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