler's
169 * /NAMES=AS_IS option. When false (the compiler's default), the shortened
170 * symbol will be upper cased. When the original symbol is 31 characters or
171 * fewer in length, no checksum will be appended and the original symbol is
172 * returned verbatim (though upper cased if the as_is_flag is false).
173 */
174
175 int
vms_shorten_symbol(const char * input_symbol,char * shortened,char as_is_flag)176 vms_shorten_symbol(const char *input_symbol, char *shortened, char as_is_flag)
177 {
178 char b32str[9];
179 UINT32 crc;
180 char *c, *symbol;
181 int symlen;
182
183 symlen = strlen(input_symbol);
184 symbol = (char *)malloc(symlen + 1);
185 if (symbol == NULL)
186 return -1;
187
188 strncpy(symbol, input_symbol, symlen);
189 symbol[symlen] = '\0';
190
191 if (!as_is_flag) {
192 for (c = symbol; *c; c++)
193 *c = toupper(*c);
194 }
195
196 if (symlen <= 31) {
197 strncpy(shortened, symbol, symlen);
198 shortened[symlen] = '\0';
199 free(symbol);
200 return 0;
201 }
202
203 /*
204 * Compute the checksum on the whole symbol.
205 */
206
207 crc = crc32(symbol);
208
209 /* The compiler does not use the inverted checksum, so we invert it
210 * back before encoding in base32.
211 */
212
213 if (u32_to_base32(~crc, (char *)&b32str) == -1) {
214 free(symbol);
215 return -1;
216 }
217
218 if (!as_is_flag) {
219 for (c = (char *)&b32str; *c; c++)
220 *c = toupper(*c);
221 }
222
223 sprintf(shortened, "%.23s%.8s", symbol, b32str);
224 shortened[31] = '\0';
225 free(symbol);
226 return 0;
227 }
228
229 #ifdef TEST_MAIN
230 int
main(int argc,char ** argv)231 main(int argc, char **argv)
232 {
233 char short_symbol[32];
234 char as_is_flag = 0;
235
236 if (argc < 2) {
237 fprintf(stderr, "Usage: %s <symbol name> [<AS_IS flag>]\n", argv[0]);
238 exit(EXIT_FAILURE);
239 }
240 if (argc > 2)
241 as_is_flag = 1;
242
243 if (vms_shorten_symbol(argv[1], (char *)&short_symbol, as_is_flag) == -1) {
244 fprintf(stderr, "Symbol shortening failed\n");
245 exit(EXIT_FAILURE);
246 }
247
248 printf("%s%s\n", (char *)&short_symbol, argv[1]);
249 }
250 #endif
251