1 /* $NetBSD: for.c,v 1.141 2021/02/04 21:33:13 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 /* 33 * Handling of .for/.endfor loops in a makefile. 34 * 35 * For loops have the form: 36 * 37 * .for <varname...> in <value...> 38 * # the body 39 * .endfor 40 * 41 * When a .for line is parsed, the following lines are copied to the body of 42 * the .for loop, until the corresponding .endfor line is reached. In this 43 * phase, the body is not yet evaluated. This also applies to any nested 44 * .for loops. 45 * 46 * After reaching the .endfor, the values from the .for line are grouped 47 * according to the number of variables. For each such group, the unexpanded 48 * body is scanned for variable expressions, and those that match the variable 49 * names are replaced with expressions of the form ${:U...} or $(:U...). 50 * After that, the body is treated like a file from an .include directive. 51 * 52 * Interface: 53 * For_Eval Evaluate the loop in the passed line. 54 * 55 * For_Run Run accumulated loop 56 */ 57 58 #include "make.h" 59 60 /* "@(#)for.c 8.1 (Berkeley) 6/6/93" */ 61 MAKE_RCSID("$NetBSD: for.c,v 1.141 2021/02/04 21:33:13 rillig Exp $"); 62 63 64 /* One of the variables to the left of the "in" in a .for loop. */ 65 typedef struct ForVar { 66 char *name; 67 size_t nameLen; 68 } ForVar; 69 70 typedef struct ForLoop { 71 Buffer body; /* Unexpanded body of the loop */ 72 Vector /* of ForVar */ vars; /* Iteration variables */ 73 Words items; /* Substitution items */ 74 Buffer curBody; /* Expanded body of the current iteration */ 75 /* Is any of the names 1 character long? If so, when the variable values 76 * are substituted, the parser must handle $V expressions as well, not 77 * only ${V} and $(V). */ 78 Boolean short_var; 79 unsigned int sub_next; /* Where to continue iterating */ 80 } ForLoop; 81 82 83 static ForLoop *accumFor; /* Loop being accumulated */ 84 static int forLevel = 0; /* Nesting level */ 85 86 87 static ForLoop * 88 ForLoop_New(void) 89 { 90 ForLoop *f = bmake_malloc(sizeof *f); 91 92 Buf_Init(&f->body); 93 Vector_Init(&f->vars, sizeof(ForVar)); 94 f->items.words = NULL; 95 f->items.freeIt = NULL; 96 Buf_Init(&f->curBody); 97 f->short_var = FALSE; 98 f->sub_next = 0; 99 100 return f; 101 } 102 103 static void 104 ForLoop_Free(ForLoop *f) 105 { 106 Buf_Done(&f->body); 107 108 while (f->vars.len > 0) { 109 ForVar *var = Vector_Pop(&f->vars); 110 free(var->name); 111 } 112 Vector_Done(&f->vars); 113 114 Words_Free(f->items); 115 Buf_Done(&f->curBody); 116 117 free(f); 118 } 119 120 static void 121 ForLoop_AddVar(ForLoop *f, const char *name, size_t len) 122 { 123 ForVar *var = Vector_Push(&f->vars); 124 var->name = bmake_strldup(name, len); 125 var->nameLen = len; 126 } 127 128 static Boolean 129 ForLoop_ParseVarnames(ForLoop *f, const char **pp) 130 { 131 const char *p = *pp; 132 133 for (;;) { 134 size_t len; 135 136 cpp_skip_whitespace(&p); 137 if (*p == '\0') { 138 Parse_Error(PARSE_FATAL, "missing `in' in for"); 139 return FALSE; 140 } 141 142 /* 143 * XXX: This allows arbitrary variable names; 144 * see directive-for.mk. 145 */ 146 for (len = 1; p[len] != '\0' && !ch_isspace(p[len]); len++) 147 continue; 148 149 if (len == 2 && p[0] == 'i' && p[1] == 'n') { 150 p += 2; 151 break; 152 } 153 if (len == 1) 154 f->short_var = TRUE; 155 156 ForLoop_AddVar(f, p, len); 157 p += len; 158 } 159 160 if (f->vars.len == 0) { 161 Parse_Error(PARSE_FATAL, "no iteration variables in for"); 162 return FALSE; 163 } 164 165 *pp = p; 166 return TRUE; 167 } 168 169 static Boolean 170 ForLoop_ParseItems(ForLoop *f, const char *p) 171 { 172 char *items; 173 174 cpp_skip_whitespace(&p); 175 176 if (Var_Subst(p, SCOPE_GLOBAL, VARE_WANTRES, &items) != VPR_OK) { 177 Parse_Error(PARSE_FATAL, "Error in .for loop items"); 178 return FALSE; 179 } 180 181 f->items = Str_Words(items, FALSE); 182 free(items); 183 184 if (f->items.len == 1 && f->items.words[0][0] == '\0') 185 f->items.len = 0; /* .for var in ${:U} */ 186 187 if (f->items.len != 0 && f->items.len % f->vars.len != 0) { 188 Parse_Error(PARSE_FATAL, 189 "Wrong number of words (%u) in .for " 190 "substitution list with %u variables", 191 (unsigned)f->items.len, (unsigned)f->vars.len); 192 return FALSE; 193 } 194 195 return TRUE; 196 } 197 198 static Boolean 199 IsFor(const char *p) 200 { 201 return p[0] == 'f' && p[1] == 'o' && p[2] == 'r' && ch_isspace(p[3]); 202 } 203 204 static Boolean 205 IsEndfor(const char *p) 206 { 207 return p[0] == 'e' && strncmp(p, "endfor", 6) == 0 && 208 (p[6] == '\0' || ch_isspace(p[6])); 209 } 210 211 /* 212 * Evaluate the for loop in the passed line. The line looks like this: 213 * .for <varname...> in <value...> 214 * 215 * Input: 216 * line Line to parse 217 * 218 * Results: 219 * 0: Not a .for statement, parse the line 220 * 1: We found a for loop 221 * -1: A .for statement with a bad syntax error, discard. 222 */ 223 int 224 For_Eval(const char *line) 225 { 226 ForLoop *f; 227 const char *p; 228 229 p = line + 1; /* skip the '.' */ 230 cpp_skip_whitespace(&p); 231 232 if (!IsFor(p)) { 233 if (IsEndfor(p)) { 234 Parse_Error(PARSE_FATAL, "for-less endfor"); 235 return -1; 236 } 237 return 0; 238 } 239 p += 3; 240 241 f = ForLoop_New(); 242 243 if (!ForLoop_ParseVarnames(f, &p)) { 244 ForLoop_Free(f); 245 return -1; 246 } 247 248 if (!ForLoop_ParseItems(f, p)) { 249 /* Continue parsing the .for loop, but don't iterate. */ 250 f->items.len = 0; 251 } 252 253 accumFor = f; 254 forLevel = 1; 255 return 1; 256 } 257 258 /* 259 * Add another line to the .for loop that is being built up. 260 * Returns FALSE when the matching .endfor is reached. 261 */ 262 Boolean 263 For_Accum(const char *line) 264 { 265 const char *p = line; 266 267 if (*p == '.') { 268 p++; 269 cpp_skip_whitespace(&p); 270 271 if (IsEndfor(p)) { 272 DEBUG1(FOR, "For: end for %d\n", forLevel); 273 if (--forLevel <= 0) 274 return FALSE; 275 } else if (IsFor(p)) { 276 forLevel++; 277 DEBUG1(FOR, "For: new loop %d\n", forLevel); 278 } 279 } 280 281 Buf_AddStr(&accumFor->body, line); 282 Buf_AddByte(&accumFor->body, '\n'); 283 return TRUE; 284 } 285 286 287 static size_t 288 for_var_len(const char *var) 289 { 290 char ch, var_start, var_end; 291 int depth; 292 size_t len; 293 294 var_start = *var; 295 if (var_start == '\0') 296 /* just escape the $ */ 297 return 0; 298 299 if (var_start == '(') 300 var_end = ')'; 301 else if (var_start == '{') 302 var_end = '}'; 303 else 304 return 1; /* Single char variable */ 305 306 depth = 1; 307 for (len = 1; (ch = var[len++]) != '\0';) { 308 if (ch == var_start) 309 depth++; 310 else if (ch == var_end && --depth == 0) 311 return len; 312 } 313 314 /* Variable end not found, escape the $ */ 315 return 0; 316 } 317 318 /* 319 * The .for loop substitutes the items as ${:U<value>...}, which means 320 * that characters that break this syntax must be backslash-escaped. 321 */ 322 static Boolean 323 NeedsEscapes(const char *value, char endc) 324 { 325 const char *p; 326 327 for (p = value; *p != '\0'; p++) { 328 if (*p == ':' || *p == '$' || *p == '\\' || *p == endc) 329 return TRUE; 330 } 331 return FALSE; 332 } 333 334 /* 335 * While expanding the body of a .for loop, write the item in the ${:U...} 336 * expression, escaping characters as needed. 337 * 338 * The result is later unescaped by ApplyModifier_Defined. 339 */ 340 static void 341 Buf_AddEscaped(Buffer *cmds, const char *item, char endc) 342 { 343 char ch; 344 345 if (!NeedsEscapes(item, endc)) { 346 Buf_AddStr(cmds, item); 347 return; 348 } 349 350 /* Escape ':', '$', '\\' and 'endc' - these will be removed later by 351 * :U processing, see ApplyModifier_Defined. */ 352 while ((ch = *item++) != '\0') { 353 if (ch == '$') { 354 size_t len = for_var_len(item); 355 if (len != 0) { 356 Buf_AddBytes(cmds, item - 1, len + 1); 357 item += len; 358 continue; 359 } 360 Buf_AddByte(cmds, '\\'); 361 } else if (ch == ':' || ch == '\\' || ch == endc) 362 Buf_AddByte(cmds, '\\'); 363 Buf_AddByte(cmds, ch); 364 } 365 } 366 367 /* 368 * While expanding the body of a .for loop, replace the variable name of an 369 * expression like ${i} or ${i:...} or $(i) or $(i:...) with ":Uvalue". 370 */ 371 static void 372 ForLoop_SubstVarLong(ForLoop *f, const char **pp, const char *bodyEnd, 373 char endc, const char **inout_mark) 374 { 375 size_t i; 376 const char *p = *pp; 377 378 for (i = 0; i < f->vars.len; i++) { 379 ForVar *forVar = Vector_Get(&f->vars, i); 380 char *varname = forVar->name; 381 size_t varnameLen = forVar->nameLen; 382 383 if (varnameLen >= (size_t)(bodyEnd - p)) 384 continue; 385 if (memcmp(p, varname, varnameLen) != 0) 386 continue; 387 /* XXX: why test for backslash here? */ 388 if (p[varnameLen] != ':' && p[varnameLen] != endc && 389 p[varnameLen] != '\\') 390 continue; 391 392 /* 393 * Found a variable match. Skip over the variable name and 394 * instead add ':U<value>' to the current body. 395 */ 396 Buf_AddBytesBetween(&f->curBody, *inout_mark, p); 397 Buf_AddStr(&f->curBody, ":U"); 398 Buf_AddEscaped(&f->curBody, 399 f->items.words[f->sub_next + i], endc); 400 401 p += varnameLen; 402 *inout_mark = p; 403 *pp = p; 404 return; 405 } 406 } 407 408 /* 409 * While expanding the body of a .for loop, replace single-character 410 * variable expressions like $i with their ${:U...} expansion. 411 */ 412 static void 413 ForLoop_SubstVarShort(ForLoop *f, const char *p, const char **inout_mark) 414 { 415 const char ch = *p; 416 ForVar *vars; 417 size_t i; 418 419 /* Skip $$ and stupid ones. */ 420 if (!f->short_var || strchr("}):$", ch) != NULL) 421 return; 422 423 vars = Vector_Get(&f->vars, 0); 424 for (i = 0; i < f->vars.len; i++) { 425 const char *varname = vars[i].name; 426 if (varname[0] == ch && varname[1] == '\0') 427 goto found; 428 } 429 return; 430 431 found: 432 /* Replace $<ch> with ${:U<value>} */ 433 Buf_AddBytesBetween(&f->curBody, *inout_mark, p), *inout_mark = p + 1; 434 Buf_AddStr(&f->curBody, "{:U"); 435 Buf_AddEscaped(&f->curBody, f->items.words[f->sub_next + i], '}'); 436 Buf_AddByte(&f->curBody, '}'); 437 } 438 439 /* 440 * Compute the body for the current iteration by copying the unexpanded body, 441 * replacing the expressions for the iteration variables on the way. 442 * 443 * Using variable expressions ensures that the .for loop can't generate 444 * syntax, and that the later parsing will still see a variable. 445 * This code assumes that the variable with the empty name will never be 446 * defined, see unit-tests/varname-empty.mk for more details. 447 * 448 * The detection of substitutions of the loop control variables is naive. 449 * Many of the modifiers use '\' to escape '$' (not '$'), so it is possible 450 * to contrive a makefile where an unwanted substitution happens. 451 */ 452 static void 453 ForLoop_SubstBody(ForLoop *f) 454 { 455 const char *p, *bodyEnd; 456 const char *mark; /* where the last replacement left off */ 457 458 Buf_Empty(&f->curBody); 459 460 mark = f->body.data; 461 bodyEnd = f->body.data + f->body.len; 462 for (p = mark; (p = strchr(p, '$')) != NULL;) { 463 if (p[1] == '{' || p[1] == '(') { 464 p += 2; 465 ForLoop_SubstVarLong(f, &p, bodyEnd, 466 p[-1] == '{' ? '}' : ')', &mark); 467 } else if (p[1] != '\0') { 468 ForLoop_SubstVarShort(f, p + 1, &mark); 469 p += 2; 470 } else 471 break; 472 } 473 474 Buf_AddBytesBetween(&f->curBody, mark, bodyEnd); 475 } 476 477 /* 478 * Compute the body for the current iteration by copying the unexpanded body, 479 * replacing the expressions for the iteration variables on the way. 480 */ 481 static char * 482 ForReadMore(void *v_arg, size_t *out_len) 483 { 484 ForLoop *f = v_arg; 485 486 if (f->sub_next == f->items.len) { 487 /* No more iterations */ 488 ForLoop_Free(f); 489 return NULL; 490 } 491 492 ForLoop_SubstBody(f); 493 DEBUG1(FOR, "For: loop body:\n%s", f->curBody.data); 494 f->sub_next += (unsigned int)f->vars.len; 495 496 *out_len = f->curBody.len; 497 return f->curBody.data; 498 } 499 500 /* Run the .for loop, imitating the actions of an include file. */ 501 void 502 For_Run(int lineno) 503 { 504 ForLoop *f = accumFor; 505 accumFor = NULL; 506 507 if (f->items.len == 0) { 508 /* 509 * Nothing to expand - possibly due to an earlier syntax 510 * error. 511 */ 512 ForLoop_Free(f); 513 return; 514 } 515 516 Parse_SetInput(NULL, lineno, -1, ForReadMore, f); 517 } 518