commands/ash/mksyntax.c

   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  * 4. Neither the name of the University nor the names of its contributors
  17  *    may be used to endorse or promote products derived from this software
  18  *    without specific prior written permission.
  19  *
  20  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
  21  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
  22  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
  23  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
  24  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
  25  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
  26  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
  27  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
  28  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
  29  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
  30  * SUCH DAMAGE.
  31  */
  32
  33 #if 0
  34 #ifndef lint
  35 static char const copyright[] =
  36 "@(#) Copyright (c) 1991, 1993\n\
  37         The Regents of the University of California.  All rights reserved.\n";
  38 #endif /* not lint */
  39
  40 #ifndef lint
  41 static char sccsid[] = "@(#)mksyntax.c  8.2 (Berkeley) 5/4/95";
  42 #endif /* not lint */
  43 #endif
  44 #include <sys/cdefs.h>
  45 /*
  46 __FBSDID("$FreeBSD: src/bin/sh/mksyntax.c,v 1.23 2004/04/06 20:06:51 markm Exp $");
  47 */
  48
  49 /*
  50  * This program creates syntax.h and syntax.c.
  51  */
  52
  53 #include <stdio.h>
  54 #include <stdlib.h>
  55 #include <string.h>
  56 #include "parser.h"
  57
  58 struct synclass {
  59         char *name;
  60         char *comment;
  61 };
  62
  63 /* Syntax classes */
  64 struct synclass synclass[] = {
  65         { "CWORD",      "character is nothing special" },
  66         { "CNL",        "newline character" },
  67         { "CBACK",      "a backslash character" },
  68         { "CSQUOTE",    "single quote" },
  69         { "CDQUOTE",    "double quote" },
  70         { "CENDQUOTE",  "a terminating quote" },
  71         { "CBQUOTE",    "backwards single quote" },
  72         { "CVAR",       "a dollar sign" },
  73         { "CENDVAR",    "a '}' character" },
  74         { "CLP",        "a left paren in arithmetic" },
  75         { "CRP",        "a right paren in arithmetic" },
  76         { "CEOF",       "end of file" },
  77         { "CCTL",       "like CWORD, except it must be escaped" },
  78         { "CSPCL",      "these terminate a word" },
  79         { NULL,         NULL }
  80 };
  81
  82
  83 /*
  84  * Syntax classes for is_ functions.  Warning:  if you add new classes
  85  * you may have to change the definition of the is_in_name macro.
  86  */
  87 struct synclass is_entry[] = {
  88         { "ISDIGIT",    "a digit" },
  89         { "ISUPPER",    "an upper case letter" },
  90         { "ISLOWER",    "a lower case letter" },
  91         { "ISUNDER",    "an underscore" },
  92         { "ISSPECL",    "the name of a special parameter" },
  93         { NULL,         NULL }
  94 };
  95
  96 static char writer[] = "\
  97 /*\n\
  98  * This file was generated by the mksyntax program.\n\
  99  */\n\
 100 \n";
 101
 102
 103 static FILE *cfile;
 104 static FILE *hfile;
 105 static char *syntax[513];
 106 static int base;
 107 static int size;        /* number of values which a char variable can have */
 108 static int nbits;       /* number of bits in a character */
 109 static int digit_contig;/* true if digits are contiguous */
 110
 111 static void filltable(char *);
 112 static void init(void);
 113 static void add(char *, char *);
 114 static void print(char *);
 115 static void output_type_macros(void);
 116 static void digit_convert(void);
 117
 118 #ifndef __unused
 119 #define __unused  __attribute__((__unused__))
 120 #endif
 121
 122 int
 123 main(int argc __unused, char **argv __unused)
 124 {
 125         char c;
 126         char d;
 127         int sign;
 128         int i;
 129         char buf[80];
 130         int pos;
 131         static char digit[] = "0123456789";
 132
 133         /* Create output files */
 134         if ((cfile = fopen("syntax.c", "w")) == NULL) {
 135                 perror("syntax.c");
 136                 exit(2);
 137         }
 138         if ((hfile = fopen("syntax.h", "w")) == NULL) {
 139                 perror("syntax.h");
 140                 exit(2);
 141         }
 142         fputs(writer, hfile);
 143         fputs(writer, cfile);
 144
 145         /* Determine the characteristics of chars. */
 146         c = -1;
 147         if (c < 0)
 148                 sign = 1;
 149         else
 150                 sign = 0;
 151         for (nbits = 1 ; ; nbits++) {
 152                 d = (1 << nbits) - 1;
 153                 if (d == c)
 154                         break;
 155         }
 156 #if 0
 157         printf("%s %d bit chars\n", sign? "signed" : "unsigned", nbits);
 158 #endif
 159         if (nbits > 9) {
 160                 fputs("Characters can't have more than 9 bits\n", stderr);
 161                 exit(2);
 162         }
 163         size = (1 << nbits) + 1;
 164         base = 1;
 165         if (sign)
 166                 base += 1 << (nbits - 1);
 167         digit_contig = 1;
 168         for (i = 0 ; i < 10 ; i++) {
 169                 if (digit[i] != '0' + i)
 170                         digit_contig = 0;
 171         }
 172
 173         fputs("#include <ctype.h>\n", hfile);
 174
 175         /* Generate the #define statements in the header file */
 176         fputs("/* Syntax classes */\n", hfile);
 177         for (i = 0 ; synclass[i].name ; i++) {
 178                 sprintf(buf, "#define %s %d", synclass[i].name, i);
 179                 fputs(buf, hfile);
 180                 for (pos = strlen(buf) ; pos < 32 ; pos = (pos + 8) & ~07)
 181                         putc('\t', hfile);
 182                 fprintf(hfile, "/* %s */\n", synclass[i].comment);
 183         }
 184         putc('\n', hfile);
 185         fputs("/* Syntax classes for is_ functions */\n", hfile);
 186         for (i = 0 ; is_entry[i].name ; i++) {
 187                 sprintf(buf, "#define %s %#o", is_entry[i].name, 1 << i);
 188                 fputs(buf, hfile);
 189                 for (pos = strlen(buf) ; pos < 32 ; pos = (pos + 8) & ~07)
 190                         putc('\t', hfile);
 191                 fprintf(hfile, "/* %s */\n", is_entry[i].comment);
 192         }
 193         putc('\n', hfile);
 194         fprintf(hfile, "#define SYNBASE %d\n", base);
 195         fprintf(hfile, "#define PEOF %d\n\n", -base);
 196         putc('\n', hfile);
 197         fputs("#define BASESYNTAX (basesyntax + SYNBASE)\n", hfile);
 198         fputs("#define DQSYNTAX (dqsyntax + SYNBASE)\n", hfile);
 199         fputs("#define SQSYNTAX (sqsyntax + SYNBASE)\n", hfile);
 200         fputs("#define ARISYNTAX (arisyntax + SYNBASE)\n", hfile);
 201         putc('\n', hfile);
 202         output_type_macros();           /* is_digit, etc. */
 203         putc('\n', hfile);
 204
 205         /* Generate the syntax tables. */
 206         fputs("#include \"shell.h\"\n", cfile);
 207         fputs("#include \"syntax.h\"\n\n", cfile);
 208         init();
 209         fputs("/* syntax table used when not in quotes */\n", cfile);
 210         add("\n", "CNL");
 211         add("\\", "CBACK");
 212         add("'", "CSQUOTE");
 213         add("\"", "CDQUOTE");
 214         add("`", "CBQUOTE");
 215         add("$", "CVAR");
 216         add("}", "CENDVAR");
 217         add("<>();&| \t", "CSPCL");
 218         print("basesyntax");
 219         init();
 220         fputs("\n/* syntax table used when in double quotes */\n", cfile);
 221         add("\n", "CNL");
 222         add("\\", "CBACK");
 223         add("\"", "CENDQUOTE");
 224         add("`", "CBQUOTE");
 225         add("$", "CVAR");
 226         add("}", "CENDVAR");
 227         /* ':/' for tilde expansion, '-' for [a\-x] pattern ranges */
 228         add("!*?[=~:/-", "CCTL");
 229         print("dqsyntax");
 230         init();
 231         fputs("\n/* syntax table used when in single quotes */\n", cfile);
 232         add("\n", "CNL");
 233         add("'", "CENDQUOTE");
 234         /* ':/' for tilde expansion, '-' for [a\-x] pattern ranges */
 235         add("!*?[=~:/-", "CCTL");
 236         print("sqsyntax");
 237         init();
 238         fputs("\n/* syntax table used when in arithmetic */\n", cfile);
 239         add("\n", "CNL");
 240         add("\\", "CBACK");
 241         add("`", "CBQUOTE");
 242         add("'", "CSQUOTE");
 243         add("\"", "CDQUOTE");
 244         add("$", "CVAR");
 245         add("}", "CENDVAR");
 246         add("(", "CLP");
 247         add(")", "CRP");
 248         print("arisyntax");
 249         filltable("0");
 250         fputs("\n/* character classification table */\n", cfile);
 251         add("0123456789", "ISDIGIT");
 252         add("abcdefghijklmnopqrstucvwxyz", "ISLOWER");
 253         add("ABCDEFGHIJKLMNOPQRSTUCVWXYZ", "ISUPPER");
 254         add("_", "ISUNDER");
 255         add("#?$!-*@", "ISSPECL");
 256         print("is_type");
 257         if (! digit_contig)
 258                 digit_convert();
 259         exit(0);
 260 }
 261
 262
 263
 264 /*
 265  * Clear the syntax table.
 266  */
 267
 268 static void
 269 filltable(char *dftval)
 270 {
 271         int i;
 272
 273         for (i = 0 ; i < size ; i++)
 274                 syntax[i] = dftval;
 275 }
 276
 277
 278 /*
 279  * Initialize the syntax table with default values.
 280  */
 281
 282 static void
 283 init(void)
 284 {
 285         filltable("CWORD");
 286         syntax[0] = "CEOF";
 287         syntax[base + CTLESC] = "CCTL";
 288         syntax[base + CTLVAR] = "CCTL";
 289         syntax[base + CTLENDVAR] = "CCTL";
 290         syntax[base + CTLBACKQ] = "CCTL";
 291         syntax[base + CTLBACKQ + CTLQUOTE] = "CCTL";
 292         syntax[base + CTLARI] = "CCTL";
 293         syntax[base + CTLENDARI] = "CCTL";
 294         syntax[base + CTLQUOTEMARK] = "CCTL";
 295 }
 296
 297
 298 /*
 299  * Add entries to the syntax table.
 300  */
 301
 302 static void
 303 add(char *p, char *type)
 304 {
 305         while (*p)
 306                 syntax[*p++ + base] = type;
 307 }
 308
 309
 310
 311 /*
 312  * Output the syntax table.
 313  */
 314
 315 static void
 316 print(char *name)
 317 {
 318         int i;
 319         int col;
 320
 321         fprintf(hfile, "extern const char %s[];\n", name);
 322         fprintf(cfile, "const char %s[%d] = {\n", name, size);
 323         col = 0;
 324         for (i = 0 ; i < size ; i++) {
 325                 if (i == 0) {
 326                         fputs("      ", cfile);
 327                 } else if ((i & 03) == 0) {
 328                         fputs(",\n      ", cfile);
 329                         col = 0;
 330                 } else {
 331                         putc(',', cfile);
 332                         while (++col < 9 * (i & 03))
 333                                 putc(' ', cfile);
 334                 }
 335                 fputs(syntax[i], cfile);
 336                 col += strlen(syntax[i]);
 337         }
 338         fputs("\n};\n", cfile);
 339 }
 340
 341
 342
 343 /*
 344  * Output character classification macros (e.g. is_digit).  If digits are
 345  * contiguous, we can test for them quickly.
 346  */
 347
 348 static char *macro[] = {
 349         "#define is_digit(c)\t((is_type+SYNBASE)[c] & ISDIGIT)",
 350         "#define is_alpha(c)\t((c) != PEOF && ((c) < CTLESC || (c) > CTLQUOTEMARK) && isalpha((unsigned char) (c)))",
 351         "#define is_name(c)\t((c) != PEOF && ((c) < CTLESC || (c) > CTLQUOTEMARK) && ((c) == '_' || isalpha((unsigned char) (c))))",
 352         "#define is_in_name(c)\t((c) != PEOF && ((c) < CTLESC || (c) > CTLQUOTEMARK) && ((c) == '_' || isalnum((unsigned char) (c))))",
 353         "#define is_special(c)\t((is_type+SYNBASE)[c] & (ISSPECL|ISDIGIT))",
 354         NULL
 355 };
 356
 357 static void
 358 output_type_macros(void)
 359 {
 360         char **pp;
 361
 362         if (digit_contig)
 363                 macro[0] = "#define is_digit(c)\t((unsigned)((c) - '0') <= 9)";
 364         for (pp = macro ; *pp ; pp++)
 365                 fprintf(hfile, "%s\n", *pp);
 366         if (digit_contig)
 367                 fputs("#define digit_val(c)\t((c) - '0')\n", hfile);
 368         else
 369                 fputs("#define digit_val(c)\t(digit_value[c])\n", hfile);
 370 }
 371
 372
 373
 374 /*
 375  * Output digit conversion table (if digits are not contiguous).
 376  */
 377
 378 static void
 379 digit_convert(void)
 380 {
 381         int maxdigit;
 382         static char digit[] = "0123456789";
 383         char *p;
 384         int i;
 385
 386         maxdigit = 0;
 387         for (p = digit ; *p ; p++)
 388                 if (*p > maxdigit)
 389                         maxdigit = *p;
 390         fputs("extern const char digit_value[];\n", hfile);
 391         fputs("\n\nconst char digit_value[] = {\n", cfile);
 392         for (i = 0 ; i <= maxdigit ; i++) {
 393                 for (p = digit ; *p && *p != i ; p++);
 394                 if (*p == '\0')
 395                         p = digit;
 396                 fprintf(cfile, "      %d,\n", (int)(p - digit));
 397         }
 398         fputs("};\n", cfile);
 399 }
 400
 401 /*
 402  * $PchId: mksyntax.c,v 1.7 2006/05/23 12:04:27 philip Exp $
 403  */