lexi.c revision 1.125
1 /* $NetBSD: lexi.c,v 1.125 2021/10/31 19:57:44 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 1.125 2021/10/31 19:57:44 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 #include <assert.h> 55 56 #include "indent.h" 57 58 /* must be sorted alphabetically, is used in binary search */ 59 static const struct keyword { 60 const char *name; 61 lexer_symbol lsym; 62 enum keyword_kind kind; 63 } keywords[] = { 64 {"_Bool", lsym_eof, kw_type}, 65 {"_Complex", lsym_eof, kw_type}, 66 {"_Imaginary", lsym_eof, kw_type}, 67 {"auto", lsym_storage_class, kw_0}, 68 {"bool", lsym_eof, kw_type}, 69 {"break", lsym_ident, kw_0}, 70 {"case", lsym_case_label, kw_0}, 71 {"char", lsym_eof, kw_type}, 72 {"complex", lsym_eof, kw_type}, 73 {"const", lsym_eof, kw_type}, 74 {"continue", lsym_ident, kw_0}, 75 {"default", lsym_case_label, kw_0}, 76 {"do", lsym_do, kw_0}, 77 {"double", lsym_eof, kw_type}, 78 {"else", lsym_else, kw_0}, 79 {"enum", lsym_eof, kw_tag}, 80 {"extern", lsym_storage_class, kw_0}, 81 {"float", lsym_eof, kw_type}, 82 {"for", lsym_for, kw_0}, 83 {"goto", lsym_ident, kw_0}, 84 {"if", lsym_if, kw_0}, 85 {"imaginary", lsym_eof, kw_type}, 86 {"inline", lsym_ident, kw_0}, 87 {"int", lsym_eof, kw_type}, 88 {"long", lsym_eof, kw_type}, 89 {"offsetof", lsym_offsetof, kw_0}, 90 {"register", lsym_storage_class, kw_0}, 91 {"restrict", lsym_ident, kw_0}, 92 {"return", lsym_ident, kw_0}, 93 {"short", lsym_eof, kw_type}, 94 {"signed", lsym_eof, kw_type}, 95 {"sizeof", lsym_sizeof, kw_0}, 96 {"static", lsym_storage_class, kw_0}, 97 {"struct", lsym_eof, kw_tag}, 98 {"switch", lsym_switch, kw_0}, 99 {"typedef", lsym_typedef, kw_0}, 100 {"union", lsym_eof, kw_tag}, 101 {"unsigned", lsym_eof, kw_type}, 102 {"void", lsym_eof, kw_type}, 103 {"volatile", lsym_eof, kw_type}, 104 {"while", lsym_while, kw_0} 105 }; 106 107 static struct { 108 const char **items; 109 unsigned int len; 110 unsigned int cap; 111 } typenames; 112 113 /* 114 * The transition table below was rewritten by hand from lx's output, given 115 * the following definitions. lx is Katherine Flavel's lexer generator. 116 * 117 * O = /[0-7]/; D = /[0-9]/; NZ = /[1-9]/; 118 * H = /[a-f0-9]/i; B = /[0-1]/; HP = /0x/i; 119 * BP = /0b/i; E = /e[+\-]?/i D+; P = /p[+\-]?/i D+; 120 * FS = /[fl]/i; IS = /u/i /(l|L|ll|LL)/? | /(l|L|ll|LL)/ /u/i?; 121 * 122 * D+ E FS? -> $float; 123 * D* "." D+ E? FS? -> $float; 124 * D+ "." E? FS? -> $float; HP H+ IS? -> $int; 125 * HP H+ P FS? -> $float; NZ D* IS? -> $int; 126 * HP H* "." H+ P FS? -> $float; "0" O* IS? -> $int; 127 * HP H+ "." P FS -> $float; BP B+ IS? -> $int; 128 */ 129 /* INDENT OFF */ 130 static const unsigned char lex_number_state[][26] = { 131 /* examples: 132 00 133 s 0xx 134 t 00xaa 135 a 11 101100xxa.. 136 r 11ee0001101lbuuxx.a.pp 137 t.01.e+008bLuxll0Ll.aa.p+0 138 states: ABCDEFGHIJKLMNOPQRSTUVWXYZ */ 139 [0] = "uuiifuufiuuiiuiiiiiuiuuuuu", /* (other) */ 140 [1] = "CEIDEHHHIJQ U Q VUVVZZZ", /* 0 */ 141 [2] = "DEIDEHHHIJQ U Q VUVVZZZ", /* 1 */ 142 [3] = "DEIDEHHHIJ U VUVVZZZ", /* 2 3 4 5 6 7 */ 143 [4] = "DEJDEHHHJJ U VUVVZZZ", /* 8 9 */ 144 [5] = " U VUVV ", /* A a C c D d */ 145 [6] = " K U VUVV ", /* B b */ 146 [7] = " FFF FF U VUVV ", /* E e */ 147 [8] = " f f U VUVV f", /* F f */ 148 [9] = " LLf fL PR Li L f", /* L */ 149 [10] = " OOf fO S P O i O f", /* l */ 150 [11] = " FFX ", /* P p */ 151 [12] = " MM M i iiM M ", /* U u */ 152 [13] = " N ", /* X x */ 153 [14] = " G Y ", /* + - */ 154 [15] = "B EE EE T W ", /* . */ 155 /* ABCDEFGHIJKLMNOPQRSTUVWXYZ */ 156 }; 157 /* INDENT ON */ 158 159 static const unsigned char lex_number_row[] = { 160 ['0'] = 1, 161 ['1'] = 2, 162 ['2'] = 3, ['3'] = 3, ['4'] = 3, ['5'] = 3, ['6'] = 3, ['7'] = 3, 163 ['8'] = 4, ['9'] = 4, 164 ['A'] = 5, ['a'] = 5, ['C'] = 5, ['c'] = 5, ['D'] = 5, ['d'] = 5, 165 ['B'] = 6, ['b'] = 6, 166 ['E'] = 7, ['e'] = 7, 167 ['F'] = 8, ['f'] = 8, 168 ['L'] = 9, 169 ['l'] = 10, 170 ['P'] = 11, ['p'] = 11, 171 ['U'] = 12, ['u'] = 12, 172 ['X'] = 13, ['x'] = 13, 173 ['+'] = 14, ['-'] = 14, 174 ['.'] = 15, 175 }; 176 177 static char 178 inbuf_peek(void) 179 { 180 return *inp.s; 181 } 182 183 void 184 inbuf_skip(void) 185 { 186 inp.s++; 187 if (inp.s >= inp.e) 188 inbuf_read_line(); 189 } 190 191 char 192 inbuf_next(void) 193 { 194 char ch = inbuf_peek(); 195 inbuf_skip(); 196 return ch; 197 } 198 199 static void 200 check_size_token(size_t desired_size) 201 { 202 if (token.e + desired_size >= token.l) 203 buf_expand(&token, desired_size); 204 } 205 206 static void 207 token_add_char(char ch) 208 { 209 check_size_token(1); 210 *token.e++ = ch; 211 } 212 213 #ifdef debug 214 static const char * 215 lsym_name(lexer_symbol sym) 216 { 217 static const char *const name[] = { 218 "eof", 219 "preprocessing", 220 "newline", 221 "form_feed", 222 "comment", 223 "lparen_or_lbracket", 224 "rparen_or_rbracket", 225 "lbrace", 226 "rbrace", 227 "period", 228 "unary_op", 229 "binary_op", 230 "postfix_op", 231 "question", 232 "colon", 233 "comma", 234 "semicolon", 235 "typedef", 236 "storage_class", 237 "type_at_paren_level_0", 238 "tag", 239 "case_label", 240 "string_prefix", 241 "sizeof", 242 "offsetof", 243 "ident", 244 "funcname", 245 "do", 246 "else", 247 "for", 248 "if", 249 "switch", 250 "while", 251 }; 252 253 return name[sym]; 254 } 255 256 static const char * 257 kw_name(enum keyword_kind kw) 258 { 259 static const char *const name[] = { 260 "0", 261 "tag", 262 "type", 263 }; 264 265 return name[kw]; 266 } 267 268 static void 269 debug_print_buf(const char *name, const struct buffer *buf) 270 { 271 if (buf->s < buf->e) { 272 debug_printf("%s ", name); 273 debug_vis_range("\"", buf->s, buf->e, "\"\n"); 274 } 275 } 276 277 #define debug_ps_bool(name) \ 278 if (ps.name != prev_ps.name) \ 279 debug_println("[%c] ps." #name, ps.name ? 'x' : ' ') 280 #define debug_ps_int(name) \ 281 if (ps.name != prev_ps.name) \ 282 debug_println("%3d ps." #name, ps.name) 283 #define debug_ps_keyword(name) \ 284 if (ps.name != kw_0) \ 285 debug_println(" ps." #name " = %s", kw_name(ps.name)) 286 287 static void 288 debug_lexi(lexer_symbol lsym) 289 { 290 /* 291 * Watch out for 'rolled back parser state' in the debug output; the 292 * differences around these are unreliable. 293 */ 294 static struct parser_state prev_ps; 295 296 debug_println(""); 297 debug_printf("line %d: %s", line_no, lsym_name(lsym)); 298 if (ps.curr_keyword != kw_0) 299 debug_printf(" %s", kw_name(ps.curr_keyword)); 300 debug_vis_range(" \"", token.s, token.e, "\"\n"); 301 302 debug_print_buf("label", &lab); 303 debug_print_buf("code", &code); 304 debug_print_buf("comment", &com); 305 306 debug_println(" ps.prev_token = %s", lsym_name(ps.prev_token)); 307 debug_ps_keyword(prev_keyword); 308 debug_ps_bool(curr_newline); 309 debug_ps_bool(curr_col_1); 310 debug_ps_bool(next_unary); 311 // procname 312 debug_ps_bool(want_blank); 313 debug_ps_int(paren_level); 314 debug_ps_int(p_l_follow); 315 // paren_indents 316 debug_ps_int(cast_mask); 317 debug_ps_int(not_cast_mask); 318 319 debug_ps_int(comment_delta); 320 debug_ps_int(n_comment_delta); 321 debug_ps_int(com_ind); 322 323 debug_ps_bool(block_init); 324 debug_ps_int(block_init_level); 325 debug_ps_bool(init_or_struct); 326 327 debug_ps_int(ind_level); 328 debug_ps_int(ind_level_follow); 329 330 debug_ps_int(decl_nest); 331 debug_ps_bool(decl_on_line); 332 debug_ps_bool(in_decl); 333 debug_ps_int(just_saw_decl); 334 debug_ps_bool(in_parameter_declaration); 335 debug_ps_bool(decl_indent_done); 336 337 debug_ps_bool(in_stmt); 338 debug_ps_bool(ind_stmt); 339 debug_ps_bool(is_case_label); 340 341 debug_ps_bool(search_stmt); 342 343 prev_ps = ps; 344 } 345 #endif 346 347 /* ARGSUSED */ 348 static lexer_symbol 349 lexi_end(lexer_symbol lsym) 350 { 351 #ifdef debug 352 debug_lexi(lsym); 353 #endif 354 return lsym; 355 } 356 357 static void 358 lex_number(void) 359 { 360 for (unsigned char s = 'A'; s != 'f' && s != 'i' && s != 'u';) { 361 unsigned char ch = (unsigned char)*inp.s; 362 if (ch >= array_length(lex_number_row) || lex_number_row[ch] == 0) 363 break; 364 365 unsigned char row = lex_number_row[ch]; 366 if (lex_number_state[row][s - 'A'] == ' ') { 367 /*- 368 * lex_number_state[0][s - 'A'] now indicates the type: 369 * f = floating, i = integer, u = unknown 370 */ 371 break; 372 } 373 374 s = lex_number_state[row][s - 'A']; 375 token_add_char(inbuf_next()); 376 } 377 } 378 379 static void 380 lex_word(void) 381 { 382 while (isalnum((unsigned char)*inp.s) || 383 *inp.s == '\\' || 384 *inp.s == '_' || *inp.s == '$') { 385 386 if (*inp.s == '\\') { 387 if (inp.s[1] == '\n') { 388 inp.s += 2; 389 if (inp.s >= inp.e) 390 inbuf_read_line(); 391 } else 392 break; 393 } 394 395 token_add_char(inbuf_next()); 396 } 397 } 398 399 static void 400 lex_char_or_string(void) 401 { 402 for (char delim = *token.s;;) { 403 if (*inp.s == '\n') { 404 diag(1, "Unterminated literal"); 405 return; 406 } 407 408 token_add_char(inbuf_next()); 409 if (token.e[-1] == delim) 410 return; 411 412 if (token.e[-1] == '\\') { 413 if (*inp.s == '\n') 414 ++line_no; 415 token_add_char(inbuf_next()); 416 } 417 } 418 } 419 420 /* Guess whether the current token is a declared type. */ 421 static bool 422 probably_typename(void) 423 { 424 if (ps.p_l_follow > 0) 425 return false; 426 if (ps.block_init || ps.in_stmt) 427 return false; 428 if (inp.s[0] == '*' && inp.s[1] != '=') 429 goto maybe; 430 if (isalpha((unsigned char)*inp.s)) 431 goto maybe; 432 return false; 433 maybe: 434 return ps.prev_token == lsym_semicolon || 435 ps.prev_token == lsym_lbrace || 436 ps.prev_token == lsym_rbrace; 437 } 438 439 static int 440 bsearch_typenames(const char *key) 441 { 442 const char **arr = typenames.items; 443 int lo = 0; 444 int hi = (int)typenames.len - 1; 445 446 while (lo <= hi) { 447 int mid = (int)((unsigned)(lo + hi) >> 1); 448 int cmp = strcmp(arr[mid], key); 449 if (cmp < 0) 450 lo = mid + 1; 451 else if (cmp > 0) 452 hi = mid - 1; 453 else 454 return mid; 455 } 456 return -(lo + 1); 457 } 458 459 static bool 460 is_typename(void) 461 { 462 if (opt.auto_typedefs && 463 token.e - token.s >= 2 && memcmp(token.e - 2, "_t", 2) == 0) 464 return true; 465 466 return bsearch_typenames(token.s) >= 0; 467 } 468 469 static int 470 cmp_keyword_by_name(const void *key, const void *elem) 471 { 472 return strcmp(key, ((const struct keyword *)elem)->name); 473 } 474 475 /* Read an alphanumeric token into 'token', or return end_of_file. */ 476 static lexer_symbol 477 lexi_alnum(void) 478 { 479 if (isdigit((unsigned char)*inp.s) || 480 (inp.s[0] == '.' && isdigit((unsigned char)inp.s[1]))) { 481 lex_number(); 482 } else if (isalnum((unsigned char)*inp.s) || 483 *inp.s == '_' || *inp.s == '$') { 484 lex_word(); 485 } else 486 return lsym_eof; /* just as a placeholder */ 487 488 *token.e = '\0'; 489 490 if (token.s[0] == 'L' && token.s[1] == '\0' && 491 (*inp.s == '"' || *inp.s == '\'')) 492 return lsym_string_prefix; 493 494 while (ch_isblank(inbuf_peek())) 495 inbuf_skip(); 496 497 if (ps.prev_token == lsym_tag && ps.p_l_follow == 0) { 498 ps.next_unary = true; 499 return lsym_type_at_paren_level_0; 500 } 501 502 /* Operator after identifier is binary unless last token was 'struct'. */ 503 ps.next_unary = ps.prev_token == lsym_tag; 504 505 const struct keyword *kw = bsearch(token.s, keywords, 506 array_length(keywords), sizeof(keywords[0]), cmp_keyword_by_name); 507 if (kw == NULL) { 508 if (is_typename()) { 509 ps.curr_keyword = kw_type; 510 ps.next_unary = true; 511 goto found_typename; 512 } 513 514 } else { /* we have a keyword */ 515 ps.curr_keyword = kw->kind; 516 ps.next_unary = true; 517 518 assert((kw->lsym == lsym_eof) != (kw->kind == kw_0)); 519 if (kw->lsym != lsym_eof) 520 return kw->lsym; 521 if (kw->kind != kw_tag && kw->kind != kw_type) 522 return lsym_ident; 523 524 found_typename: 525 if (ps.p_l_follow > 0) { 526 /* inside parentheses: cast, param list, offsetof or sizeof */ 527 ps.cast_mask |= (1 << ps.p_l_follow) & ~ps.not_cast_mask; 528 } 529 if (ps.prev_token != lsym_period && ps.prev_token != lsym_unary_op) { 530 if (kw != NULL && kw->kind == kw_tag) 531 return lsym_tag; 532 if (ps.p_l_follow == 0) 533 return lsym_type_at_paren_level_0; 534 } 535 } 536 537 if (*inp.s == '(' && ps.tos <= 1 && ps.ind_level == 0 && 538 !ps.in_parameter_declaration && !ps.block_init) { 539 540 for (const char *p = inp.s; p < inp.e;) 541 if (*p++ == ')' && (*p == ';' || *p == ',')) 542 goto no_function_definition; 543 544 strncpy(ps.procname, token.s, sizeof ps.procname - 1); 545 if (ps.in_decl) 546 ps.in_parameter_declaration = true; 547 return lsym_funcname; 548 no_function_definition:; 549 550 } else if (probably_typename()) { 551 ps.curr_keyword = kw_type; 552 ps.next_unary = true; 553 return lsym_type_at_paren_level_0; 554 } 555 556 return lsym_ident; /* the ident is not in the list */ 557 } 558 559 /* Reads the next token, placing it in the global variable "token". */ 560 lexer_symbol 561 lexi(void) 562 { 563 token.e = token.s; 564 ps.curr_col_1 = ps.curr_newline; 565 ps.curr_newline = false; 566 ps.prev_keyword = ps.curr_keyword; 567 ps.curr_keyword = kw_0; 568 569 while (ch_isblank(*inp.s)) { 570 ps.curr_col_1 = false; 571 inbuf_skip(); 572 } 573 574 lexer_symbol alnum_lsym = lexi_alnum(); 575 if (alnum_lsym != lsym_eof) 576 return lexi_end(alnum_lsym); 577 578 /* Scan a non-alphanumeric token */ 579 580 check_size_token(3); /* for things like "<<=" */ 581 *token.e++ = inbuf_next(); 582 *token.e = '\0'; 583 584 lexer_symbol lsym; 585 bool unary_delim = false; /* whether the current token forces a 586 * following operator to be unary */ 587 588 switch (*token.s) { 589 case '\n': 590 unary_delim = ps.next_unary; 591 ps.curr_newline = true; 592 /* if data has been exhausted, the newline is a dummy. */ 593 lsym = had_eof ? lsym_eof : lsym_newline; 594 break; 595 596 case '\'': 597 case '"': 598 lex_char_or_string(); 599 lsym = lsym_ident; 600 break; 601 602 case '(': 603 case '[': 604 unary_delim = true; 605 lsym = lsym_lparen_or_lbracket; 606 break; 607 608 case ')': 609 case ']': 610 lsym = lsym_rparen_or_rbracket; 611 break; 612 613 case '#': 614 unary_delim = ps.next_unary; 615 lsym = lsym_preprocessing; 616 break; 617 618 case '?': 619 unary_delim = true; 620 lsym = lsym_question; 621 break; 622 623 case ':': 624 lsym = lsym_colon; 625 unary_delim = true; 626 break; 627 628 case ';': 629 unary_delim = true; 630 lsym = lsym_semicolon; 631 break; 632 633 case '{': 634 unary_delim = true; 635 lsym = lsym_lbrace; 636 break; 637 638 case '}': 639 unary_delim = true; 640 lsym = lsym_rbrace; 641 break; 642 643 case '\f': 644 unary_delim = ps.next_unary; 645 ps.curr_newline = true; 646 lsym = lsym_form_feed; 647 break; 648 649 case ',': 650 unary_delim = true; 651 lsym = lsym_comma; 652 break; 653 654 case '.': 655 unary_delim = false; 656 lsym = lsym_period; 657 break; 658 659 case '-': 660 case '+': 661 lsym = ps.next_unary ? lsym_unary_op : lsym_binary_op; 662 unary_delim = true; 663 664 if (*inp.s == token.s[0]) { /* ++, -- */ 665 *token.e++ = *inp.s++; 666 if (ps.prev_token == lsym_ident || 667 ps.prev_token == lsym_rparen_or_rbracket) { 668 lsym = ps.next_unary ? lsym_unary_op : lsym_postfix_op; 669 unary_delim = false; 670 } 671 672 } else if (*inp.s == '=') { /* += */ 673 *token.e++ = *inp.s++; 674 675 } else if (*inp.s == '>') { /* -> */ 676 *token.e++ = *inp.s++; 677 unary_delim = false; 678 lsym = lsym_unary_op; 679 ps.want_blank = false; 680 } 681 break; 682 683 case '=': 684 if (ps.init_or_struct) 685 ps.block_init = true; 686 if (*inp.s == '=') { /* == */ 687 *token.e++ = *inp.s++; 688 *token.e = '\0'; 689 } 690 lsym = lsym_binary_op; 691 unary_delim = true; 692 break; 693 694 case '>': 695 case '<': 696 case '!': /* ops like <, <<, <=, !=, etc */ 697 if (*inp.s == '>' || *inp.s == '<' || *inp.s == '=') 698 *token.e++ = inbuf_next(); 699 if (*inp.s == '=') 700 *token.e++ = *inp.s++; 701 lsym = ps.next_unary ? lsym_unary_op : lsym_binary_op; 702 unary_delim = true; 703 break; 704 705 case '*': 706 unary_delim = true; 707 if (!ps.next_unary) { 708 if (*inp.s == '=') 709 *token.e++ = *inp.s++; 710 lsym = lsym_binary_op; 711 break; 712 } 713 714 while (*inp.s == '*' || isspace((unsigned char)*inp.s)) { 715 if (*inp.s == '*') 716 token_add_char('*'); 717 inbuf_skip(); 718 } 719 720 if (ps.in_decl) { 721 char *tp = inp.s; 722 723 while (isalpha((unsigned char)*tp) || 724 isspace((unsigned char)*tp)) { 725 if (++tp >= inp.e) 726 inbuf_read_line(); 727 } 728 if (*tp == '(') 729 ps.procname[0] = ' '; 730 } 731 732 lsym = lsym_unary_op; 733 break; 734 735 default: 736 if (token.s[0] == '/' && (*inp.s == '*' || *inp.s == '/')) { 737 /* it is start of comment */ 738 *token.e++ = inbuf_next(); 739 740 lsym = lsym_comment; 741 unary_delim = ps.next_unary; 742 break; 743 } 744 745 while (token.e[-1] == *inp.s || *inp.s == '=') { 746 /* handle '||', '&&', etc., and also things as in 'int *****i' */ 747 token_add_char(inbuf_next()); 748 } 749 750 lsym = ps.next_unary ? lsym_unary_op : lsym_binary_op; 751 unary_delim = true; 752 } 753 754 if (inp.s >= inp.e) /* check for input buffer empty */ 755 inbuf_read_line(); 756 757 ps.next_unary = unary_delim; 758 759 check_size_token(1); 760 *token.e = '\0'; 761 762 return lexi_end(lsym); 763 } 764 765 void 766 add_typename(const char *name) 767 { 768 if (typenames.len >= typenames.cap) { 769 typenames.cap = 16 + 2 * typenames.cap; 770 typenames.items = xrealloc(typenames.items, 771 sizeof(typenames.items[0]) * typenames.cap); 772 } 773 774 int pos = bsearch_typenames(name); 775 if (pos >= 0) 776 return; /* already in the list */ 777 778 pos = -(pos + 1); 779 memmove(typenames.items + pos + 1, typenames.items + pos, 780 sizeof(typenames.items[0]) * (typenames.len++ - (unsigned)pos)); 781 typenames.items[pos] = xstrdup(name); 782 } 783
Indexes created Wed Sep 24 05:09:52 GMT 2025