lexi.c revision 1.113
1 /* $NetBSD: lexi.c,v 1.113 2021/10/29 21:31:29 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 1.113 2021/10/29 21:31:29 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 <sys/param.h> 52 #include <assert.h> 53 #include <stdio.h> 54 #include <ctype.h> 55 #include <stdlib.h> 56 #include <string.h> 57 58 #include "indent.h" 59 60 /* must be sorted alphabetically, is used in binary search */ 61 static const struct keyword { 62 const char *name; 63 enum keyword_kind kind; 64 } keywords[] = { 65 {"_Bool", kw_type}, 66 {"_Complex", kw_type}, 67 {"_Imaginary", kw_type}, 68 {"auto", kw_storage_class}, 69 {"bool", kw_type}, 70 {"break", kw_jump}, 71 {"case", kw_case_or_default}, 72 {"char", kw_type}, 73 {"complex", kw_type}, 74 {"const", kw_type}, 75 {"continue", kw_jump}, 76 {"default", kw_case_or_default}, 77 {"do", kw_do}, 78 {"double", kw_type}, 79 {"else", kw_else}, 80 {"enum", kw_struct_or_union_or_enum}, 81 {"extern", kw_storage_class}, 82 {"float", kw_type}, 83 {"for", kw_for}, 84 {"goto", kw_jump}, 85 {"if", kw_if}, 86 {"imaginary", kw_type}, 87 {"inline", kw_inline_or_restrict}, 88 {"int", kw_type}, 89 {"long", kw_type}, 90 {"offsetof", kw_offsetof}, 91 {"register", kw_storage_class}, 92 {"restrict", kw_inline_or_restrict}, 93 {"return", kw_jump}, 94 {"short", kw_type}, 95 {"signed", kw_type}, 96 {"sizeof", kw_sizeof}, 97 {"static", kw_storage_class}, 98 {"struct", kw_struct_or_union_or_enum}, 99 {"switch", kw_switch}, 100 {"typedef", kw_typedef}, 101 {"union", kw_struct_or_union_or_enum}, 102 {"unsigned", kw_type}, 103 {"void", kw_type}, 104 {"volatile", kw_type}, 105 {"while", kw_while} 106 }; 107 108 static struct { 109 const char **items; 110 unsigned int len; 111 unsigned int cap; 112 } typenames; 113 114 /* 115 * The transition table below was rewritten by hand from lx's output, given 116 * the following definitions. lx is Katherine Flavel's lexer generator. 117 * 118 * O = /[0-7]/; D = /[0-9]/; NZ = /[1-9]/; 119 * H = /[a-f0-9]/i; B = /[0-1]/; HP = /0x/i; 120 * BP = /0b/i; E = /e[+\-]?/i D+; P = /p[+\-]?/i D+; 121 * FS = /[fl]/i; IS = /u/i /(l|L|ll|LL)/? | /(l|L|ll|LL)/ /u/i?; 122 * 123 * D+ E FS? -> $float; 124 * D* "." D+ E? FS? -> $float; 125 * D+ "." E? FS? -> $float; HP H+ IS? -> $int; 126 * HP H+ P FS? -> $float; NZ D* IS? -> $int; 127 * HP H* "." H+ P FS? -> $float; "0" O* IS? -> $int; 128 * HP H+ "." P FS -> $float; BP B+ IS? -> $int; 129 */ 130 /* INDENT OFF */ 131 static const unsigned char lex_number_state[][26] = { 132 /* examples: 133 00 134 s 0xx 135 t 00xaa 136 a 11 101100xxa.. 137 r 11ee0001101lbuuxx.a.pp 138 t.01.e+008bLuxll0Ll.aa.p+0 139 states: ABCDEFGHIJKLMNOPQRSTUVWXYZ */ 140 [0] = "uuiifuufiuuiiuiiiiiuiuuuuu", /* (other) */ 141 [1] = "CEIDEHHHIJQ U Q VUVVZZZ", /* 0 */ 142 [2] = "DEIDEHHHIJQ U Q VUVVZZZ", /* 1 */ 143 [3] = "DEIDEHHHIJ U VUVVZZZ", /* 2 3 4 5 6 7 */ 144 [4] = "DEJDEHHHJJ U VUVVZZZ", /* 8 9 */ 145 [5] = " U VUVV ", /* A a C c D d */ 146 [6] = " K U VUVV ", /* B b */ 147 [7] = " FFF FF U VUVV ", /* E e */ 148 [8] = " f f U VUVV f", /* F f */ 149 [9] = " LLf fL PR Li L f", /* L */ 150 [10] = " OOf fO S P O i O f", /* l */ 151 [11] = " FFX ", /* P p */ 152 [12] = " MM M i iiM M ", /* U u */ 153 [13] = " N ", /* X x */ 154 [14] = " G Y ", /* + - */ 155 [15] = "B EE EE T W ", /* . */ 156 /* ABCDEFGHIJKLMNOPQRSTUVWXYZ */ 157 }; 158 /* INDENT ON */ 159 160 static const uint8_t lex_number_row[] = { 161 ['0'] = 1, 162 ['1'] = 2, 163 ['2'] = 3, ['3'] = 3, ['4'] = 3, ['5'] = 3, ['6'] = 3, ['7'] = 3, 164 ['8'] = 4, ['9'] = 4, 165 ['A'] = 5, ['a'] = 5, ['C'] = 5, ['c'] = 5, ['D'] = 5, ['d'] = 5, 166 ['B'] = 6, ['b'] = 6, 167 ['E'] = 7, ['e'] = 7, 168 ['F'] = 8, ['f'] = 8, 169 ['L'] = 9, 170 ['l'] = 10, 171 ['P'] = 11, ['p'] = 11, 172 ['U'] = 12, ['u'] = 12, 173 ['X'] = 13, ['x'] = 13, 174 ['+'] = 14, ['-'] = 14, 175 ['.'] = 15, 176 }; 177 178 static char 179 inbuf_peek(void) 180 { 181 return *inp.s; 182 } 183 184 void 185 inbuf_skip(void) 186 { 187 inp.s++; 188 if (inp.s >= inp.e) 189 inbuf_read_line(); 190 } 191 192 char 193 inbuf_next(void) 194 { 195 char ch = inbuf_peek(); 196 inbuf_skip(); 197 return ch; 198 } 199 200 static void 201 check_size_token(size_t desired_size) 202 { 203 if (token.e + desired_size >= token.l) 204 buf_expand(&token, desired_size); 205 } 206 207 static void 208 token_add_char(char ch) 209 { 210 check_size_token(1); 211 *token.e++ = ch; 212 } 213 214 static int 215 cmp_keyword_by_name(const void *key, const void *elem) 216 { 217 return strcmp(key, ((const struct keyword *)elem)->name); 218 } 219 220 #ifdef debug 221 static const char * 222 lsym_name(lexer_symbol sym) 223 { 224 static const char *const name[] = { 225 "eof", 226 "preprocessing", 227 "newline", 228 "form_feed", 229 "comment", 230 "lparen_or_lbracket", 231 "rparen_or_rbracket", 232 "lbrace", 233 "rbrace", 234 "period", 235 "unary_op", 236 "binary_op", 237 "postfix_op", 238 "question", 239 "colon", 240 "comma", 241 "semicolon", 242 "typedef", 243 "storage_class", 244 "type", 245 "tag", 246 "case_label", 247 "string_prefix", 248 "ident", 249 "funcname", 250 "do", 251 "else", 252 "for", 253 "if", 254 "switch", 255 "while", 256 }; 257 258 assert(array_length(name) == (int)lsym_while + 1); 259 260 return name[sym]; 261 } 262 263 static const char * 264 kw_name(enum keyword_kind kw) 265 { 266 static const char *name[] = { 267 "0", 268 "offsetof", 269 "sizeof", 270 "struct_or_union_or_enum", 271 "type", 272 "for", 273 "if", 274 "while", 275 "do", 276 "else", 277 "switch", 278 "case_or_default", 279 "jump", 280 "storage_class", 281 "typedef", 282 "inline_or_restrict", 283 }; 284 285 return name[kw]; 286 } 287 288 static void 289 debug_print_buf(const char *name, const struct buffer *buf) 290 { 291 if (buf->s < buf->e) { 292 debug_printf("%s ", name); 293 debug_vis_range("\"", buf->s, buf->e, "\"\n"); 294 } 295 } 296 297 #define debug_ps_bool(name) \ 298 if (ps.name != prev_ps.name) \ 299 debug_println("[%c] ps." #name, ps.name ? 'x' : ' ') 300 #define debug_ps_int(name) \ 301 if (ps.name != prev_ps.name) \ 302 debug_println("%3d ps." #name, ps.name) 303 #define debug_ps_keyword(name) \ 304 if (ps.name != kw_0) \ 305 debug_println(" ps." #name " = %s", kw_name(ps.name)) 306 307 static void 308 debug_lexi(lexer_symbol lsym) 309 { 310 /* 311 * Watch out for 'rolled back parser state' in the debug output; the 312 * differences around these are unreliable. 313 */ 314 static struct parser_state prev_ps; 315 316 debug_println(""); 317 debug_printf("line %d\n", line_no); 318 debug_print_buf("label", &lab); 319 debug_print_buf("code", &code); 320 debug_print_buf("comment", &com); 321 debug_printf("lexi returns '%s'", lsym_name(lsym)); 322 debug_vis_range(" \"", token.s, token.e, "\"\n"); 323 324 // prev_token 325 debug_ps_bool(prev_newline); 326 debug_ps_bool(prev_col_1); 327 debug_ps_keyword(prev_keyword); 328 debug_ps_keyword(curr_keyword); 329 debug_ps_bool(next_unary); 330 // procname 331 debug_ps_bool(want_blank); 332 debug_ps_int(paren_level); 333 debug_ps_int(p_l_follow); 334 // paren_indents 335 debug_ps_int(cast_mask); 336 debug_ps_int(not_cast_mask); 337 338 debug_ps_int(comment_delta); 339 debug_ps_int(n_comment_delta); 340 debug_ps_int(com_ind); 341 342 debug_ps_bool(block_init); 343 debug_ps_int(block_init_level); 344 debug_ps_bool(init_or_struct); 345 346 debug_ps_int(ind_level); 347 debug_ps_int(ind_level_follow); 348 349 debug_ps_int(decl_nest); 350 debug_ps_bool(decl_on_line); 351 debug_ps_bool(in_decl); 352 debug_ps_int(just_saw_decl); 353 debug_ps_bool(in_parameter_declaration); 354 debug_ps_bool(decl_indent_done); 355 356 debug_ps_bool(in_stmt); 357 debug_ps_bool(ind_stmt); 358 debug_ps_bool(is_case_label); 359 360 debug_ps_bool(search_stmt); 361 362 prev_ps = ps; 363 } 364 #endif 365 366 /* ARGSUSED */ 367 static lexer_symbol 368 lexi_end(lexer_symbol lsym) 369 { 370 #ifdef debug 371 debug_lexi(lsym); 372 #endif 373 return lsym; 374 } 375 376 static void 377 lex_number(void) 378 { 379 for (uint8_t s = 'A'; s != 'f' && s != 'i' && s != 'u';) { 380 uint8_t ch = (uint8_t)*inp.s; 381 if (ch >= array_length(lex_number_row) || lex_number_row[ch] == 0) 382 break; 383 384 uint8_t row = lex_number_row[ch]; 385 if (lex_number_state[row][s - 'A'] == ' ') { 386 /*- 387 * lex_number_state[0][s - 'A'] now indicates the type: 388 * f = floating, i = integer, u = unknown 389 */ 390 break; 391 } 392 393 s = lex_number_state[row][s - 'A']; 394 token_add_char(inbuf_next()); 395 } 396 } 397 398 static void 399 lex_word(void) 400 { 401 while (isalnum((unsigned char)*inp.s) || 402 *inp.s == '\\' || 403 *inp.s == '_' || *inp.s == '$') { 404 405 if (*inp.s == '\\') { 406 if (inp.s[1] == '\n') { 407 inp.s += 2; 408 if (inp.s >= inp.e) 409 inbuf_read_line(); 410 } else 411 break; 412 } 413 414 token_add_char(inbuf_next()); 415 } 416 } 417 418 static void 419 lex_char_or_string(void) 420 { 421 for (char delim = *token.s;;) { 422 if (*inp.s == '\n') { 423 diag(1, "Unterminated literal"); 424 return; 425 } 426 427 token_add_char(inbuf_next()); 428 if (token.e[-1] == delim) 429 return; 430 431 if (token.e[-1] == '\\') { 432 if (*inp.s == '\n') 433 ++line_no; 434 token_add_char(inbuf_next()); 435 } 436 } 437 } 438 439 /* Guess whether the current token is a declared type. */ 440 static bool 441 probably_typename(void) 442 { 443 if (ps.p_l_follow > 0) 444 return false; 445 if (ps.block_init || ps.in_stmt) 446 return false; 447 if (inp.s[0] == '*' && inp.s[1] != '=') 448 goto maybe; 449 if (isalpha((unsigned char)*inp.s)) 450 goto maybe; 451 return false; 452 maybe: 453 return ps.prev_token == lsym_semicolon || 454 ps.prev_token == lsym_lbrace || 455 ps.prev_token == lsym_rbrace; 456 } 457 458 static int 459 bsearch_typenames(const char *key) 460 { 461 const char **arr = typenames.items; 462 int lo = 0; 463 int hi = (int)typenames.len - 1; 464 465 while (lo <= hi) { 466 int mid = (int)((unsigned)(lo + hi) >> 1); 467 int cmp = strcmp(arr[mid], key); 468 if (cmp < 0) 469 lo = mid + 1; 470 else if (cmp > 0) 471 hi = mid - 1; 472 else 473 return mid; 474 } 475 return -(lo + 1); 476 } 477 478 static bool 479 is_typename(void) 480 { 481 if (opt.auto_typedefs && 482 token.e - token.s >= 2 && memcmp(token.e - 2, "_t", 2) == 0) 483 return true; 484 485 return bsearch_typenames(token.s) >= 0; 486 } 487 488 /* Read an alphanumeric token into 'token', or return end_of_file. */ 489 static lexer_symbol 490 lexi_alnum(void) 491 { 492 if (isdigit((unsigned char)*inp.s) || 493 (inp.s[0] == '.' && isdigit((unsigned char)inp.s[1]))) { 494 lex_number(); 495 } else if (isalnum((unsigned char)*inp.s) || 496 *inp.s == '_' || *inp.s == '$') { 497 lex_word(); 498 } else 499 return lsym_eof; /* just as a placeholder */ 500 501 *token.e = '\0'; 502 503 if (token.s[0] == 'L' && token.s[1] == '\0' && 504 (*inp.s == '"' || *inp.s == '\'')) 505 return lsym_string_prefix; 506 507 while (ch_isblank(inbuf_peek())) 508 inbuf_skip(); 509 510 if (ps.prev_token == lsym_tag && ps.p_l_follow == 0) { 511 ps.next_unary = true; 512 return lsym_type; 513 } 514 515 /* Operator after identifier is binary unless last token was 'struct'. */ 516 ps.next_unary = ps.prev_token == lsym_tag; 517 518 const struct keyword *kw = bsearch(token.s, keywords, 519 array_length(keywords), sizeof(keywords[0]), cmp_keyword_by_name); 520 if (kw == NULL) { 521 if (is_typename()) { 522 ps.curr_keyword = kw_type; 523 ps.next_unary = true; 524 goto found_typename; 525 } 526 527 } else { /* we have a keyword */ 528 ps.curr_keyword = kw->kind; 529 ps.next_unary = true; 530 531 switch (kw->kind) { 532 case kw_switch: 533 return lsym_switch; 534 535 case kw_case_or_default: 536 return lsym_case_label; 537 538 case kw_struct_or_union_or_enum: 539 case kw_type: 540 found_typename: 541 if (ps.p_l_follow > 0) { 542 /* inside parentheses: cast, param list, offsetof or sizeof */ 543 ps.cast_mask |= (1 << ps.p_l_follow) & ~ps.not_cast_mask; 544 } 545 if (ps.prev_token == lsym_period || 546 ps.prev_token == lsym_unary_op) 547 break; 548 if (kw != NULL && kw->kind == kw_struct_or_union_or_enum) 549 return lsym_tag; 550 if (ps.p_l_follow > 0) 551 break; 552 return lsym_type; 553 554 case kw_for: 555 return lsym_for; 556 557 case kw_if: 558 return lsym_if; 559 560 case kw_while: 561 return lsym_while; 562 563 case kw_do: 564 return lsym_do; 565 566 case kw_else: 567 return lsym_else; 568 569 case kw_storage_class: 570 return lsym_storage_class; 571 572 case kw_typedef: 573 return lsym_typedef; 574 575 default: /* all others are treated like any other 576 * identifier */ 577 return lsym_ident; 578 } 579 } 580 581 if (*inp.s == '(' && ps.tos <= 1 && ps.ind_level == 0 && 582 !ps.in_parameter_declaration && !ps.block_init) { 583 584 for (const char *p = inp.s; p < inp.e;) 585 if (*p++ == ')' && (*p == ';' || *p == ',')) 586 goto not_proc; 587 588 strncpy(ps.procname, token.s, sizeof ps.procname - 1); 589 if (ps.in_decl) 590 ps.in_parameter_declaration = true; 591 return lsym_funcname; 592 not_proc:; 593 594 } else if (probably_typename()) { 595 ps.curr_keyword = kw_type; 596 ps.next_unary = true; 597 return lsym_type; 598 } 599 600 if (ps.prev_token == lsym_type) /* if this is a declared variable, 601 * then following sign is unary */ 602 ps.next_unary = true; /* will make "int a -1" work */ 603 604 return lsym_ident; /* the ident is not in the list */ 605 } 606 607 /* Reads the next token, placing it in the global variable "token". */ 608 lexer_symbol 609 lexi(void) 610 { 611 token.e = token.s; 612 ps.prev_col_1 = ps.prev_newline; 613 ps.prev_newline = false; 614 ps.prev_keyword = ps.curr_keyword; 615 ps.curr_keyword = kw_0; 616 617 while (ch_isblank(*inp.s)) { 618 ps.prev_col_1 = false; 619 inbuf_skip(); 620 } 621 622 lexer_symbol alnum_lsym = lexi_alnum(); 623 if (alnum_lsym != lsym_eof) 624 return lexi_end(alnum_lsym); 625 626 /* Scan a non-alphanumeric token */ 627 628 check_size_token(3); /* for things like "<<=" */ 629 *token.e++ = inbuf_next(); 630 *token.e = '\0'; 631 632 lexer_symbol lsym; 633 bool unary_delim = false; /* whether the current token forces a 634 * following operator to be unary */ 635 636 switch (*token.s) { 637 case '\n': 638 unary_delim = ps.next_unary; 639 ps.prev_newline = true; 640 /* if data has been exhausted, the newline is a dummy. */ 641 lsym = had_eof ? lsym_eof : lsym_newline; 642 break; 643 644 case '\'': 645 case '"': 646 lex_char_or_string(); 647 lsym = lsym_ident; 648 break; 649 650 case '(': 651 case '[': 652 unary_delim = true; 653 lsym = lsym_lparen_or_lbracket; 654 break; 655 656 case ')': 657 case ']': 658 lsym = lsym_rparen_or_rbracket; 659 break; 660 661 case '#': 662 unary_delim = ps.next_unary; 663 lsym = lsym_preprocessing; 664 break; 665 666 case '?': 667 unary_delim = true; 668 lsym = lsym_question; 669 break; 670 671 case ':': 672 lsym = lsym_colon; 673 unary_delim = true; 674 break; 675 676 case ';': 677 unary_delim = true; 678 lsym = lsym_semicolon; 679 break; 680 681 case '{': 682 unary_delim = true; 683 lsym = lsym_lbrace; 684 break; 685 686 case '}': 687 unary_delim = true; 688 lsym = lsym_rbrace; 689 break; 690 691 case '\f': 692 unary_delim = ps.next_unary; 693 ps.prev_newline = true; 694 lsym = lsym_form_feed; 695 break; 696 697 case ',': 698 unary_delim = true; 699 lsym = lsym_comma; 700 break; 701 702 case '.': 703 unary_delim = false; 704 lsym = lsym_period; 705 break; 706 707 case '-': 708 case '+': 709 lsym = ps.next_unary ? lsym_unary_op : lsym_binary_op; 710 unary_delim = true; 711 712 if (*inp.s == token.s[0]) { /* ++, -- */ 713 *token.e++ = *inp.s++; 714 if (ps.prev_token == lsym_ident || 715 ps.prev_token == lsym_rparen_or_rbracket) { 716 lsym = ps.next_unary ? lsym_unary_op : lsym_postfix_op; 717 unary_delim = false; 718 } 719 720 } else if (*inp.s == '=') { /* += */ 721 *token.e++ = *inp.s++; 722 723 } else if (*inp.s == '>') { /* -> */ 724 *token.e++ = *inp.s++; 725 unary_delim = false; 726 lsym = lsym_unary_op; 727 ps.want_blank = false; 728 } 729 break; 730 731 case '=': 732 if (ps.init_or_struct) 733 ps.block_init = true; 734 if (*inp.s == '=') { /* == */ 735 *token.e++ = *inp.s++; 736 *token.e = '\0'; 737 } 738 lsym = lsym_binary_op; 739 unary_delim = true; 740 break; 741 742 case '>': 743 case '<': 744 case '!': /* ops like <, <<, <=, !=, etc */ 745 if (*inp.s == '>' || *inp.s == '<' || *inp.s == '=') 746 *token.e++ = inbuf_next(); 747 if (*inp.s == '=') 748 *token.e++ = *inp.s++; 749 lsym = ps.next_unary ? lsym_unary_op : lsym_binary_op; 750 unary_delim = true; 751 break; 752 753 case '*': 754 unary_delim = true; 755 if (!ps.next_unary) { 756 if (*inp.s == '=') 757 *token.e++ = *inp.s++; 758 lsym = lsym_binary_op; 759 break; 760 } 761 762 while (*inp.s == '*' || isspace((unsigned char)*inp.s)) { 763 if (*inp.s == '*') 764 token_add_char('*'); 765 inbuf_skip(); 766 } 767 768 if (ps.in_decl) { 769 char *tp = inp.s; 770 771 while (isalpha((unsigned char)*tp) || 772 isspace((unsigned char)*tp)) { 773 if (++tp >= inp.e) 774 inbuf_read_line(); 775 } 776 if (*tp == '(') 777 ps.procname[0] = ' '; 778 } 779 780 lsym = lsym_unary_op; 781 break; 782 783 default: 784 if (token.s[0] == '/' && (*inp.s == '*' || *inp.s == '/')) { 785 /* it is start of comment */ 786 *token.e++ = inbuf_next(); 787 788 lsym = lsym_comment; 789 unary_delim = ps.next_unary; 790 break; 791 } 792 793 while (token.e[-1] == *inp.s || *inp.s == '=') { 794 /* handle '||', '&&', etc., and also things as in 'int *****i' */ 795 token_add_char(inbuf_next()); 796 } 797 798 lsym = ps.next_unary ? lsym_unary_op : lsym_binary_op; 799 unary_delim = true; 800 } 801 802 if (inp.s >= inp.e) /* check for input buffer empty */ 803 inbuf_read_line(); 804 805 ps.next_unary = unary_delim; 806 807 check_size_token(1); 808 *token.e = '\0'; 809 810 return lexi_end(lsym); 811 } 812 813 void 814 add_typename(const char *name) 815 { 816 if (typenames.len >= typenames.cap) { 817 typenames.cap = 16 + 2 * typenames.cap; 818 typenames.items = xrealloc(typenames.items, 819 sizeof(typenames.items[0]) * typenames.cap); 820 } 821 822 int pos = bsearch_typenames(name); 823 if (pos >= 0) 824 return; /* already in the list */ 825 826 pos = -(pos + 1); 827 memmove(typenames.items + pos + 1, typenames.items + pos, 828 sizeof(typenames.items[0]) * (typenames.len++ - (unsigned)pos)); 829 typenames.items[pos] = xstrdup(name); 830 } 831
Indexes created Sun Sep 28 16:09:52 GMT 2025