1 /*- 2 * Copyright (c) 1991, 1993 3 * The Regents of the University of California. All rights reserved. 4 * 5 * This code is derived from software contributed to Berkeley by 6 * Kenneth Almquist. 7 * 8 * Redistribution and use in source and binary forms, with or without 9 * modification, are permitted provided that the following conditions 10 * are met: 11 * 1. Redistributions of source code must retain the above copyright 12 * notice, this list of conditions and the following disclaimer. 13 * 2. Redistributions in binary form must reproduce the above copyright 14 * notice, this list of conditions and the following disclaimer in the 15 * documentation and/or other materials provided with the distribution. 16 * 3. All advertising materials mentioning features or use of this software 17 * must display the following acknowledgement: 18 * This product includes software developed by the University of 19 * California, Berkeley and its contributors. 20 * 4. Neither the name of the University nor the names of its contributors 21 * may be used to endorse or promote products derived from this software 22 * without specific prior written permission. 23 * 24 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 25 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 26 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 27 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 28 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 29 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 30 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 31 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 32 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 33 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 34 * SUCH DAMAGE. 35 * 36 * @(#) Copyright (c) 1991, 1993 The Regents of the University of California. All rights reserved. 37 * @(#)mksyntax.c 8.2 (Berkeley) 5/4/95 38 * $FreeBSD: src/bin/sh/mksyntax.c,v 1.35 2011/05/05 20:55:55 jilles Exp $ 39 */ 40 41 /* 42 * This program creates syntax.h and syntax.c. 43 */ 44 45 #include <stdio.h> 46 #include <stdlib.h> 47 #include <string.h> 48 #include "parser.h" 49 50 51 struct synclass { 52 const char *name; 53 const char *comment; 54 }; 55 56 /* Syntax classes */ 57 struct synclass synclass[] = { 58 { "CWORD", "character is nothing special" }, 59 { "CNL", "newline character" }, 60 { "CBACK", "a backslash character" }, 61 { "CSBACK", "a backslash character in single quotes" }, 62 { "CSQUOTE", "single quote" }, 63 { "CDQUOTE", "double quote" }, 64 { "CENDQUOTE", "a terminating quote" }, 65 { "CBQUOTE", "backwards single quote" }, 66 { "CVAR", "a dollar sign" }, 67 { "CENDVAR", "a '}' character" }, 68 { "CLP", "a left paren in arithmetic" }, 69 { "CRP", "a right paren in arithmetic" }, 70 { "CEOF", "end of file" }, 71 { "CCTL", "like CWORD, except it must be escaped" }, 72 { "CSPCL", "these terminate a word" }, 73 { "CIGN", "character should be ignored" }, 74 { NULL, NULL } 75 }; 76 77 78 /* 79 * Syntax classes for is_ functions. Warning: if you add new classes 80 * you may have to change the definition of the is_in_name macro. 81 */ 82 struct synclass is_entry[] = { 83 { "ISDIGIT", "a digit" }, 84 { "ISUPPER", "an upper case letter" }, 85 { "ISLOWER", "a lower case letter" }, 86 { "ISUNDER", "an underscore" }, 87 { "ISSPECL", "the name of a special parameter" }, 88 { NULL, NULL } 89 }; 90 91 static char writer[] = "\ 92 /*\n\ 93 * This file was generated by the mksyntax program.\n\ 94 */\n\ 95 \n"; 96 97 98 static FILE *cfile; 99 static FILE *hfile; 100 static const char *syntax[513]; 101 static int base; 102 static int size; /* number of values which a char variable can have */ 103 static int nbits; /* number of bits in a character */ 104 static int digit_contig;/* true if digits are contiguous */ 105 106 static void filltable(const char *); 107 static void init(void); 108 static void add(const char *, const char *); 109 static void print(const char *); 110 static void output_type_macros(void); 111 static void digit_convert(void); 112 113 int 114 main(int argc __unused, char **argv __unused) 115 { 116 char c; 117 char d; 118 int sign; 119 int i; 120 char buf[80]; 121 int pos; 122 static char digit[] = "0123456789"; 123 124 /* Create output files */ 125 if ((cfile = fopen("syntax.c", "w")) == NULL) { 126 perror("syntax.c"); 127 exit(2); 128 } 129 if ((hfile = fopen("syntax.h", "w")) == NULL) { 130 perror("syntax.h"); 131 exit(2); 132 } 133 fputs(writer, hfile); 134 fputs(writer, cfile); 135 136 /* Determine the characteristics of chars. */ 137 c = -1; 138 sign = (c > 0) ? 0 : 1; 139 for (nbits = 1 ; ; nbits++) { 140 d = (1 << nbits) - 1; 141 if (d == c) 142 break; 143 } 144 #if 0 145 printf("%s %d bit chars\n", sign? "signed" : "unsigned", nbits); 146 #endif 147 if (nbits > 9) { 148 fputs("Characters can't have more than 9 bits\n", stderr); 149 exit(2); 150 } 151 size = (1 << nbits) + 1; 152 base = 1; 153 if (sign) 154 base += 1 << (nbits - 1); 155 digit_contig = 1; 156 for (i = 0 ; i < 10 ; i++) { 157 if (digit[i] != '0' + i) 158 digit_contig = 0; 159 } 160 161 fputs("#include <sys/cdefs.h>\n", hfile); 162 fputs("#include <ctype.h>\n", hfile); 163 164 /* Generate the #define statements in the header file */ 165 fputs("/* Syntax classes */\n", hfile); 166 for (i = 0 ; synclass[i].name ; i++) { 167 sprintf(buf, "#define %s %d", synclass[i].name, i); 168 fputs(buf, hfile); 169 for (pos = strlen(buf) ; pos < 32 ; pos = (pos + 8) & ~07) 170 putc('\t', hfile); 171 fprintf(hfile, "/* %s */\n", synclass[i].comment); 172 } 173 putc('\n', hfile); 174 fputs("/* Syntax classes for is_ functions */\n", hfile); 175 for (i = 0 ; is_entry[i].name ; i++) { 176 sprintf(buf, "#define %s %#o", is_entry[i].name, 1 << i); 177 fputs(buf, hfile); 178 for (pos = strlen(buf) ; pos < 32 ; pos = (pos + 8) & ~07) 179 putc('\t', hfile); 180 fprintf(hfile, "/* %s */\n", is_entry[i].comment); 181 } 182 putc('\n', hfile); 183 fprintf(hfile, "#define SYNBASE %d\n", base); 184 fprintf(hfile, "#define PEOF %d\n\n", -base); 185 putc('\n', hfile); 186 fputs("#define BASESYNTAX (basesyntax + SYNBASE)\n", hfile); 187 fputs("#define DQSYNTAX (dqsyntax + SYNBASE)\n", hfile); 188 fputs("#define SQSYNTAX (sqsyntax + SYNBASE)\n", hfile); 189 fputs("#define ARISYNTAX (arisyntax + SYNBASE)\n", hfile); 190 putc('\n', hfile); 191 output_type_macros(); /* is_digit, etc. */ 192 putc('\n', hfile); 193 194 /* Generate the syntax tables. */ 195 fputs("#include \"shell.h\"\n", cfile); 196 fputs("#include \"syntax.h\"\n\n", cfile); 197 init(); 198 fputs("/* syntax table used when not in quotes */\n", cfile); 199 add("\n", "CNL"); 200 add("\\", "CBACK"); 201 add("'", "CSQUOTE"); 202 add("\"", "CDQUOTE"); 203 add("`", "CBQUOTE"); 204 add("$", "CVAR"); 205 add("}", "CENDVAR"); 206 add("<>();&| \t", "CSPCL"); 207 print("basesyntax"); 208 init(); 209 fputs("\n/* syntax table used when in double quotes */\n", cfile); 210 add("\n", "CNL"); 211 add("\\", "CBACK"); 212 add("\"", "CENDQUOTE"); 213 add("`", "CBQUOTE"); 214 add("$", "CVAR"); 215 add("}", "CENDVAR"); 216 /* ':/' for tilde expansion, '-' for [a\-x] pattern ranges */ 217 add("!*?[=~:/-", "CCTL"); 218 print("dqsyntax"); 219 init(); 220 fputs("\n/* syntax table used when in single quotes */\n", cfile); 221 add("\n", "CNL"); 222 add("\\", "CSBACK"); 223 add("'", "CENDQUOTE"); 224 /* ':/' for tilde expansion, '-' for [a\-x] pattern ranges */ 225 add("!*?[=~:/-", "CCTL"); 226 print("sqsyntax"); 227 init(); 228 fputs("\n/* syntax table used when in arithmetic */\n", cfile); 229 add("\n", "CNL"); 230 add("\\", "CBACK"); 231 add("`", "CBQUOTE"); 232 add("\"", "CIGN"); 233 add("$", "CVAR"); 234 add("}", "CENDVAR"); 235 add("(", "CLP"); 236 add(")", "CRP"); 237 print("arisyntax"); 238 filltable("0"); 239 fputs("\n/* character classification table */\n", cfile); 240 add("0123456789", "ISDIGIT"); 241 add("abcdefghijklmnopqrstuvwxyz", "ISLOWER"); 242 add("ABCDEFGHIJKLMNOPQRSTUVWXYZ", "ISUPPER"); 243 add("_", "ISUNDER"); 244 add("#?$!-*@", "ISSPECL"); 245 print("is_type"); 246 if (! digit_contig) 247 digit_convert(); 248 exit(0); 249 } 250 251 252 253 /* 254 * Clear the syntax table. 255 */ 256 257 static void 258 filltable(const char *dftval) 259 { 260 int i; 261 262 for (i = 0 ; i < size ; i++) 263 syntax[i] = dftval; 264 } 265 266 267 /* 268 * Initialize the syntax table with default values. 269 */ 270 271 static void 272 init(void) 273 { 274 filltable("CWORD"); 275 syntax[0] = "CEOF"; 276 syntax[base + CTLESC] = "CCTL"; 277 syntax[base + CTLVAR] = "CCTL"; 278 syntax[base + CTLENDVAR] = "CCTL"; 279 syntax[base + CTLBACKQ] = "CCTL"; 280 syntax[base + CTLBACKQ + CTLQUOTE] = "CCTL"; 281 syntax[base + CTLARI] = "CCTL"; 282 syntax[base + CTLENDARI] = "CCTL"; 283 syntax[base + CTLQUOTEMARK] = "CCTL"; 284 syntax[base + CTLQUOTEEND] = "CCTL"; 285 } 286 287 288 /* 289 * Add entries to the syntax table. 290 */ 291 292 static void 293 add(const char *p, const char *type) 294 { 295 while (*p) 296 syntax[*p++ + base] = type; 297 } 298 299 300 301 /* 302 * Output the syntax table. 303 */ 304 305 static void 306 print(const char *name) 307 { 308 int i; 309 int col; 310 311 fprintf(hfile, "extern const char %s[];\n", name); 312 fprintf(cfile, "const char %s[%d] = {\n", name, size); 313 col = 0; 314 for (i = 0 ; i < size ; i++) { 315 if (i == 0) { 316 fputs(" ", cfile); 317 } else if ((i & 03) == 0) { 318 fputs(",\n ", cfile); 319 col = 0; 320 } else { 321 putc(',', cfile); 322 while (++col < 9 * (i & 03)) 323 putc(' ', cfile); 324 } 325 fputs(syntax[i], cfile); 326 col += strlen(syntax[i]); 327 } 328 fputs("\n};\n", cfile); 329 } 330 331 332 333 /* 334 * Output character classification macros (e.g. is_digit). If digits are 335 * contiguous, we can test for them quickly. 336 */ 337 338 static const char *macro[] = { 339 "#define is_digit(c)\t((is_type+SYNBASE)[(int)c] & ISDIGIT)", 340 "#define is_eof(c)\t((c) == PEOF)", 341 "#define is_alpha(c)\t((is_type+SYNBASE)[(int)c] & (ISUPPER|ISLOWER))", 342 "#define is_name(c)\t((is_type+SYNBASE)[(int)c] & (ISUPPER|ISLOWER|ISUNDER))", 343 "#define is_in_name(c)\t((is_type+SYNBASE)[(int)c] & (ISUPPER|ISLOWER|ISUNDER|ISDIGIT))", 344 "#define is_special(c)\t((is_type+SYNBASE)[(int)c] & (ISSPECL|ISDIGIT))", 345 NULL 346 }; 347 348 static void 349 output_type_macros(void) 350 { 351 const char **pp; 352 353 if (digit_contig) 354 macro[0] = "#define is_digit(c)\t((unsigned int)((c) - '0') <= 9)"; 355 for (pp = macro ; *pp ; pp++) 356 fprintf(hfile, "%s\n", *pp); 357 if (digit_contig) 358 fputs("#define digit_val(c)\t((c) - '0')\n", hfile); 359 else 360 fputs("#define digit_val(c)\t(digit_value[c])\n", hfile); 361 } 362 363 364 365 /* 366 * Output digit conversion table (if digits are not contiguous). 367 */ 368 369 static void 370 digit_convert(void) 371 { 372 int maxdigit; 373 static char digit[] = "0123456789"; 374 char *p; 375 int i; 376 377 maxdigit = 0; 378 for (p = digit ; *p ; p++) 379 if (*p > maxdigit) 380 maxdigit = *p; 381 fputs("extern const char digit_value[];\n", hfile); 382 fputs("\n\nconst char digit_value[] = {\n", cfile); 383 for (i = 0 ; i <= maxdigit ; i++) { 384 for (p = digit ; *p && *p != i ; p++); 385 if (*p == '\0') 386 p = digit; 387 fprintf(cfile, " %td,\n", p - digit); 388 } 389 fputs("};\n", cfile); 390 } 391