lexi.c revision 1.115
1 /* $NetBSD: lexi.c,v 1.115 2021/10/30 22:15:51 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 1.115 2021/10/30 22:15:51 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_jump}, 68 {"case", kw_case_or_default}, 69 {"char", kw_type}, 70 {"complex", kw_type}, 71 {"const", kw_type}, 72 {"continue", kw_jump}, 73 {"default", kw_case_or_default}, 74 {"do", kw_do}, 75 {"double", kw_type}, 76 {"else", kw_else}, 77 {"enum", kw_struct_or_union_or_enum}, 78 {"extern", kw_storage_class}, 79 {"float", kw_type}, 80 {"for", kw_for}, 81 {"goto", kw_jump}, 82 {"if", kw_if}, 83 {"imaginary", kw_type}, 84 {"inline", kw_inline_or_restrict}, 85 {"int", kw_type}, 86 {"long", kw_type}, 87 {"offsetof", kw_offsetof}, 88 {"register", kw_storage_class}, 89 {"restrict", kw_inline_or_restrict}, 90 {"return", kw_jump}, 91 {"short", kw_type}, 92 {"signed", kw_type}, 93 {"sizeof", kw_sizeof}, 94 {"static", kw_storage_class}, 95 {"struct", kw_struct_or_union_or_enum}, 96 {"switch", kw_switch}, 97 {"typedef", kw_typedef}, 98 {"union", kw_struct_or_union_or_enum}, 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 "struct_or_union_or_enum", 260 "type", 261 "for", 262 "if", 263 "while", 264 "do", 265 "else", 266 "switch", 267 "case_or_default", 268 "jump", 269 "storage_class", 270 "typedef", 271 "inline_or_restrict", 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\n", line_no); 307 debug_print_buf("label", &lab); 308 debug_print_buf("code", &code); 309 debug_print_buf("comment", &com); 310 debug_printf("lexi: %s", lsym_name(lsym)); 311 debug_vis_range(" \"", token.s, token.e, "\"\n"); 312 313 // prev_token 314 debug_ps_bool(prev_newline); 315 debug_ps_bool(prev_col_1); 316 debug_ps_keyword(prev_keyword); 317 debug_ps_keyword(curr_keyword); 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 switch (kw->kind) { 527 case kw_switch: 528 return lsym_switch; 529 530 case kw_case_or_default: 531 return lsym_case_label; 532 533 case kw_struct_or_union_or_enum: 534 case kw_type: 535 found_typename: 536 if (ps.p_l_follow > 0) { 537 /* inside parentheses: cast, param list, offsetof or sizeof */ 538 ps.cast_mask |= (1 << ps.p_l_follow) & ~ps.not_cast_mask; 539 } 540 if (ps.prev_token == lsym_period || 541 ps.prev_token == lsym_unary_op) 542 break; 543 if (kw != NULL && kw->kind == kw_struct_or_union_or_enum) 544 return lsym_tag; 545 if (ps.p_l_follow > 0) 546 break; 547 return lsym_type; 548 549 case kw_for: 550 return lsym_for; 551 552 case kw_if: 553 return lsym_if; 554 555 case kw_while: 556 return lsym_while; 557 558 case kw_do: 559 return lsym_do; 560 561 case kw_else: 562 return lsym_else; 563 564 case kw_storage_class: 565 return lsym_storage_class; 566 567 case kw_typedef: 568 return lsym_typedef; 569 570 default: /* all others are treated like any other 571 * identifier */ 572 return lsym_ident; 573 } 574 } 575 576 if (*inp.s == '(' && ps.tos <= 1 && ps.ind_level == 0 && 577 !ps.in_parameter_declaration && !ps.block_init) { 578 579 for (const char *p = inp.s; p < inp.e;) 580 if (*p++ == ')' && (*p == ';' || *p == ',')) 581 goto not_proc; 582 583 strncpy(ps.procname, token.s, sizeof ps.procname - 1); 584 if (ps.in_decl) 585 ps.in_parameter_declaration = true; 586 return lsym_funcname; 587 not_proc:; 588 589 } else if (probably_typename()) { 590 ps.curr_keyword = kw_type; 591 ps.next_unary = true; 592 return lsym_type; 593 } 594 595 if (ps.prev_token == lsym_type) /* if this is a declared variable, 596 * then following sign is unary */ 597 ps.next_unary = true; /* will make "int a -1" work */ 598 599 return lsym_ident; /* the ident is not in the list */ 600 } 601 602 /* Reads the next token, placing it in the global variable "token". */ 603 lexer_symbol 604 lexi(void) 605 { 606 token.e = token.s; 607 ps.prev_col_1 = ps.prev_newline; 608 ps.prev_newline = false; 609 ps.prev_keyword = ps.curr_keyword; 610 ps.curr_keyword = kw_0; 611 612 while (ch_isblank(*inp.s)) { 613 ps.prev_col_1 = false; 614 inbuf_skip(); 615 } 616 617 lexer_symbol alnum_lsym = lexi_alnum(); 618 if (alnum_lsym != lsym_eof) 619 return lexi_end(alnum_lsym); 620 621 /* Scan a non-alphanumeric token */ 622 623 check_size_token(3); /* for things like "<<=" */ 624 *token.e++ = inbuf_next(); 625 *token.e = '\0'; 626 627 lexer_symbol lsym; 628 bool unary_delim = false; /* whether the current token forces a 629 * following operator to be unary */ 630 631 switch (*token.s) { 632 case '\n': 633 unary_delim = ps.next_unary; 634 ps.prev_newline = true; 635 /* if data has been exhausted, the newline is a dummy. */ 636 lsym = had_eof ? lsym_eof : lsym_newline; 637 break; 638 639 case '\'': 640 case '"': 641 lex_char_or_string(); 642 lsym = lsym_ident; 643 break; 644 645 case '(': 646 case '[': 647 unary_delim = true; 648 lsym = lsym_lparen_or_lbracket; 649 break; 650 651 case ')': 652 case ']': 653 lsym = lsym_rparen_or_rbracket; 654 break; 655 656 case '#': 657 unary_delim = ps.next_unary; 658 lsym = lsym_preprocessing; 659 break; 660 661 case '?': 662 unary_delim = true; 663 lsym = lsym_question; 664 break; 665 666 case ':': 667 lsym = lsym_colon; 668 unary_delim = true; 669 break; 670 671 case ';': 672 unary_delim = true; 673 lsym = lsym_semicolon; 674 break; 675 676 case '{': 677 unary_delim = true; 678 lsym = lsym_lbrace; 679 break; 680 681 case '}': 682 unary_delim = true; 683 lsym = lsym_rbrace; 684 break; 685 686 case '\f': 687 unary_delim = ps.next_unary; 688 ps.prev_newline = true; 689 lsym = lsym_form_feed; 690 break; 691 692 case ',': 693 unary_delim = true; 694 lsym = lsym_comma; 695 break; 696 697 case '.': 698 unary_delim = false; 699 lsym = lsym_period; 700 break; 701 702 case '-': 703 case '+': 704 lsym = ps.next_unary ? lsym_unary_op : lsym_binary_op; 705 unary_delim = true; 706 707 if (*inp.s == token.s[0]) { /* ++, -- */ 708 *token.e++ = *inp.s++; 709 if (ps.prev_token == lsym_ident || 710 ps.prev_token == lsym_rparen_or_rbracket) { 711 lsym = ps.next_unary ? lsym_unary_op : lsym_postfix_op; 712 unary_delim = false; 713 } 714 715 } else if (*inp.s == '=') { /* += */ 716 *token.e++ = *inp.s++; 717 718 } else if (*inp.s == '>') { /* -> */ 719 *token.e++ = *inp.s++; 720 unary_delim = false; 721 lsym = lsym_unary_op; 722 ps.want_blank = false; 723 } 724 break; 725 726 case '=': 727 if (ps.init_or_struct) 728 ps.block_init = true; 729 if (*inp.s == '=') { /* == */ 730 *token.e++ = *inp.s++; 731 *token.e = '\0'; 732 } 733 lsym = lsym_binary_op; 734 unary_delim = true; 735 break; 736 737 case '>': 738 case '<': 739 case '!': /* ops like <, <<, <=, !=, etc */ 740 if (*inp.s == '>' || *inp.s == '<' || *inp.s == '=') 741 *token.e++ = inbuf_next(); 742 if (*inp.s == '=') 743 *token.e++ = *inp.s++; 744 lsym = ps.next_unary ? lsym_unary_op : lsym_binary_op; 745 unary_delim = true; 746 break; 747 748 case '*': 749 unary_delim = true; 750 if (!ps.next_unary) { 751 if (*inp.s == '=') 752 *token.e++ = *inp.s++; 753 lsym = lsym_binary_op; 754 break; 755 } 756 757 while (*inp.s == '*' || isspace((unsigned char)*inp.s)) { 758 if (*inp.s == '*') 759 token_add_char('*'); 760 inbuf_skip(); 761 } 762 763 if (ps.in_decl) { 764 char *tp = inp.s; 765 766 while (isalpha((unsigned char)*tp) || 767 isspace((unsigned char)*tp)) { 768 if (++tp >= inp.e) 769 inbuf_read_line(); 770 } 771 if (*tp == '(') 772 ps.procname[0] = ' '; 773 } 774 775 lsym = lsym_unary_op; 776 break; 777 778 default: 779 if (token.s[0] == '/' && (*inp.s == '*' || *inp.s == '/')) { 780 /* it is start of comment */ 781 *token.e++ = inbuf_next(); 782 783 lsym = lsym_comment; 784 unary_delim = ps.next_unary; 785 break; 786 } 787 788 while (token.e[-1] == *inp.s || *inp.s == '=') { 789 /* handle '||', '&&', etc., and also things as in 'int *****i' */ 790 token_add_char(inbuf_next()); 791 } 792 793 lsym = ps.next_unary ? lsym_unary_op : lsym_binary_op; 794 unary_delim = true; 795 } 796 797 if (inp.s >= inp.e) /* check for input buffer empty */ 798 inbuf_read_line(); 799 800 ps.next_unary = unary_delim; 801 802 check_size_token(1); 803 *token.e = '\0'; 804 805 return lexi_end(lsym); 806 } 807 808 void 809 add_typename(const char *name) 810 { 811 if (typenames.len >= typenames.cap) { 812 typenames.cap = 16 + 2 * typenames.cap; 813 typenames.items = xrealloc(typenames.items, 814 sizeof(typenames.items[0]) * typenames.cap); 815 } 816 817 int pos = bsearch_typenames(name); 818 if (pos >= 0) 819 return; /* already in the list */ 820 821 pos = -(pos + 1); 822 memmove(typenames.items + pos + 1, typenames.items + pos, 823 sizeof(typenames.items[0]) * (typenames.len++ - (unsigned)pos)); 824 typenames.items[pos] = xstrdup(name); 825 } 826
Indexes created Fri Sep 26 20:09:58 GMT 2025