lexi.c revision 1.119
1 /* $NetBSD: lexi.c,v 1.119 2021/10/31 09:52:37 rillig Exp $ */ 2 3 /*- 4 * SPDX-License-Identifier: BSD-4-Clause 5 * 6 * Copyright (c) 1985 Sun Microsystems, Inc. 7 * Copyright (c) 1980, 1993 8 * The Regents of the University of California. All rights reserved. 9 * All rights reserved. 10 * 11 * Redistribution and use in source and binary forms, with or without 12 * modification, are permitted provided that the following conditions 13 * are met: 14 * 1. Redistributions of source code must retain the above copyright 15 * notice, this list of conditions and the following disclaimer. 16 * 2. Redistributions in binary form must reproduce the above copyright 17 * notice, this list of conditions and the following disclaimer in the 18 * documentation and/or other materials provided with the distribution. 19 * 3. All advertising materials mentioning features or use of this software 20 * must display the following acknowledgement: 21 * This product includes software developed by the University of 22 * California, Berkeley and its contributors. 23 * 4. Neither the name of the University nor the names of its contributors 24 * may be used to endorse or promote products derived from this software 25 * without specific prior written permission. 26 * 27 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 28 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 29 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 30 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 31 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 32 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 33 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 34 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 35 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 36 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 37 * SUCH DAMAGE. 38 */ 39 40 #if 0 41 static char sccsid[] = "@(#)lexi.c 8.1 (Berkeley) 6/6/93"; 42 #endif 43 44 #include <sys/cdefs.h> 45 #if defined(__NetBSD__) 46 __RCSID("$NetBSD: lexi.c,v 1.119 2021/10/31 09:52:37 rillig Exp $"); 47 #elif defined(__FreeBSD__) 48 __FBSDID("$FreeBSD: head/usr.bin/indent/lexi.c 337862 2018-08-15 18:19:45Z pstef $"); 49 #endif 50 51 #include <ctype.h> 52 #include <stdlib.h> 53 #include <string.h> 54 55 #include "indent.h" 56 57 /* must be sorted alphabetically, is used in binary search */ 58 static const struct keyword { 59 const char *name; 60 enum keyword_kind kind; 61 } keywords[] = { 62 {"_Bool", kw_type}, 63 {"_Complex", kw_type}, 64 {"_Imaginary", kw_type}, 65 {"auto", kw_storage_class}, 66 {"bool", kw_type}, 67 {"break", kw_other}, 68 {"case", kw_case_or_default}, 69 {"char", kw_type}, 70 {"complex", kw_type}, 71 {"const", kw_type}, 72 {"continue", kw_other}, 73 {"default", kw_case_or_default}, 74 {"do", kw_do}, 75 {"double", kw_type}, 76 {"else", kw_else}, 77 {"enum", kw_tag}, 78 {"extern", kw_storage_class}, 79 {"float", kw_type}, 80 {"for", kw_for}, 81 {"goto", kw_other}, 82 {"if", kw_if}, 83 {"imaginary", kw_type}, 84 {"inline", kw_other}, 85 {"int", kw_type}, 86 {"long", kw_type}, 87 {"offsetof", kw_offsetof}, 88 {"register", kw_storage_class}, 89 {"restrict", kw_other}, 90 {"return", kw_other}, 91 {"short", kw_type}, 92 {"signed", kw_type}, 93 {"sizeof", kw_sizeof}, 94 {"static", kw_storage_class}, 95 {"struct", kw_tag}, 96 {"switch", kw_switch}, 97 {"typedef", kw_typedef}, 98 {"union", kw_tag}, 99 {"unsigned", kw_type}, 100 {"void", kw_type}, 101 {"volatile", kw_type}, 102 {"while", kw_while} 103 }; 104 105 static struct { 106 const char **items; 107 unsigned int len; 108 unsigned int cap; 109 } typenames; 110 111 /* 112 * The transition table below was rewritten by hand from lx's output, given 113 * the following definitions. lx is Katherine Flavel's lexer generator. 114 * 115 * O = /[0-7]/; D = /[0-9]/; NZ = /[1-9]/; 116 * H = /[a-f0-9]/i; B = /[0-1]/; HP = /0x/i; 117 * BP = /0b/i; E = /e[+\-]?/i D+; P = /p[+\-]?/i D+; 118 * FS = /[fl]/i; IS = /u/i /(l|L|ll|LL)/? | /(l|L|ll|LL)/ /u/i?; 119 * 120 * D+ E FS? -> $float; 121 * D* "." D+ E? FS? -> $float; 122 * D+ "." E? FS? -> $float; HP H+ IS? -> $int; 123 * HP H+ P FS? -> $float; NZ D* IS? -> $int; 124 * HP H* "." H+ P FS? -> $float; "0" O* IS? -> $int; 125 * HP H+ "." P FS -> $float; BP B+ IS? -> $int; 126 */ 127 /* INDENT OFF */ 128 static const unsigned char lex_number_state[][26] = { 129 /* examples: 130 00 131 s 0xx 132 t 00xaa 133 a 11 101100xxa.. 134 r 11ee0001101lbuuxx.a.pp 135 t.01.e+008bLuxll0Ll.aa.p+0 136 states: ABCDEFGHIJKLMNOPQRSTUVWXYZ */ 137 [0] = "uuiifuufiuuiiuiiiiiuiuuuuu", /* (other) */ 138 [1] = "CEIDEHHHIJQ U Q VUVVZZZ", /* 0 */ 139 [2] = "DEIDEHHHIJQ U Q VUVVZZZ", /* 1 */ 140 [3] = "DEIDEHHHIJ U VUVVZZZ", /* 2 3 4 5 6 7 */ 141 [4] = "DEJDEHHHJJ U VUVVZZZ", /* 8 9 */ 142 [5] = " U VUVV ", /* A a C c D d */ 143 [6] = " K U VUVV ", /* B b */ 144 [7] = " FFF FF U VUVV ", /* E e */ 145 [8] = " f f U VUVV f", /* F f */ 146 [9] = " LLf fL PR Li L f", /* L */ 147 [10] = " OOf fO S P O i O f", /* l */ 148 [11] = " FFX ", /* P p */ 149 [12] = " MM M i iiM M ", /* U u */ 150 [13] = " N ", /* X x */ 151 [14] = " G Y ", /* + - */ 152 [15] = "B EE EE T W ", /* . */ 153 /* ABCDEFGHIJKLMNOPQRSTUVWXYZ */ 154 }; 155 /* INDENT ON */ 156 157 static const unsigned char lex_number_row[] = { 158 ['0'] = 1, 159 ['1'] = 2, 160 ['2'] = 3, ['3'] = 3, ['4'] = 3, ['5'] = 3, ['6'] = 3, ['7'] = 3, 161 ['8'] = 4, ['9'] = 4, 162 ['A'] = 5, ['a'] = 5, ['C'] = 5, ['c'] = 5, ['D'] = 5, ['d'] = 5, 163 ['B'] = 6, ['b'] = 6, 164 ['E'] = 7, ['e'] = 7, 165 ['F'] = 8, ['f'] = 8, 166 ['L'] = 9, 167 ['l'] = 10, 168 ['P'] = 11, ['p'] = 11, 169 ['U'] = 12, ['u'] = 12, 170 ['X'] = 13, ['x'] = 13, 171 ['+'] = 14, ['-'] = 14, 172 ['.'] = 15, 173 }; 174 175 static char 176 inbuf_peek(void) 177 { 178 return *inp.s; 179 } 180 181 void 182 inbuf_skip(void) 183 { 184 inp.s++; 185 if (inp.s >= inp.e) 186 inbuf_read_line(); 187 } 188 189 char 190 inbuf_next(void) 191 { 192 char ch = inbuf_peek(); 193 inbuf_skip(); 194 return ch; 195 } 196 197 static void 198 check_size_token(size_t desired_size) 199 { 200 if (token.e + desired_size >= token.l) 201 buf_expand(&token, desired_size); 202 } 203 204 static void 205 token_add_char(char ch) 206 { 207 check_size_token(1); 208 *token.e++ = ch; 209 } 210 211 #ifdef debug 212 static const char * 213 lsym_name(lexer_symbol sym) 214 { 215 static const char *const name[] = { 216 "eof", 217 "preprocessing", 218 "newline", 219 "form_feed", 220 "comment", 221 "lparen_or_lbracket", 222 "rparen_or_rbracket", 223 "lbrace", 224 "rbrace", 225 "period", 226 "unary_op", 227 "binary_op", 228 "postfix_op", 229 "question", 230 "colon", 231 "comma", 232 "semicolon", 233 "typedef", 234 "storage_class", 235 "type", 236 "tag", 237 "case_label", 238 "string_prefix", 239 "ident", 240 "funcname", 241 "do", 242 "else", 243 "for", 244 "if", 245 "switch", 246 "while", 247 }; 248 249 return name[sym]; 250 } 251 252 static const char * 253 kw_name(enum keyword_kind kw) 254 { 255 static const char *const name[] = { 256 "0", 257 "offsetof", 258 "sizeof", 259 "tag", 260 "type", 261 "for", 262 "if", 263 "while", 264 "do", 265 "else", 266 "switch", 267 "case_or_default", 268 "storage_class", 269 "typedef", 270 "other", 271 }; 272 273 return name[kw]; 274 } 275 276 static void 277 debug_print_buf(const char *name, const struct buffer *buf) 278 { 279 if (buf->s < buf->e) { 280 debug_printf("%s ", name); 281 debug_vis_range("\"", buf->s, buf->e, "\"\n"); 282 } 283 } 284 285 #define debug_ps_bool(name) \ 286 if (ps.name != prev_ps.name) \ 287 debug_println("[%c] ps." #name, ps.name ? 'x' : ' ') 288 #define debug_ps_int(name) \ 289 if (ps.name != prev_ps.name) \ 290 debug_println("%3d ps." #name, ps.name) 291 #define debug_ps_keyword(name) \ 292 if (ps.name != kw_0) \ 293 debug_println(" ps." #name " = %s", kw_name(ps.name)) 294 295 static void 296 debug_lexi(lexer_symbol lsym) 297 { 298 /* 299 * Watch out for 'rolled back parser state' in the debug output; the 300 * differences around these are unreliable. 301 */ 302 static struct parser_state prev_ps; 303 304 debug_println(""); 305 debug_printf("line %d: %s", line_no, lsym_name(lsym)); 306 debug_vis_range(" \"", token.s, token.e, "\"\n"); 307 debug_print_buf("label", &lab); 308 debug_print_buf("code", &code); 309 debug_print_buf("comment", &com); 310 311 // prev_token 312 debug_ps_keyword(prev_keyword); 313 debug_ps_keyword(curr_keyword); 314 debug_ps_bool(curr_newline); 315 debug_ps_bool(curr_col_1); 316 debug_ps_bool(next_unary); 317 // procname 318 debug_ps_bool(want_blank); 319 debug_ps_int(paren_level); 320 debug_ps_int(p_l_follow); 321 // paren_indents 322 debug_ps_int(cast_mask); 323 debug_ps_int(not_cast_mask); 324 325 debug_ps_int(comment_delta); 326 debug_ps_int(n_comment_delta); 327 debug_ps_int(com_ind); 328 329 debug_ps_bool(block_init); 330 debug_ps_int(block_init_level); 331 debug_ps_bool(init_or_struct); 332 333 debug_ps_int(ind_level); 334 debug_ps_int(ind_level_follow); 335 336 debug_ps_int(decl_nest); 337 debug_ps_bool(decl_on_line); 338 debug_ps_bool(in_decl); 339 debug_ps_int(just_saw_decl); 340 debug_ps_bool(in_parameter_declaration); 341 debug_ps_bool(decl_indent_done); 342 343 debug_ps_bool(in_stmt); 344 debug_ps_bool(ind_stmt); 345 debug_ps_bool(is_case_label); 346 347 debug_ps_bool(search_stmt); 348 349 prev_ps = ps; 350 } 351 #endif 352 353 /* ARGSUSED */ 354 static lexer_symbol 355 lexi_end(lexer_symbol lsym) 356 { 357 #ifdef debug 358 debug_lexi(lsym); 359 #endif 360 return lsym; 361 } 362 363 static void 364 lex_number(void) 365 { 366 for (unsigned char s = 'A'; s != 'f' && s != 'i' && s != 'u';) { 367 unsigned char ch = (unsigned char)*inp.s; 368 if (ch >= array_length(lex_number_row) || lex_number_row[ch] == 0) 369 break; 370 371 unsigned char row = lex_number_row[ch]; 372 if (lex_number_state[row][s - 'A'] == ' ') { 373 /*- 374 * lex_number_state[0][s - 'A'] now indicates the type: 375 * f = floating, i = integer, u = unknown 376 */ 377 break; 378 } 379 380 s = lex_number_state[row][s - 'A']; 381 token_add_char(inbuf_next()); 382 } 383 } 384 385 static void 386 lex_word(void) 387 { 388 while (isalnum((unsigned char)*inp.s) || 389 *inp.s == '\\' || 390 *inp.s == '_' || *inp.s == '$') { 391 392 if (*inp.s == '\\') { 393 if (inp.s[1] == '\n') { 394 inp.s += 2; 395 if (inp.s >= inp.e) 396 inbuf_read_line(); 397 } else 398 break; 399 } 400 401 token_add_char(inbuf_next()); 402 } 403 } 404 405 static void 406 lex_char_or_string(void) 407 { 408 for (char delim = *token.s;;) { 409 if (*inp.s == '\n') { 410 diag(1, "Unterminated literal"); 411 return; 412 } 413 414 token_add_char(inbuf_next()); 415 if (token.e[-1] == delim) 416 return; 417 418 if (token.e[-1] == '\\') { 419 if (*inp.s == '\n') 420 ++line_no; 421 token_add_char(inbuf_next()); 422 } 423 } 424 } 425 426 /* Guess whether the current token is a declared type. */ 427 static bool 428 probably_typename(void) 429 { 430 if (ps.p_l_follow > 0) 431 return false; 432 if (ps.block_init || ps.in_stmt) 433 return false; 434 if (inp.s[0] == '*' && inp.s[1] != '=') 435 goto maybe; 436 if (isalpha((unsigned char)*inp.s)) 437 goto maybe; 438 return false; 439 maybe: 440 return ps.prev_token == lsym_semicolon || 441 ps.prev_token == lsym_lbrace || 442 ps.prev_token == lsym_rbrace; 443 } 444 445 static int 446 bsearch_typenames(const char *key) 447 { 448 const char **arr = typenames.items; 449 int lo = 0; 450 int hi = (int)typenames.len - 1; 451 452 while (lo <= hi) { 453 int mid = (int)((unsigned)(lo + hi) >> 1); 454 int cmp = strcmp(arr[mid], key); 455 if (cmp < 0) 456 lo = mid + 1; 457 else if (cmp > 0) 458 hi = mid - 1; 459 else 460 return mid; 461 } 462 return -(lo + 1); 463 } 464 465 static bool 466 is_typename(void) 467 { 468 if (opt.auto_typedefs && 469 token.e - token.s >= 2 && memcmp(token.e - 2, "_t", 2) == 0) 470 return true; 471 472 return bsearch_typenames(token.s) >= 0; 473 } 474 475 static int 476 cmp_keyword_by_name(const void *key, const void *elem) 477 { 478 return strcmp(key, ((const struct keyword *)elem)->name); 479 } 480 481 /* Read an alphanumeric token into 'token', or return end_of_file. */ 482 static lexer_symbol 483 lexi_alnum(void) 484 { 485 if (isdigit((unsigned char)*inp.s) || 486 (inp.s[0] == '.' && isdigit((unsigned char)inp.s[1]))) { 487 lex_number(); 488 } else if (isalnum((unsigned char)*inp.s) || 489 *inp.s == '_' || *inp.s == '$') { 490 lex_word(); 491 } else 492 return lsym_eof; /* just as a placeholder */ 493 494 *token.e = '\0'; 495 496 if (token.s[0] == 'L' && token.s[1] == '\0' && 497 (*inp.s == '"' || *inp.s == '\'')) 498 return lsym_string_prefix; 499 500 while (ch_isblank(inbuf_peek())) 501 inbuf_skip(); 502 503 if (ps.prev_token == lsym_tag && ps.p_l_follow == 0) { 504 ps.next_unary = true; 505 return lsym_type; 506 } 507 508 /* Operator after identifier is binary unless last token was 'struct'. */ 509 ps.next_unary = ps.prev_token == lsym_tag; 510 511 const struct keyword *kw = bsearch(token.s, keywords, 512 array_length(keywords), sizeof(keywords[0]), cmp_keyword_by_name); 513 if (kw == NULL) { 514 if (is_typename()) { 515 ps.curr_keyword = kw_type; 516 ps.next_unary = true; 517 goto found_typename; 518 } 519 520 } else { /* we have a keyword */ 521 ps.curr_keyword = kw->kind; 522 ps.next_unary = true; 523 524 /* INDENT OFF */ 525 switch (kw->kind) { 526 case kw_tag: 527 case kw_type: goto found_typename; 528 case kw_case_or_default: return lsym_case_label; 529 case kw_for: return lsym_for; 530 case kw_if: return lsym_if; 531 case kw_else: return lsym_else; 532 case kw_switch: return lsym_switch; 533 case kw_while: return lsym_while; 534 case kw_do: return lsym_do; 535 case kw_storage_class: return lsym_storage_class; 536 case kw_typedef: return lsym_typedef; 537 default: return lsym_ident; 538 } 539 /* INDENT ON */ 540 541 found_typename: 542 if (ps.p_l_follow > 0) { 543 /* inside parentheses: cast, param list, offsetof or sizeof */ 544 ps.cast_mask |= (1 << ps.p_l_follow) & ~ps.not_cast_mask; 545 } 546 if (ps.prev_token != lsym_period && ps.prev_token != lsym_unary_op) { 547 if (kw != NULL && kw->kind == kw_tag) 548 return lsym_tag; 549 if (ps.p_l_follow == 0) 550 return lsym_type; 551 } 552 } 553 554 if (*inp.s == '(' && ps.tos <= 1 && ps.ind_level == 0 && 555 !ps.in_parameter_declaration && !ps.block_init) { 556 557 for (const char *p = inp.s; p < inp.e;) 558 if (*p++ == ')' && (*p == ';' || *p == ',')) 559 goto no_function_definition; 560 561 strncpy(ps.procname, token.s, sizeof ps.procname - 1); 562 if (ps.in_decl) 563 ps.in_parameter_declaration = true; 564 return lsym_funcname; 565 no_function_definition:; 566 567 } else if (probably_typename()) { 568 ps.curr_keyword = kw_type; 569 ps.next_unary = true; 570 return lsym_type; 571 } 572 573 if (ps.prev_token == lsym_type) /* if this is a declared variable, 574 * then following sign is unary */ 575 ps.next_unary = true; /* will make "int a -1" work */ 576 577 return lsym_ident; /* the ident is not in the list */ 578 } 579 580 /* Reads the next token, placing it in the global variable "token". */ 581 lexer_symbol 582 lexi(void) 583 { 584 token.e = token.s; 585 ps.curr_col_1 = ps.curr_newline; 586 ps.curr_newline = false; 587 ps.prev_keyword = ps.curr_keyword; 588 ps.curr_keyword = kw_0; 589 590 while (ch_isblank(*inp.s)) { 591 ps.curr_col_1 = false; 592 inbuf_skip(); 593 } 594 595 lexer_symbol alnum_lsym = lexi_alnum(); 596 if (alnum_lsym != lsym_eof) 597 return lexi_end(alnum_lsym); 598 599 /* Scan a non-alphanumeric token */ 600 601 check_size_token(3); /* for things like "<<=" */ 602 *token.e++ = inbuf_next(); 603 *token.e = '\0'; 604 605 lexer_symbol lsym; 606 bool unary_delim = false; /* whether the current token forces a 607 * following operator to be unary */ 608 609 switch (*token.s) { 610 case '\n': 611 unary_delim = ps.next_unary; 612 ps.curr_newline = true; 613 /* if data has been exhausted, the newline is a dummy. */ 614 lsym = had_eof ? lsym_eof : lsym_newline; 615 break; 616 617 case '\'': 618 case '"': 619 lex_char_or_string(); 620 lsym = lsym_ident; 621 break; 622 623 case '(': 624 case '[': 625 unary_delim = true; 626 lsym = lsym_lparen_or_lbracket; 627 break; 628 629 case ')': 630 case ']': 631 lsym = lsym_rparen_or_rbracket; 632 break; 633 634 case '#': 635 unary_delim = ps.next_unary; 636 lsym = lsym_preprocessing; 637 break; 638 639 case '?': 640 unary_delim = true; 641 lsym = lsym_question; 642 break; 643 644 case ':': 645 lsym = lsym_colon; 646 unary_delim = true; 647 break; 648 649 case ';': 650 unary_delim = true; 651 lsym = lsym_semicolon; 652 break; 653 654 case '{': 655 unary_delim = true; 656 lsym = lsym_lbrace; 657 break; 658 659 case '}': 660 unary_delim = true; 661 lsym = lsym_rbrace; 662 break; 663 664 case '\f': 665 unary_delim = ps.next_unary; 666 ps.curr_newline = true; 667 lsym = lsym_form_feed; 668 break; 669 670 case ',': 671 unary_delim = true; 672 lsym = lsym_comma; 673 break; 674 675 case '.': 676 unary_delim = false; 677 lsym = lsym_period; 678 break; 679 680 case '-': 681 case '+': 682 lsym = ps.next_unary ? lsym_unary_op : lsym_binary_op; 683 unary_delim = true; 684 685 if (*inp.s == token.s[0]) { /* ++, -- */ 686 *token.e++ = *inp.s++; 687 if (ps.prev_token == lsym_ident || 688 ps.prev_token == lsym_rparen_or_rbracket) { 689 lsym = ps.next_unary ? lsym_unary_op : lsym_postfix_op; 690 unary_delim = false; 691 } 692 693 } else if (*inp.s == '=') { /* += */ 694 *token.e++ = *inp.s++; 695 696 } else if (*inp.s == '>') { /* -> */ 697 *token.e++ = *inp.s++; 698 unary_delim = false; 699 lsym = lsym_unary_op; 700 ps.want_blank = false; 701 } 702 break; 703 704 case '=': 705 if (ps.init_or_struct) 706 ps.block_init = true; 707 if (*inp.s == '=') { /* == */ 708 *token.e++ = *inp.s++; 709 *token.e = '\0'; 710 } 711 lsym = lsym_binary_op; 712 unary_delim = true; 713 break; 714 715 case '>': 716 case '<': 717 case '!': /* ops like <, <<, <=, !=, etc */ 718 if (*inp.s == '>' || *inp.s == '<' || *inp.s == '=') 719 *token.e++ = inbuf_next(); 720 if (*inp.s == '=') 721 *token.e++ = *inp.s++; 722 lsym = ps.next_unary ? lsym_unary_op : lsym_binary_op; 723 unary_delim = true; 724 break; 725 726 case '*': 727 unary_delim = true; 728 if (!ps.next_unary) { 729 if (*inp.s == '=') 730 *token.e++ = *inp.s++; 731 lsym = lsym_binary_op; 732 break; 733 } 734 735 while (*inp.s == '*' || isspace((unsigned char)*inp.s)) { 736 if (*inp.s == '*') 737 token_add_char('*'); 738 inbuf_skip(); 739 } 740 741 if (ps.in_decl) { 742 char *tp = inp.s; 743 744 while (isalpha((unsigned char)*tp) || 745 isspace((unsigned char)*tp)) { 746 if (++tp >= inp.e) 747 inbuf_read_line(); 748 } 749 if (*tp == '(') 750 ps.procname[0] = ' '; 751 } 752 753 lsym = lsym_unary_op; 754 break; 755 756 default: 757 if (token.s[0] == '/' && (*inp.s == '*' || *inp.s == '/')) { 758 /* it is start of comment */ 759 *token.e++ = inbuf_next(); 760 761 lsym = lsym_comment; 762 unary_delim = ps.next_unary; 763 break; 764 } 765 766 while (token.e[-1] == *inp.s || *inp.s == '=') { 767 /* handle '||', '&&', etc., and also things as in 'int *****i' */ 768 token_add_char(inbuf_next()); 769 } 770 771 lsym = ps.next_unary ? lsym_unary_op : lsym_binary_op; 772 unary_delim = true; 773 } 774 775 if (inp.s >= inp.e) /* check for input buffer empty */ 776 inbuf_read_line(); 777 778 ps.next_unary = unary_delim; 779 780 check_size_token(1); 781 *token.e = '\0'; 782 783 return lexi_end(lsym); 784 } 785 786 void 787 add_typename(const char *name) 788 { 789 if (typenames.len >= typenames.cap) { 790 typenames.cap = 16 + 2 * typenames.cap; 791 typenames.items = xrealloc(typenames.items, 792 sizeof(typenames.items[0]) * typenames.cap); 793 } 794 795 int pos = bsearch_typenames(name); 796 if (pos >= 0) 797 return; /* already in the list */ 798 799 pos = -(pos + 1); 800 memmove(typenames.items + pos + 1, typenames.items + pos, 801 sizeof(typenames.items[0]) * (typenames.len++ - (unsigned)pos)); 802 typenames.items[pos] = xstrdup(name); 803 } 804
Indexes created Wed Sep 24 22:09:59 GMT 2025