lexi.c revision 1.121
1 /* $NetBSD: lexi.c,v 1.121 2021/10/31 10:09:43 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 1.121 2021/10/31 10:09:43 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 "offsetof", 241 "ident", 242 "funcname", 243 "do", 244 "else", 245 "for", 246 "if", 247 "switch", 248 "while", 249 }; 250 251 return name[sym]; 252 } 253 254 static const char * 255 kw_name(enum keyword_kind kw) 256 { 257 static const char *const name[] = { 258 "0", 259 "offsetof", 260 "sizeof", 261 "tag", 262 "type", 263 "for", 264 "if", 265 "while", 266 "do", 267 "else", 268 "switch", 269 "case_or_default", 270 "storage_class", 271 "typedef", 272 "other", 273 }; 274 275 return name[kw]; 276 } 277 278 static void 279 debug_print_buf(const char *name, const struct buffer *buf) 280 { 281 if (buf->s < buf->e) { 282 debug_printf("%s ", name); 283 debug_vis_range("\"", buf->s, buf->e, "\"\n"); 284 } 285 } 286 287 #define debug_ps_bool(name) \ 288 if (ps.name != prev_ps.name) \ 289 debug_println("[%c] ps." #name, ps.name ? 'x' : ' ') 290 #define debug_ps_int(name) \ 291 if (ps.name != prev_ps.name) \ 292 debug_println("%3d ps." #name, ps.name) 293 #define debug_ps_keyword(name) \ 294 if (ps.name != kw_0) \ 295 debug_println(" ps." #name " = %s", kw_name(ps.name)) 296 297 static void 298 debug_lexi(lexer_symbol lsym) 299 { 300 /* 301 * Watch out for 'rolled back parser state' in the debug output; the 302 * differences around these are unreliable. 303 */ 304 static struct parser_state prev_ps; 305 306 debug_println(""); 307 debug_printf("line %d: %s", line_no, lsym_name(lsym)); 308 debug_vis_range(" \"", token.s, token.e, "\"\n"); 309 debug_print_buf("label", &lab); 310 debug_print_buf("code", &code); 311 debug_print_buf("comment", &com); 312 313 // prev_token 314 debug_ps_keyword(prev_keyword); 315 debug_ps_keyword(curr_keyword); 316 debug_ps_bool(curr_newline); 317 debug_ps_bool(curr_col_1); 318 debug_ps_bool(next_unary); 319 // procname 320 debug_ps_bool(want_blank); 321 debug_ps_int(paren_level); 322 debug_ps_int(p_l_follow); 323 // paren_indents 324 debug_ps_int(cast_mask); 325 debug_ps_int(not_cast_mask); 326 327 debug_ps_int(comment_delta); 328 debug_ps_int(n_comment_delta); 329 debug_ps_int(com_ind); 330 331 debug_ps_bool(block_init); 332 debug_ps_int(block_init_level); 333 debug_ps_bool(init_or_struct); 334 335 debug_ps_int(ind_level); 336 debug_ps_int(ind_level_follow); 337 338 debug_ps_int(decl_nest); 339 debug_ps_bool(decl_on_line); 340 debug_ps_bool(in_decl); 341 debug_ps_int(just_saw_decl); 342 debug_ps_bool(in_parameter_declaration); 343 debug_ps_bool(decl_indent_done); 344 345 debug_ps_bool(in_stmt); 346 debug_ps_bool(ind_stmt); 347 debug_ps_bool(is_case_label); 348 349 debug_ps_bool(search_stmt); 350 351 prev_ps = ps; 352 } 353 #endif 354 355 /* ARGSUSED */ 356 static lexer_symbol 357 lexi_end(lexer_symbol lsym) 358 { 359 #ifdef debug 360 debug_lexi(lsym); 361 #endif 362 return lsym; 363 } 364 365 static void 366 lex_number(void) 367 { 368 for (unsigned char s = 'A'; s != 'f' && s != 'i' && s != 'u';) { 369 unsigned char ch = (unsigned char)*inp.s; 370 if (ch >= array_length(lex_number_row) || lex_number_row[ch] == 0) 371 break; 372 373 unsigned char row = lex_number_row[ch]; 374 if (lex_number_state[row][s - 'A'] == ' ') { 375 /*- 376 * lex_number_state[0][s - 'A'] now indicates the type: 377 * f = floating, i = integer, u = unknown 378 */ 379 break; 380 } 381 382 s = lex_number_state[row][s - 'A']; 383 token_add_char(inbuf_next()); 384 } 385 } 386 387 static void 388 lex_word(void) 389 { 390 while (isalnum((unsigned char)*inp.s) || 391 *inp.s == '\\' || 392 *inp.s == '_' || *inp.s == '$') { 393 394 if (*inp.s == '\\') { 395 if (inp.s[1] == '\n') { 396 inp.s += 2; 397 if (inp.s >= inp.e) 398 inbuf_read_line(); 399 } else 400 break; 401 } 402 403 token_add_char(inbuf_next()); 404 } 405 } 406 407 static void 408 lex_char_or_string(void) 409 { 410 for (char delim = *token.s;;) { 411 if (*inp.s == '\n') { 412 diag(1, "Unterminated literal"); 413 return; 414 } 415 416 token_add_char(inbuf_next()); 417 if (token.e[-1] == delim) 418 return; 419 420 if (token.e[-1] == '\\') { 421 if (*inp.s == '\n') 422 ++line_no; 423 token_add_char(inbuf_next()); 424 } 425 } 426 } 427 428 /* Guess whether the current token is a declared type. */ 429 static bool 430 probably_typename(void) 431 { 432 if (ps.p_l_follow > 0) 433 return false; 434 if (ps.block_init || ps.in_stmt) 435 return false; 436 if (inp.s[0] == '*' && inp.s[1] != '=') 437 goto maybe; 438 if (isalpha((unsigned char)*inp.s)) 439 goto maybe; 440 return false; 441 maybe: 442 return ps.prev_token == lsym_semicolon || 443 ps.prev_token == lsym_lbrace || 444 ps.prev_token == lsym_rbrace; 445 } 446 447 static int 448 bsearch_typenames(const char *key) 449 { 450 const char **arr = typenames.items; 451 int lo = 0; 452 int hi = (int)typenames.len - 1; 453 454 while (lo <= hi) { 455 int mid = (int)((unsigned)(lo + hi) >> 1); 456 int cmp = strcmp(arr[mid], key); 457 if (cmp < 0) 458 lo = mid + 1; 459 else if (cmp > 0) 460 hi = mid - 1; 461 else 462 return mid; 463 } 464 return -(lo + 1); 465 } 466 467 static bool 468 is_typename(void) 469 { 470 if (opt.auto_typedefs && 471 token.e - token.s >= 2 && memcmp(token.e - 2, "_t", 2) == 0) 472 return true; 473 474 return bsearch_typenames(token.s) >= 0; 475 } 476 477 static int 478 cmp_keyword_by_name(const void *key, const void *elem) 479 { 480 return strcmp(key, ((const struct keyword *)elem)->name); 481 } 482 483 /* Read an alphanumeric token into 'token', or return end_of_file. */ 484 static lexer_symbol 485 lexi_alnum(void) 486 { 487 if (isdigit((unsigned char)*inp.s) || 488 (inp.s[0] == '.' && isdigit((unsigned char)inp.s[1]))) { 489 lex_number(); 490 } else if (isalnum((unsigned char)*inp.s) || 491 *inp.s == '_' || *inp.s == '$') { 492 lex_word(); 493 } else 494 return lsym_eof; /* just as a placeholder */ 495 496 *token.e = '\0'; 497 498 if (token.s[0] == 'L' && token.s[1] == '\0' && 499 (*inp.s == '"' || *inp.s == '\'')) 500 return lsym_string_prefix; 501 502 while (ch_isblank(inbuf_peek())) 503 inbuf_skip(); 504 505 if (ps.prev_token == lsym_tag && ps.p_l_follow == 0) { 506 ps.next_unary = true; 507 return lsym_type; 508 } 509 510 /* Operator after identifier is binary unless last token was 'struct'. */ 511 ps.next_unary = ps.prev_token == lsym_tag; 512 513 const struct keyword *kw = bsearch(token.s, keywords, 514 array_length(keywords), sizeof(keywords[0]), cmp_keyword_by_name); 515 if (kw == NULL) { 516 if (is_typename()) { 517 ps.curr_keyword = kw_type; 518 ps.next_unary = true; 519 goto found_typename; 520 } 521 522 } else { /* we have a keyword */ 523 ps.curr_keyword = kw->kind; 524 ps.next_unary = true; 525 526 /* INDENT OFF */ 527 switch (kw->kind) { 528 case kw_tag: 529 case kw_type: goto found_typename; 530 case kw_case_or_default: return lsym_case_label; 531 case kw_for: return lsym_for; 532 case kw_if: return lsym_if; 533 case kw_else: return lsym_else; 534 case kw_switch: return lsym_switch; 535 case kw_while: return lsym_while; 536 case kw_do: return lsym_do; 537 case kw_storage_class: return lsym_storage_class; 538 case kw_typedef: return lsym_typedef; 539 case kw_offsetof: return lsym_offsetof; 540 case kw_sizeof: return lsym_sizeof; 541 default: return lsym_ident; 542 } 543 /* INDENT ON */ 544 545 found_typename: 546 if (ps.p_l_follow > 0) { 547 /* inside parentheses: cast, param list, offsetof or sizeof */ 548 ps.cast_mask |= (1 << ps.p_l_follow) & ~ps.not_cast_mask; 549 } 550 if (ps.prev_token != lsym_period && ps.prev_token != lsym_unary_op) { 551 if (kw != NULL && kw->kind == kw_tag) 552 return lsym_tag; 553 if (ps.p_l_follow == 0) 554 return lsym_type; 555 } 556 } 557 558 if (*inp.s == '(' && ps.tos <= 1 && ps.ind_level == 0 && 559 !ps.in_parameter_declaration && !ps.block_init) { 560 561 for (const char *p = inp.s; p < inp.e;) 562 if (*p++ == ')' && (*p == ';' || *p == ',')) 563 goto no_function_definition; 564 565 strncpy(ps.procname, token.s, sizeof ps.procname - 1); 566 if (ps.in_decl) 567 ps.in_parameter_declaration = true; 568 return lsym_funcname; 569 no_function_definition:; 570 571 } else if (probably_typename()) { 572 ps.curr_keyword = kw_type; 573 ps.next_unary = true; 574 return lsym_type; 575 } 576 577 if (ps.prev_token == lsym_type) /* if this is a declared variable, 578 * then following sign is unary */ 579 ps.next_unary = true; /* will make "int a -1" work */ 580 581 return lsym_ident; /* the ident is not in the list */ 582 } 583 584 /* Reads the next token, placing it in the global variable "token". */ 585 lexer_symbol 586 lexi(void) 587 { 588 token.e = token.s; 589 ps.curr_col_1 = ps.curr_newline; 590 ps.curr_newline = false; 591 ps.prev_keyword = ps.curr_keyword; 592 ps.curr_keyword = kw_0; 593 594 while (ch_isblank(*inp.s)) { 595 ps.curr_col_1 = false; 596 inbuf_skip(); 597 } 598 599 lexer_symbol alnum_lsym = lexi_alnum(); 600 if (alnum_lsym != lsym_eof) 601 return lexi_end(alnum_lsym); 602 603 /* Scan a non-alphanumeric token */ 604 605 check_size_token(3); /* for things like "<<=" */ 606 *token.e++ = inbuf_next(); 607 *token.e = '\0'; 608 609 lexer_symbol lsym; 610 bool unary_delim = false; /* whether the current token forces a 611 * following operator to be unary */ 612 613 switch (*token.s) { 614 case '\n': 615 unary_delim = ps.next_unary; 616 ps.curr_newline = true; 617 /* if data has been exhausted, the newline is a dummy. */ 618 lsym = had_eof ? lsym_eof : lsym_newline; 619 break; 620 621 case '\'': 622 case '"': 623 lex_char_or_string(); 624 lsym = lsym_ident; 625 break; 626 627 case '(': 628 case '[': 629 unary_delim = true; 630 lsym = lsym_lparen_or_lbracket; 631 break; 632 633 case ')': 634 case ']': 635 lsym = lsym_rparen_or_rbracket; 636 break; 637 638 case '#': 639 unary_delim = ps.next_unary; 640 lsym = lsym_preprocessing; 641 break; 642 643 case '?': 644 unary_delim = true; 645 lsym = lsym_question; 646 break; 647 648 case ':': 649 lsym = lsym_colon; 650 unary_delim = true; 651 break; 652 653 case ';': 654 unary_delim = true; 655 lsym = lsym_semicolon; 656 break; 657 658 case '{': 659 unary_delim = true; 660 lsym = lsym_lbrace; 661 break; 662 663 case '}': 664 unary_delim = true; 665 lsym = lsym_rbrace; 666 break; 667 668 case '\f': 669 unary_delim = ps.next_unary; 670 ps.curr_newline = true; 671 lsym = lsym_form_feed; 672 break; 673 674 case ',': 675 unary_delim = true; 676 lsym = lsym_comma; 677 break; 678 679 case '.': 680 unary_delim = false; 681 lsym = lsym_period; 682 break; 683 684 case '-': 685 case '+': 686 lsym = ps.next_unary ? lsym_unary_op : lsym_binary_op; 687 unary_delim = true; 688 689 if (*inp.s == token.s[0]) { /* ++, -- */ 690 *token.e++ = *inp.s++; 691 if (ps.prev_token == lsym_ident || 692 ps.prev_token == lsym_rparen_or_rbracket) { 693 lsym = ps.next_unary ? lsym_unary_op : lsym_postfix_op; 694 unary_delim = false; 695 } 696 697 } else if (*inp.s == '=') { /* += */ 698 *token.e++ = *inp.s++; 699 700 } else if (*inp.s == '>') { /* -> */ 701 *token.e++ = *inp.s++; 702 unary_delim = false; 703 lsym = lsym_unary_op; 704 ps.want_blank = false; 705 } 706 break; 707 708 case '=': 709 if (ps.init_or_struct) 710 ps.block_init = true; 711 if (*inp.s == '=') { /* == */ 712 *token.e++ = *inp.s++; 713 *token.e = '\0'; 714 } 715 lsym = lsym_binary_op; 716 unary_delim = true; 717 break; 718 719 case '>': 720 case '<': 721 case '!': /* ops like <, <<, <=, !=, etc */ 722 if (*inp.s == '>' || *inp.s == '<' || *inp.s == '=') 723 *token.e++ = inbuf_next(); 724 if (*inp.s == '=') 725 *token.e++ = *inp.s++; 726 lsym = ps.next_unary ? lsym_unary_op : lsym_binary_op; 727 unary_delim = true; 728 break; 729 730 case '*': 731 unary_delim = true; 732 if (!ps.next_unary) { 733 if (*inp.s == '=') 734 *token.e++ = *inp.s++; 735 lsym = lsym_binary_op; 736 break; 737 } 738 739 while (*inp.s == '*' || isspace((unsigned char)*inp.s)) { 740 if (*inp.s == '*') 741 token_add_char('*'); 742 inbuf_skip(); 743 } 744 745 if (ps.in_decl) { 746 char *tp = inp.s; 747 748 while (isalpha((unsigned char)*tp) || 749 isspace((unsigned char)*tp)) { 750 if (++tp >= inp.e) 751 inbuf_read_line(); 752 } 753 if (*tp == '(') 754 ps.procname[0] = ' '; 755 } 756 757 lsym = lsym_unary_op; 758 break; 759 760 default: 761 if (token.s[0] == '/' && (*inp.s == '*' || *inp.s == '/')) { 762 /* it is start of comment */ 763 *token.e++ = inbuf_next(); 764 765 lsym = lsym_comment; 766 unary_delim = ps.next_unary; 767 break; 768 } 769 770 while (token.e[-1] == *inp.s || *inp.s == '=') { 771 /* handle '||', '&&', etc., and also things as in 'int *****i' */ 772 token_add_char(inbuf_next()); 773 } 774 775 lsym = ps.next_unary ? lsym_unary_op : lsym_binary_op; 776 unary_delim = true; 777 } 778 779 if (inp.s >= inp.e) /* check for input buffer empty */ 780 inbuf_read_line(); 781 782 ps.next_unary = unary_delim; 783 784 check_size_token(1); 785 *token.e = '\0'; 786 787 return lexi_end(lsym); 788 } 789 790 void 791 add_typename(const char *name) 792 { 793 if (typenames.len >= typenames.cap) { 794 typenames.cap = 16 + 2 * typenames.cap; 795 typenames.items = xrealloc(typenames.items, 796 sizeof(typenames.items[0]) * typenames.cap); 797 } 798 799 int pos = bsearch_typenames(name); 800 if (pos >= 0) 801 return; /* already in the list */ 802 803 pos = -(pos + 1); 804 memmove(typenames.items + pos + 1, typenames.items + pos, 805 sizeof(typenames.items[0]) * (typenames.len++ - (unsigned)pos)); 806 typenames.items[pos] = xstrdup(name); 807 } 808
Indexes created Mon Sep 29 21:09:56 GMT 2025