contrib/bmake/for.c

   1 /*      $NetBSD: for.c,v 1.54 2020/07/03 08:13:23 rillig Exp $  */
   2
   3 /*
   4  * Copyright (c) 1992, The Regents of the University of California.
   5  * All rights reserved.
   6  *
   7  * Redistribution and use in source and binary forms, with or without
   8  * modification, are permitted provided that the following conditions
   9  * are met:
  10  * 1. Redistributions of source code must retain the above copyright
  11  *    notice, this list of conditions and the following disclaimer.
  12  * 2. Redistributions in binary form must reproduce the above copyright
  13  *    notice, this list of conditions and the following disclaimer in the
  14  *    documentation and/or other materials provided with the distribution.
  15  * 3. Neither the name of the University nor the names of its contributors
  16  *    may be used to endorse or promote products derived from this software
  17  *    without specific prior written permission.
  18  *
  19  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
  20  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
  21  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
  22  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
  23  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
  24  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
  25  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
  26  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
  27  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
  28  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
  29  * SUCH DAMAGE.
  30  */
  31
  32 #ifndef MAKE_NATIVE
  33 static char rcsid[] = "$NetBSD: for.c,v 1.54 2020/07/03 08:13:23 rillig Exp $";
  34 #else
  35 #include <sys/cdefs.h>
  36 #ifndef lint
  37 #if 0
  38 static char sccsid[] = "@(#)for.c       8.1 (Berkeley) 6/6/93";
  39 #else
  40 __RCSID("$NetBSD: for.c,v 1.54 2020/07/03 08:13:23 rillig Exp $");
  41 #endif
  42 #endif /* not lint */
  43 #endif
  44
  45 /*-
  46  * for.c --
  47  *      Functions to handle loops in a makefile.
  48  *
  49  * Interface:
  50  *      For_Eval        Evaluate the loop in the passed line.
  51  *      For_Run         Run accumulated loop
  52  *
  53  */
  54
  55 #include    <assert.h>
  56 #include    <ctype.h>
  57
  58 #include    "make.h"
  59 #include    "hash.h"
  60 #include    "dir.h"
  61 #include    "buf.h"
  62 #include    "strlist.h"
  63
  64 #define FOR_SUB_ESCAPE_CHAR  1
  65 #define FOR_SUB_ESCAPE_BRACE 2
  66 #define FOR_SUB_ESCAPE_PAREN 4
  67
  68 /*
  69  * For statements are of the form:
  70  *
  71  * .for <variable> in <varlist>
  72  * ...
  73  * .endfor
  74  *
  75  * The trick is to look for the matching end inside for for loop
  76  * To do that, we count the current nesting level of the for loops.
  77  * and the .endfor statements, accumulating all the statements between
  78  * the initial .for loop and the matching .endfor;
  79  * then we evaluate the for loop for each variable in the varlist.
  80  *
  81  * Note that any nested fors are just passed through; they get handled
  82  * recursively in For_Eval when we're expanding the enclosing for in
  83  * For_Run.
  84  */
  85
  86 static int        forLevel = 0;         /* Nesting level        */
  87
  88 /*
  89  * State of a for loop.
  90  */
  91 typedef struct _For {
  92     Buffer        buf;                  /* Body of loop         */
  93     strlist_t     vars;                 /* Iteration variables  */
  94     strlist_t     items;                /* Substitution items */
  95     char          *parse_buf;
  96     int           short_var;
  97     int           sub_next;
  98 } For;
  99
 100 static For        *accumFor;            /* Loop being accumulated */
 101
 102 \f
 103
 104 static char *
 105 make_str(const char *ptr, int len)
 106 {
 107         char *new_ptr;
 108
 109         new_ptr = bmake_malloc(len + 1);
 110         memcpy(new_ptr, ptr, len);
 111         new_ptr[len] = 0;
 112         return new_ptr;
 113 }
 114
 115 static void
 116 For_Free(For *arg)
 117 {
 118     Buf_Destroy(&arg->buf, TRUE);
 119     strlist_clean(&arg->vars);
 120     strlist_clean(&arg->items);
 121     free(arg->parse_buf);
 122
 123     free(arg);
 124 }
 125
 126 /*-
 127  *-----------------------------------------------------------------------
 128  * For_Eval --
 129  *      Evaluate the for loop in the passed line. The line
 130  *      looks like this:
 131  *          .for <variable> in <varlist>
 132  *
 133  * Input:
 134  *      line            Line to parse
 135  *
 136  * Results:
 137  *      0: Not a .for statement, parse the line
 138  *      1: We found a for loop
 139  *     -1: A .for statement with a bad syntax error, discard.
 140  *
 141  * Side Effects:
 142  *      None.
 143  *
 144  *-----------------------------------------------------------------------
 145  */
 146 int
 147 For_Eval(char *line)
 148 {
 149     For *new_for;
 150     char *ptr = line, *sub;
 151     int len;
 152     int escapes;
 153     unsigned char ch;
 154     char **words, *word_buf;
 155     int n, nwords;
 156
 157     /* Skip the '.' and any following whitespace */
 158     for (ptr++; *ptr && isspace((unsigned char) *ptr); ptr++)
 159         continue;
 160
 161     /*
 162      * If we are not in a for loop quickly determine if the statement is
 163      * a for.
 164      */
 165     if (ptr[0] != 'f' || ptr[1] != 'o' || ptr[2] != 'r' ||
 166             !isspace((unsigned char) ptr[3])) {
 167         if (ptr[0] == 'e' && strncmp(ptr+1, "ndfor", 5) == 0) {
 168             Parse_Error(PARSE_FATAL, "for-less endfor");
 169             return -1;
 170         }
 171         return 0;
 172     }
 173     ptr += 3;
 174
 175     /*
 176      * we found a for loop, and now we are going to parse it.
 177      */
 178
 179     new_for = bmake_malloc(sizeof *new_for);
 180     memset(new_for, 0, sizeof *new_for);
 181
 182     /* Grab the variables. Terminate on "in". */
 183     for (;; ptr += len) {
 184         while (*ptr && isspace((unsigned char) *ptr))
 185             ptr++;
 186         if (*ptr == '\0') {
 187             Parse_Error(PARSE_FATAL, "missing `in' in for");
 188             For_Free(new_for);
 189             return -1;
 190         }
 191         for (len = 1; ptr[len] && !isspace((unsigned char)ptr[len]); len++)
 192             continue;
 193         if (len == 2 && ptr[0] == 'i' && ptr[1] == 'n') {
 194             ptr += 2;
 195             break;
 196         }
 197         if (len == 1)
 198             new_for->short_var = 1;
 199         strlist_add_str(&new_for->vars, make_str(ptr, len), len);
 200     }
 201
 202     if (strlist_num(&new_for->vars) == 0) {
 203         Parse_Error(PARSE_FATAL, "no iteration variables in for");
 204         For_Free(new_for);
 205         return -1;
 206     }
 207
 208     while (*ptr && isspace((unsigned char) *ptr))
 209         ptr++;
 210
 211     /*
 212      * Make a list with the remaining words
 213      * The values are substituted as ${:U<value>...} so we must \ escape
 214      * characters that break that syntax.
 215      * Variables are fully expanded - so it is safe for escape $.
 216      * We can't do the escapes here - because we don't know whether
 217      * we are substuting into ${...} or $(...).
 218      */
 219     sub = Var_Subst(NULL, ptr, VAR_GLOBAL, VARF_WANTRES);
 220
 221     /*
 222      * Split into words allowing for quoted strings.
 223      */
 224     words = brk_string(sub, &nwords, FALSE, &word_buf);
 225
 226     free(sub);
 227
 228     if (words != NULL) {
 229         for (n = 0; n < nwords; n++) {
 230             ptr = words[n];
 231             if (!*ptr)
 232                 continue;
 233             escapes = 0;
 234             while ((ch = *ptr++)) {
 235                 switch(ch) {
 236                 case ':':
 237                 case '$':
 238                 case '\\':
 239                     escapes |= FOR_SUB_ESCAPE_CHAR;
 240                     break;
 241                 case ')':
 242                     escapes |= FOR_SUB_ESCAPE_PAREN;
 243                     break;
 244                 case /*{*/ '}':
 245                     escapes |= FOR_SUB_ESCAPE_BRACE;
 246                     break;
 247                 }
 248             }
 249             /*
 250              * We have to dup words[n] to maintain the semantics of
 251              * strlist.
 252              */
 253             strlist_add_str(&new_for->items, bmake_strdup(words[n]), escapes);
 254         }
 255
 256         free(words);
 257         free(word_buf);
 258
 259         if ((len = strlist_num(&new_for->items)) > 0 &&
 260             len % (n = strlist_num(&new_for->vars))) {
 261             Parse_Error(PARSE_FATAL,
 262                         "Wrong number of words (%d) in .for substitution list"
 263                         " with %d vars", len, n);
 264             /*
 265              * Return 'success' so that the body of the .for loop is
 266              * accumulated.
 267              * Remove all items so that the loop doesn't iterate.
 268              */
 269             strlist_clean(&new_for->items);
 270         }
 271     }
 272
 273     Buf_Init(&new_for->buf, 0);
 274     accumFor = new_for;
 275     forLevel = 1;
 276     return 1;
 277 }
 278
 279 /*
 280  * Add another line to a .for loop.
 281  * Returns 0 when the matching .endfor is reached.
 282  */
 283
 284 int
 285 For_Accum(char *line)
 286 {
 287     char *ptr = line;
 288
 289     if (*ptr == '.') {
 290
 291         for (ptr++; *ptr && isspace((unsigned char) *ptr); ptr++)
 292             continue;
 293
 294         if (strncmp(ptr, "endfor", 6) == 0 &&
 295                 (isspace((unsigned char) ptr[6]) || !ptr[6])) {
 296             if (DEBUG(FOR))
 297                 (void)fprintf(debug_file, "For: end for %d\n", forLevel);
 298             if (--forLevel <= 0)
 299                 return 0;
 300         } else if (strncmp(ptr, "for", 3) == 0 &&
 301                  isspace((unsigned char) ptr[3])) {
 302             forLevel++;
 303             if (DEBUG(FOR))
 304                 (void)fprintf(debug_file, "For: new loop %d\n", forLevel);
 305         }
 306     }
 307
 308     Buf_AddBytes(&accumFor->buf, strlen(line), line);
 309     Buf_AddByte(&accumFor->buf, '\n');
 310     return 1;
 311 }
 312
 313 \f
 314 /*-
 315  *-----------------------------------------------------------------------
 316  * For_Run --
 317  *      Run the for loop, imitating the actions of an include file
 318  *
 319  * Results:
 320  *      None.
 321  *
 322  * Side Effects:
 323  *      None.
 324  *
 325  *-----------------------------------------------------------------------
 326  */
 327
 328 static int
 329 for_var_len(const char *var)
 330 {
 331     char ch, var_start, var_end;
 332     int depth;
 333     int len;
 334
 335     var_start = *var;
 336     if (var_start == 0)
 337         /* just escape the $ */
 338         return 0;
 339
 340     if (var_start == '(')
 341         var_end = ')';
 342     else if (var_start == '{')
 343         var_end = '}';
 344     else
 345         /* Single char variable */
 346         return 1;
 347
 348     depth = 1;
 349     for (len = 1; (ch = var[len++]) != 0;) {
 350         if (ch == var_start)
 351             depth++;
 352         else if (ch == var_end && --depth == 0)
 353             return len;
 354     }
 355
 356     /* Variable end not found, escape the $ */
 357     return 0;
 358 }
 359
 360 static void
 361 for_substitute(Buffer *cmds, strlist_t *items, unsigned int item_no, char ech)
 362 {
 363     const char *item = strlist_str(items, item_no);
 364     int len;
 365     char ch;
 366
 367     /* If there were no escapes, or the only escape is the other variable
 368      * terminator, then just substitute the full string */
 369     if (!(strlist_info(items, item_no) &
 370             (ech == ')' ? ~FOR_SUB_ESCAPE_BRACE : ~FOR_SUB_ESCAPE_PAREN))) {
 371         Buf_AddBytes(cmds, strlen(item), item);
 372         return;
 373     }
 374
 375     /* Escape ':', '$', '\\' and 'ech' - removed by :U processing */
 376     while ((ch = *item++) != 0) {
 377         if (ch == '$') {
 378             len = for_var_len(item);
 379             if (len != 0) {
 380                 Buf_AddBytes(cmds, len + 1, item - 1);
 381                 item += len;
 382                 continue;
 383             }
 384             Buf_AddByte(cmds, '\\');
 385         } else if (ch == ':' || ch == '\\' || ch == ech)
 386             Buf_AddByte(cmds, '\\');
 387         Buf_AddByte(cmds, ch);
 388     }
 389 }
 390
 391 static char *
 392 For_Iterate(void *v_arg, size_t *ret_len)
 393 {
 394     For *arg = v_arg;
 395     int i, len;
 396     char *var;
 397     char *cp;
 398     char *cmd_cp;
 399     char *body_end;
 400     char ch;
 401     Buffer cmds;
 402
 403     if (arg->sub_next + strlist_num(&arg->vars) > strlist_num(&arg->items)) {
 404         /* No more iterations */
 405         For_Free(arg);
 406         return NULL;
 407     }
 408
 409     free(arg->parse_buf);
 410     arg->parse_buf = NULL;
 411
 412     /*
 413      * Scan the for loop body and replace references to the loop variables
 414      * with variable references that expand to the required text.
 415      * Using variable expansions ensures that the .for loop can't generate
 416      * syntax, and that the later parsing will still see a variable.
 417      * We assume that the null variable will never be defined.
 418      *
 419      * The detection of substitions of the loop control variable is naive.
 420      * Many of the modifiers use \ to escape $ (not $) so it is possible
 421      * to contrive a makefile where an unwanted substitution happens.
 422      */
 423
 424     cmd_cp = Buf_GetAll(&arg->buf, &len);
 425     body_end = cmd_cp + len;
 426     Buf_Init(&cmds, len + 256);
 427     for (cp = cmd_cp; (cp = strchr(cp, '$')) != NULL;) {
 428         char ech;
 429         ch = *++cp;
 430         if ((ch == '(' && (ech = ')', 1)) || (ch == '{' && (ech = '}', 1))) {
 431             cp++;
 432             /* Check variable name against the .for loop variables */
 433             STRLIST_FOREACH(var, &arg->vars, i) {
 434                 len = strlist_info(&arg->vars, i);
 435                 if (memcmp(cp, var, len) != 0)
 436                     continue;
 437                 if (cp[len] != ':' && cp[len] != ech && cp[len] != '\\')
 438                     continue;
 439                 /* Found a variable match. Replace with :U<value> */
 440                 Buf_AddBytes(&cmds, cp - cmd_cp, cmd_cp);
 441                 Buf_AddBytes(&cmds, 2, ":U");
 442                 cp += len;
 443                 cmd_cp = cp;
 444                 for_substitute(&cmds, &arg->items, arg->sub_next + i, ech);
 445                 break;
 446             }
 447             continue;
 448         }
 449         if (ch == 0)
 450             break;
 451         /* Probably a single character name, ignore $$ and stupid ones. {*/
 452         if (!arg->short_var || strchr("}):$", ch) != NULL) {
 453             cp++;
 454             continue;
 455         }
 456         STRLIST_FOREACH(var, &arg->vars, i) {
 457             if (var[0] != ch || var[1] != 0)
 458                 continue;
 459             /* Found a variable match. Replace with ${:U<value>} */
 460             Buf_AddBytes(&cmds, cp - cmd_cp, cmd_cp);
 461             Buf_AddBytes(&cmds, 3, "{:U");
 462             cmd_cp = ++cp;
 463             for_substitute(&cmds, &arg->items, arg->sub_next + i, /*{*/ '}');
 464             Buf_AddBytes(&cmds, 1, "}");
 465             break;
 466         }
 467     }
 468     Buf_AddBytes(&cmds, body_end - cmd_cp, cmd_cp);
 469
 470     cp = Buf_Destroy(&cmds, FALSE);
 471     if (DEBUG(FOR))
 472         (void)fprintf(debug_file, "For: loop body:\n%s", cp);
 473
 474     arg->sub_next += strlist_num(&arg->vars);
 475
 476     arg->parse_buf = cp;
 477     *ret_len = strlen(cp);
 478     return cp;
 479 }
 480
 481 void
 482 For_Run(int lineno)
 483 {
 484     For *arg;
 485
 486     arg = accumFor;
 487     accumFor = NULL;
 488
 489     if (strlist_num(&arg->items) == 0) {
 490         /* Nothing to expand - possibly due to an earlier syntax error. */
 491         For_Free(arg);
 492         return;
 493     }
 494
 495     Parse_SetInput(NULL, lineno, -1, For_Iterate, arg);
 496 }