lexi.c revision 1.120
1 /* $NetBSD: lexi.c,v 1.120 2021/10/31 10:00:37 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 1.120 2021/10/31 10:00:37 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 "sizeof", 240 "ident", 241 "funcname", 242 "do", 243 "else", 244 "for", 245 "if", 246 "switch", 247 "while", 248 }; 249 250 return name[sym]; 251 } 252 253 static const char * 254 kw_name(enum keyword_kind kw) 255 { 256 static const char *const name[] = { 257 "0", 258 "offsetof", 259 "sizeof", 260 "tag", 261 "type", 262 "for", 263 "if", 264 "while", 265 "do", 266 "else", 267 "switch", 268 "case_or_default", 269 "storage_class", 270 "typedef", 271 "other", 272 }; 273 274 return name[kw]; 275 } 276 277 static void 278 debug_print_buf(const char *name, const struct buffer *buf) 279 { 280 if (buf->s < buf->e) { 281 debug_printf("%s ", name); 282 debug_vis_range("\"", buf->s, buf->e, "\"\n"); 283 } 284 } 285 286 #define debug_ps_bool(name) \ 287 if (ps.name != prev_ps.name) \ 288 debug_println("[%c] ps." #name, ps.name ? 'x' : ' ') 289 #define debug_ps_int(name) \ 290 if (ps.name != prev_ps.name) \ 291 debug_println("%3d ps." #name, ps.name) 292 #define debug_ps_keyword(name) \ 293 if (ps.name != kw_0) \ 294 debug_println(" ps." #name " = %s", kw_name(ps.name)) 295 296 static void 297 debug_lexi(lexer_symbol lsym) 298 { 299 /* 300 * Watch out for 'rolled back parser state' in the debug output; the 301 * differences around these are unreliable. 302 */ 303 static struct parser_state prev_ps; 304 305 debug_println(""); 306 debug_printf("line %d: %s", line_no, lsym_name(lsym)); 307 debug_vis_range(" \"", token.s, token.e, "\"\n"); 308 debug_print_buf("label", &lab); 309 debug_print_buf("code", &code); 310 debug_print_buf("comment", &com); 311 312 // prev_token 313 debug_ps_keyword(prev_keyword); 314 debug_ps_keyword(curr_keyword); 315 debug_ps_bool(curr_newline); 316 debug_ps_bool(curr_col_1); 317 debug_ps_bool(next_unary); 318 // procname 319 debug_ps_bool(want_blank); 320 debug_ps_int(paren_level); 321 debug_ps_int(p_l_follow); 322 // paren_indents 323 debug_ps_int(cast_mask); 324 debug_ps_int(not_cast_mask); 325 326 debug_ps_int(comment_delta); 327 debug_ps_int(n_comment_delta); 328 debug_ps_int(com_ind); 329 330 debug_ps_bool(block_init); 331 debug_ps_int(block_init_level); 332 debug_ps_bool(init_or_struct); 333 334 debug_ps_int(ind_level); 335 debug_ps_int(ind_level_follow); 336 337 debug_ps_int(decl_nest); 338 debug_ps_bool(decl_on_line); 339 debug_ps_bool(in_decl); 340 debug_ps_int(just_saw_decl); 341 debug_ps_bool(in_parameter_declaration); 342 debug_ps_bool(decl_indent_done); 343 344 debug_ps_bool(in_stmt); 345 debug_ps_bool(ind_stmt); 346 debug_ps_bool(is_case_label); 347 348 debug_ps_bool(search_stmt); 349 350 prev_ps = ps; 351 } 352 #endif 353 354 /* ARGSUSED */ 355 static lexer_symbol 356 lexi_end(lexer_symbol lsym) 357 { 358 #ifdef debug 359 debug_lexi(lsym); 360 #endif 361 return lsym; 362 } 363 364 static void 365 lex_number(void) 366 { 367 for (unsigned char s = 'A'; s != 'f' && s != 'i' && s != 'u';) { 368 unsigned char ch = (unsigned char)*inp.s; 369 if (ch >= array_length(lex_number_row) || lex_number_row[ch] == 0) 370 break; 371 372 unsigned char row = lex_number_row[ch]; 373 if (lex_number_state[row][s - 'A'] == ' ') { 374 /*- 375 * lex_number_state[0][s - 'A'] now indicates the type: 376 * f = floating, i = integer, u = unknown 377 */ 378 break; 379 } 380 381 s = lex_number_state[row][s - 'A']; 382 token_add_char(inbuf_next()); 383 } 384 } 385 386 static void 387 lex_word(void) 388 { 389 while (isalnum((unsigned char)*inp.s) || 390 *inp.s == '\\' || 391 *inp.s == '_' || *inp.s == '$') { 392 393 if (*inp.s == '\\') { 394 if (inp.s[1] == '\n') { 395 inp.s += 2; 396 if (inp.s >= inp.e) 397 inbuf_read_line(); 398 } else 399 break; 400 } 401 402 token_add_char(inbuf_next()); 403 } 404 } 405 406 static void 407 lex_char_or_string(void) 408 { 409 for (char delim = *token.s;;) { 410 if (*inp.s == '\n') { 411 diag(1, "Unterminated literal"); 412 return; 413 } 414 415 token_add_char(inbuf_next()); 416 if (token.e[-1] == delim) 417 return; 418 419 if (token.e[-1] == '\\') { 420 if (*inp.s == '\n') 421 ++line_no; 422 token_add_char(inbuf_next()); 423 } 424 } 425 } 426 427 /* Guess whether the current token is a declared type. */ 428 static bool 429 probably_typename(void) 430 { 431 if (ps.p_l_follow > 0) 432 return false; 433 if (ps.block_init || ps.in_stmt) 434 return false; 435 if (inp.s[0] == '*' && inp.s[1] != '=') 436 goto maybe; 437 if (isalpha((unsigned char)*inp.s)) 438 goto maybe; 439 return false; 440 maybe: 441 return ps.prev_token == lsym_semicolon || 442 ps.prev_token == lsym_lbrace || 443 ps.prev_token == lsym_rbrace; 444 } 445 446 static int 447 bsearch_typenames(const char *key) 448 { 449 const char **arr = typenames.items; 450 int lo = 0; 451 int hi = (int)typenames.len - 1; 452 453 while (lo <= hi) { 454 int mid = (int)((unsigned)(lo + hi) >> 1); 455 int cmp = strcmp(arr[mid], key); 456 if (cmp < 0) 457 lo = mid + 1; 458 else if (cmp > 0) 459 hi = mid - 1; 460 else 461 return mid; 462 } 463 return -(lo + 1); 464 } 465 466 static bool 467 is_typename(void) 468 { 469 if (opt.auto_typedefs && 470 token.e - token.s >= 2 && memcmp(token.e - 2, "_t", 2) == 0) 471 return true; 472 473 return bsearch_typenames(token.s) >= 0; 474 } 475 476 static int 477 cmp_keyword_by_name(const void *key, const void *elem) 478 { 479 return strcmp(key, ((const struct keyword *)elem)->name); 480 } 481 482 /* Read an alphanumeric token into 'token', or return end_of_file. */ 483 static lexer_symbol 484 lexi_alnum(void) 485 { 486 if (isdigit((unsigned char)*inp.s) || 487 (inp.s[0] == '.' && isdigit((unsigned char)inp.s[1]))) { 488 lex_number(); 489 } else if (isalnum((unsigned char)*inp.s) || 490 *inp.s == '_' || *inp.s == '$') { 491 lex_word(); 492 } else 493 return lsym_eof; /* just as a placeholder */ 494 495 *token.e = '\0'; 496 497 if (token.s[0] == 'L' && token.s[1] == '\0' && 498 (*inp.s == '"' || *inp.s == '\'')) 499 return lsym_string_prefix; 500 501 while (ch_isblank(inbuf_peek())) 502 inbuf_skip(); 503 504 if (ps.prev_token == lsym_tag && ps.p_l_follow == 0) { 505 ps.next_unary = true; 506 return lsym_type; 507 } 508 509 /* Operator after identifier is binary unless last token was 'struct'. */ 510 ps.next_unary = ps.prev_token == lsym_tag; 511 512 const struct keyword *kw = bsearch(token.s, keywords, 513 array_length(keywords), sizeof(keywords[0]), cmp_keyword_by_name); 514 if (kw == NULL) { 515 if (is_typename()) { 516 ps.curr_keyword = kw_type; 517 ps.next_unary = true; 518 goto found_typename; 519 } 520 521 } else { /* we have a keyword */ 522 ps.curr_keyword = kw->kind; 523 ps.next_unary = true; 524 525 /* INDENT OFF */ 526 switch (kw->kind) { 527 case kw_tag: 528 case kw_type: goto found_typename; 529 case kw_case_or_default: return lsym_case_label; 530 case kw_for: return lsym_for; 531 case kw_if: return lsym_if; 532 case kw_else: return lsym_else; 533 case kw_switch: return lsym_switch; 534 case kw_while: return lsym_while; 535 case kw_do: return lsym_do; 536 case kw_storage_class: return lsym_storage_class; 537 case kw_typedef: return lsym_typedef; 538 case kw_sizeof: return lsym_sizeof; 539 default: return lsym_ident; 540 } 541 /* INDENT ON */ 542 543 found_typename: 544 if (ps.p_l_follow > 0) { 545 /* inside parentheses: cast, param list, offsetof or sizeof */ 546 ps.cast_mask |= (1 << ps.p_l_follow) & ~ps.not_cast_mask; 547 } 548 if (ps.prev_token != lsym_period && ps.prev_token != lsym_unary_op) { 549 if (kw != NULL && kw->kind == kw_tag) 550 return lsym_tag; 551 if (ps.p_l_follow == 0) 552 return lsym_type; 553 } 554 } 555 556 if (*inp.s == '(' && ps.tos <= 1 && ps.ind_level == 0 && 557 !ps.in_parameter_declaration && !ps.block_init) { 558 559 for (const char *p = inp.s; p < inp.e;) 560 if (*p++ == ')' && (*p == ';' || *p == ',')) 561 goto no_function_definition; 562 563 strncpy(ps.procname, token.s, sizeof ps.procname - 1); 564 if (ps.in_decl) 565 ps.in_parameter_declaration = true; 566 return lsym_funcname; 567 no_function_definition:; 568 569 } else if (probably_typename()) { 570 ps.curr_keyword = kw_type; 571 ps.next_unary = true; 572 return lsym_type; 573 } 574 575 if (ps.prev_token == lsym_type) /* if this is a declared variable, 576 * then following sign is unary */ 577 ps.next_unary = true; /* will make "int a -1" work */ 578 579 return lsym_ident; /* the ident is not in the list */ 580 } 581 582 /* Reads the next token, placing it in the global variable "token". */ 583 lexer_symbol 584 lexi(void) 585 { 586 token.e = token.s; 587 ps.curr_col_1 = ps.curr_newline; 588 ps.curr_newline = false; 589 ps.prev_keyword = ps.curr_keyword; 590 ps.curr_keyword = kw_0; 591 592 while (ch_isblank(*inp.s)) { 593 ps.curr_col_1 = false; 594 inbuf_skip(); 595 } 596 597 lexer_symbol alnum_lsym = lexi_alnum(); 598 if (alnum_lsym != lsym_eof) 599 return lexi_end(alnum_lsym); 600 601 /* Scan a non-alphanumeric token */ 602 603 check_size_token(3); /* for things like "<<=" */ 604 *token.e++ = inbuf_next(); 605 *token.e = '\0'; 606 607 lexer_symbol lsym; 608 bool unary_delim = false; /* whether the current token forces a 609 * following operator to be unary */ 610 611 switch (*token.s) { 612 case '\n': 613 unary_delim = ps.next_unary; 614 ps.curr_newline = true; 615 /* if data has been exhausted, the newline is a dummy. */ 616 lsym = had_eof ? lsym_eof : lsym_newline; 617 break; 618 619 case '\'': 620 case '"': 621 lex_char_or_string(); 622 lsym = lsym_ident; 623 break; 624 625 case '(': 626 case '[': 627 unary_delim = true; 628 lsym = lsym_lparen_or_lbracket; 629 break; 630 631 case ')': 632 case ']': 633 lsym = lsym_rparen_or_rbracket; 634 break; 635 636 case '#': 637 unary_delim = ps.next_unary; 638 lsym = lsym_preprocessing; 639 break; 640 641 case '?': 642 unary_delim = true; 643 lsym = lsym_question; 644 break; 645 646 case ':': 647 lsym = lsym_colon; 648 unary_delim = true; 649 break; 650 651 case ';': 652 unary_delim = true; 653 lsym = lsym_semicolon; 654 break; 655 656 case '{': 657 unary_delim = true; 658 lsym = lsym_lbrace; 659 break; 660 661 case '}': 662 unary_delim = true; 663 lsym = lsym_rbrace; 664 break; 665 666 case '\f': 667 unary_delim = ps.next_unary; 668 ps.curr_newline = true; 669 lsym = lsym_form_feed; 670 break; 671 672 case ',': 673 unary_delim = true; 674 lsym = lsym_comma; 675 break; 676 677 case '.': 678 unary_delim = false; 679 lsym = lsym_period; 680 break; 681 682 case '-': 683 case '+': 684 lsym = ps.next_unary ? lsym_unary_op : lsym_binary_op; 685 unary_delim = true; 686 687 if (*inp.s == token.s[0]) { /* ++, -- */ 688 *token.e++ = *inp.s++; 689 if (ps.prev_token == lsym_ident || 690 ps.prev_token == lsym_rparen_or_rbracket) { 691 lsym = ps.next_unary ? lsym_unary_op : lsym_postfix_op; 692 unary_delim = false; 693 } 694 695 } else if (*inp.s == '=') { /* += */ 696 *token.e++ = *inp.s++; 697 698 } else if (*inp.s == '>') { /* -> */ 699 *token.e++ = *inp.s++; 700 unary_delim = false; 701 lsym = lsym_unary_op; 702 ps.want_blank = false; 703 } 704 break; 705 706 case '=': 707 if (ps.init_or_struct) 708 ps.block_init = true; 709 if (*inp.s == '=') { /* == */ 710 *token.e++ = *inp.s++; 711 *token.e = '\0'; 712 } 713 lsym = lsym_binary_op; 714 unary_delim = true; 715 break; 716 717 case '>': 718 case '<': 719 case '!': /* ops like <, <<, <=, !=, etc */ 720 if (*inp.s == '>' || *inp.s == '<' || *inp.s == '=') 721 *token.e++ = inbuf_next(); 722 if (*inp.s == '=') 723 *token.e++ = *inp.s++; 724 lsym = ps.next_unary ? lsym_unary_op : lsym_binary_op; 725 unary_delim = true; 726 break; 727 728 case '*': 729 unary_delim = true; 730 if (!ps.next_unary) { 731 if (*inp.s == '=') 732 *token.e++ = *inp.s++; 733 lsym = lsym_binary_op; 734 break; 735 } 736 737 while (*inp.s == '*' || isspace((unsigned char)*inp.s)) { 738 if (*inp.s == '*') 739 token_add_char('*'); 740 inbuf_skip(); 741 } 742 743 if (ps.in_decl) { 744 char *tp = inp.s; 745 746 while (isalpha((unsigned char)*tp) || 747 isspace((unsigned char)*tp)) { 748 if (++tp >= inp.e) 749 inbuf_read_line(); 750 } 751 if (*tp == '(') 752 ps.procname[0] = ' '; 753 } 754 755 lsym = lsym_unary_op; 756 break; 757 758 default: 759 if (token.s[0] == '/' && (*inp.s == '*' || *inp.s == '/')) { 760 /* it is start of comment */ 761 *token.e++ = inbuf_next(); 762 763 lsym = lsym_comment; 764 unary_delim = ps.next_unary; 765 break; 766 } 767 768 while (token.e[-1] == *inp.s || *inp.s == '=') { 769 /* handle '||', '&&', etc., and also things as in 'int *****i' */ 770 token_add_char(inbuf_next()); 771 } 772 773 lsym = ps.next_unary ? lsym_unary_op : lsym_binary_op; 774 unary_delim = true; 775 } 776 777 if (inp.s >= inp.e) /* check for input buffer empty */ 778 inbuf_read_line(); 779 780 ps.next_unary = unary_delim; 781 782 check_size_token(1); 783 *token.e = '\0'; 784 785 return lexi_end(lsym); 786 } 787 788 void 789 add_typename(const char *name) 790 { 791 if (typenames.len >= typenames.cap) { 792 typenames.cap = 16 + 2 * typenames.cap; 793 typenames.items = xrealloc(typenames.items, 794 sizeof(typenames.items[0]) * typenames.cap); 795 } 796 797 int pos = bsearch_typenames(name); 798 if (pos >= 0) 799 return; /* already in the list */ 800 801 pos = -(pos + 1); 802 memmove(typenames.items + pos + 1, typenames.items + pos, 803 sizeof(typenames.items[0]) * (typenames.len++ - (unsigned)pos)); 804 typenames.items[pos] = xstrdup(name); 805 } 806
Indexes created Sun Sep 28 12:09:53 GMT 2025