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