Amit Daniel Kachhap | e6a01f5 | 2011-07-20 11:45:59 +0530 | [diff] [blame] | 1 | /**************************************************************************** |
| 2 | * Copyright (c) 1998-2006,2008 Free Software Foundation, Inc. * |
| 3 | * * |
| 4 | * Permission is hereby granted, free of charge, to any person obtaining a * |
| 5 | * copy of this software and associated documentation files (the * |
| 6 | * "Software"), to deal in the Software without restriction, including * |
| 7 | * without limitation the rights to use, copy, modify, merge, publish, * |
| 8 | * distribute, distribute with modifications, sublicense, and/or sell * |
| 9 | * copies of the Software, and to permit persons to whom the Software is * |
| 10 | * furnished to do so, subject to the following conditions: * |
| 11 | * * |
| 12 | * The above copyright notice and this permission notice shall be included * |
| 13 | * in all copies or substantial portions of the Software. * |
| 14 | * * |
| 15 | * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS * |
| 16 | * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF * |
| 17 | * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. * |
| 18 | * IN NO EVENT SHALL THE ABOVE COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, * |
| 19 | * DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR * |
| 20 | * OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR * |
| 21 | * THE USE OR OTHER DEALINGS IN THE SOFTWARE. * |
| 22 | * * |
| 23 | * Except as contained in this notice, the name(s) of the above copyright * |
| 24 | * holders shall not be used in advertising or otherwise to promote the * |
| 25 | * sale, use or other dealings in this Software without prior written * |
| 26 | * authorization. * |
| 27 | ****************************************************************************/ |
| 28 | |
| 29 | /**************************************************************************** |
| 30 | * Author: Zeyd M. Ben-Halim <zmbenhal@netcom.com> 1992,1995 * |
| 31 | * and: Eric S. Raymond <esr@snark.thyrsus.com> * |
| 32 | * and: Thomas E. Dickey 1996 on * |
| 33 | ****************************************************************************/ |
| 34 | |
| 35 | /* |
| 36 | * comp_scan.c --- Lexical scanner for terminfo compiler. |
| 37 | * |
| 38 | * _nc_reset_input() |
| 39 | * _nc_get_token() |
| 40 | * _nc_panic_mode() |
| 41 | * int _nc_syntax; |
| 42 | * int _nc_curr_line; |
| 43 | * long _nc_curr_file_pos; |
| 44 | * long _nc_comment_start; |
| 45 | * long _nc_comment_end; |
| 46 | */ |
| 47 | |
| 48 | #include <curses.priv.h> |
| 49 | |
| 50 | #include <ctype.h> |
| 51 | #include <term_entry.h> |
| 52 | #include <tic.h> |
| 53 | |
| 54 | MODULE_ID("$Id: comp_scan.c,v 1.83 2008/08/16 19:22:55 tom Exp $") |
| 55 | |
| 56 | /* |
| 57 | * Maximum length of string capability we'll accept before raising an error. |
| 58 | * Yes, there is a real capability in /etc/termcap this long, an "is". |
| 59 | */ |
| 60 | #define MAXCAPLEN 600 |
| 61 | |
| 62 | #define iswhite(ch) (ch == ' ' || ch == '\t') |
| 63 | |
| 64 | NCURSES_EXPORT_VAR(int) |
| 65 | _nc_syntax = 0; /* termcap or terminfo? */ |
| 66 | NCURSES_EXPORT_VAR(long) |
| 67 | _nc_curr_file_pos = 0; /* file offset of current line */ |
| 68 | NCURSES_EXPORT_VAR(long) |
| 69 | _nc_comment_start = 0; /* start of comment range before name */ |
| 70 | NCURSES_EXPORT_VAR(long) |
| 71 | _nc_comment_end = 0; /* end of comment range before name */ |
| 72 | NCURSES_EXPORT_VAR(long) |
| 73 | _nc_start_line = 0; /* start line of current entry */ |
| 74 | |
| 75 | NCURSES_EXPORT_VAR(struct token) |
| 76 | _nc_curr_token = |
| 77 | { |
| 78 | 0, 0, 0 |
| 79 | }; |
| 80 | |
| 81 | /***************************************************************************** |
| 82 | * |
| 83 | * Token-grabbing machinery |
| 84 | * |
| 85 | *****************************************************************************/ |
| 86 | |
| 87 | static bool first_column; /* See 'next_char()' below */ |
| 88 | static bool had_newline; |
| 89 | static char separator; /* capability separator */ |
| 90 | static int pushtype; /* type of pushback token */ |
| 91 | static char *pushname; |
| 92 | |
| 93 | #if NCURSES_EXT_FUNCS |
| 94 | NCURSES_EXPORT_VAR(bool) |
| 95 | _nc_disable_period = FALSE; /* used by tic -a option */ |
| 96 | #endif |
| 97 | |
| 98 | /***************************************************************************** |
| 99 | * |
| 100 | * Character-stream handling |
| 101 | * |
| 102 | *****************************************************************************/ |
| 103 | |
| 104 | #define LEXBUFSIZ 1024 |
| 105 | |
| 106 | static char *bufptr; /* otherwise, the input buffer pointer */ |
| 107 | static char *bufstart; /* start of buffer so we can compute offsets */ |
| 108 | static FILE *yyin; /* scanner's input file descriptor */ |
| 109 | |
| 110 | /* |
| 111 | * _nc_reset_input() |
| 112 | * |
| 113 | * Resets the input-reading routines. Used on initialization, |
| 114 | * or after a seek has been done. Exactly one argument must be |
| 115 | * non-null. |
| 116 | */ |
| 117 | |
| 118 | NCURSES_EXPORT(void) |
| 119 | _nc_reset_input(FILE *fp, char *buf) |
| 120 | { |
| 121 | pushtype = NO_PUSHBACK; |
| 122 | if (pushname != 0) |
| 123 | pushname[0] = '\0'; |
| 124 | yyin = fp; |
| 125 | bufstart = bufptr = buf; |
| 126 | _nc_curr_file_pos = 0L; |
| 127 | if (fp != 0) |
| 128 | _nc_curr_line = 0; |
| 129 | _nc_curr_col = 0; |
| 130 | } |
| 131 | |
| 132 | /* |
| 133 | * int last_char() |
| 134 | * |
| 135 | * Returns the final nonblank character on the current input buffer |
| 136 | */ |
| 137 | static int |
| 138 | last_char(void) |
| 139 | { |
| 140 | size_t len = strlen(bufptr); |
| 141 | while (len--) { |
| 142 | if (!isspace(UChar(bufptr[len]))) |
| 143 | return bufptr[len]; |
| 144 | } |
| 145 | return 0; |
| 146 | } |
| 147 | |
| 148 | /* |
| 149 | * int next_char() |
| 150 | * |
| 151 | * Returns the next character in the input stream. Comments and leading |
| 152 | * white space are stripped. |
| 153 | * |
| 154 | * The global state variable 'firstcolumn' is set TRUE if the character |
| 155 | * returned is from the first column of the input line. |
| 156 | * |
| 157 | * The global variable _nc_curr_line is incremented for each new line. |
| 158 | * The global variable _nc_curr_file_pos is set to the file offset of the |
| 159 | * beginning of each line. |
| 160 | */ |
| 161 | |
| 162 | static int |
| 163 | next_char(void) |
| 164 | { |
| 165 | static char *result; |
| 166 | static size_t allocated; |
| 167 | int the_char; |
| 168 | |
| 169 | if (!yyin) { |
| 170 | if (result != 0) { |
| 171 | FreeAndNull(result); |
| 172 | FreeAndNull(pushname); |
| 173 | allocated = 0; |
| 174 | } |
| 175 | /* |
| 176 | * An string with an embedded null will truncate the input. This is |
| 177 | * intentional (we don't read binary files here). |
| 178 | */ |
| 179 | if (bufptr == 0 || *bufptr == '\0') |
| 180 | return (EOF); |
| 181 | if (*bufptr == '\n') { |
| 182 | _nc_curr_line++; |
| 183 | _nc_curr_col = 0; |
| 184 | } else if (*bufptr == '\t') { |
| 185 | _nc_curr_col = (_nc_curr_col | 7); |
| 186 | } |
| 187 | } else if (!bufptr || !*bufptr) { |
| 188 | /* |
| 189 | * In theory this could be recoded to do its I/O one character at a |
| 190 | * time, saving the buffer space. In practice, this turns out to be |
| 191 | * quite hard to get completely right. Try it and see. If you |
| 192 | * succeed, don't forget to hack push_back() correspondingly. |
| 193 | */ |
| 194 | size_t used; |
| 195 | size_t len; |
| 196 | |
| 197 | do { |
| 198 | bufstart = 0; |
| 199 | used = 0; |
| 200 | do { |
| 201 | if (used + (LEXBUFSIZ / 4) >= allocated) { |
| 202 | allocated += (allocated + LEXBUFSIZ); |
| 203 | result = typeRealloc(char, allocated, result); |
| 204 | if (result == 0) |
| 205 | return (EOF); |
| 206 | bufstart = result; |
| 207 | } |
| 208 | if (used == 0) |
| 209 | _nc_curr_file_pos = ftell(yyin); |
| 210 | |
| 211 | if (fgets(result + used, (int) (allocated - used), yyin) != 0) { |
| 212 | bufstart = result; |
| 213 | if (used == 0) { |
| 214 | _nc_curr_line++; |
| 215 | _nc_curr_col = 0; |
| 216 | } |
| 217 | } else { |
| 218 | if (used != 0) |
| 219 | strcat(result, "\n"); |
| 220 | } |
| 221 | if ((bufptr = bufstart) != 0) { |
| 222 | used = strlen(bufptr); |
| 223 | while (iswhite(*bufptr)) { |
| 224 | if (*bufptr == '\t') { |
| 225 | _nc_curr_col = (_nc_curr_col | 7) + 1; |
| 226 | } else { |
| 227 | _nc_curr_col++; |
| 228 | } |
| 229 | bufptr++; |
| 230 | } |
| 231 | |
| 232 | /* |
| 233 | * Treat a trailing <cr><lf> the same as a <newline> so we |
| 234 | * can read files on OS/2, etc. |
| 235 | */ |
| 236 | if ((len = strlen(bufptr)) > 1) { |
| 237 | if (bufptr[len - 1] == '\n' |
| 238 | && bufptr[len - 2] == '\r') { |
| 239 | len--; |
| 240 | bufptr[len - 1] = '\n'; |
| 241 | bufptr[len] = '\0'; |
| 242 | } |
| 243 | } |
| 244 | } else { |
| 245 | return (EOF); |
| 246 | } |
| 247 | } while (bufptr[len - 1] != '\n'); /* complete a line */ |
| 248 | } while (result[0] == '#'); /* ignore comments */ |
| 249 | } else if (*bufptr == '\t') { |
| 250 | _nc_curr_col = (_nc_curr_col | 7); |
| 251 | } |
| 252 | |
| 253 | first_column = (bufptr == bufstart); |
| 254 | if (first_column) |
| 255 | had_newline = FALSE; |
| 256 | |
| 257 | _nc_curr_col++; |
| 258 | the_char = *bufptr++; |
| 259 | return UChar(the_char); |
| 260 | } |
| 261 | |
| 262 | static void |
| 263 | push_back(char c) |
| 264 | /* push a character back onto the input stream */ |
| 265 | { |
| 266 | if (bufptr == bufstart) |
| 267 | _nc_syserr_abort("Can't backspace off beginning of line"); |
| 268 | *--bufptr = c; |
| 269 | _nc_curr_col--; |
| 270 | } |
| 271 | |
| 272 | static long |
| 273 | stream_pos(void) |
| 274 | /* return our current character position in the input stream */ |
| 275 | { |
| 276 | return (yyin ? ftell(yyin) : (bufptr ? bufptr - bufstart : 0)); |
| 277 | } |
| 278 | |
| 279 | static bool |
| 280 | end_of_stream(void) |
| 281 | /* are we at end of input? */ |
| 282 | { |
| 283 | return ((yyin ? feof(yyin) : (bufptr && *bufptr == '\0')) |
| 284 | ? TRUE : FALSE); |
| 285 | } |
| 286 | |
| 287 | /* Assume we may be looking at a termcap-style continuation */ |
| 288 | static NCURSES_INLINE int |
| 289 | eat_escaped_newline(int ch) |
| 290 | { |
| 291 | if (ch == '\\') |
| 292 | while ((ch = next_char()) == '\n' || iswhite(ch)) |
| 293 | continue; |
| 294 | return ch; |
| 295 | } |
| 296 | |
| 297 | #define TOK_BUF_SIZE MAX_ENTRY_SIZE |
| 298 | |
| 299 | #define OkToAdd() \ |
| 300 | ((tok_ptr - tok_buf) < (TOK_BUF_SIZE - 2)) |
| 301 | |
| 302 | #define AddCh(ch) \ |
| 303 | *tok_ptr++ = (char) ch; \ |
| 304 | *tok_ptr = '\0' |
| 305 | |
| 306 | /* |
| 307 | * int |
| 308 | * get_token() |
| 309 | * |
| 310 | * Scans the input for the next token, storing the specifics in the |
| 311 | * global structure 'curr_token' and returning one of the following: |
| 312 | * |
| 313 | * NAMES A line beginning in column 1. 'name' |
| 314 | * will be set to point to everything up to but |
| 315 | * not including the first separator on the line. |
| 316 | * BOOLEAN An entry consisting of a name followed by |
| 317 | * a separator. 'name' will be set to point to |
| 318 | * the name of the capability. |
| 319 | * NUMBER An entry of the form |
| 320 | * name#digits, |
| 321 | * 'name' will be set to point to the capability |
| 322 | * name and 'valnumber' to the number given. |
| 323 | * STRING An entry of the form |
| 324 | * name=characters, |
| 325 | * 'name' is set to the capability name and |
| 326 | * 'valstring' to the string of characters, with |
| 327 | * input translations done. |
| 328 | * CANCEL An entry of the form |
| 329 | * name@, |
| 330 | * 'name' is set to the capability name and |
| 331 | * 'valnumber' to -1. |
| 332 | * EOF The end of the file has been reached. |
| 333 | * |
| 334 | * A `separator' is either a comma or a semicolon, depending on whether |
| 335 | * we are in termcap or terminfo mode. |
| 336 | * |
| 337 | */ |
| 338 | |
| 339 | NCURSES_EXPORT(int) |
| 340 | _nc_get_token(bool silent) |
| 341 | { |
| 342 | static const char terminfo_punct[] = "@%&*!#"; |
| 343 | static char *tok_buf; |
| 344 | |
| 345 | char *after_list; |
| 346 | char *after_name; |
| 347 | char *numchk; |
| 348 | char *tok_ptr; |
| 349 | char *s; |
| 350 | char numbuf[80]; |
| 351 | int ch; |
| 352 | int dot_flag = FALSE; |
| 353 | int type; |
| 354 | long number; |
| 355 | long token_start; |
| 356 | unsigned found; |
| 357 | #ifdef TRACE |
| 358 | int old_line; |
| 359 | int old_col; |
| 360 | #endif |
| 361 | |
| 362 | if (pushtype != NO_PUSHBACK) { |
| 363 | int retval = pushtype; |
| 364 | |
| 365 | _nc_set_type(pushname != 0 ? pushname : ""); |
| 366 | DEBUG(3, ("pushed-back token: `%s', class %d", |
| 367 | _nc_curr_token.tk_name, pushtype)); |
| 368 | |
| 369 | pushtype = NO_PUSHBACK; |
| 370 | if (pushname != 0) |
| 371 | pushname[0] = '\0'; |
| 372 | |
| 373 | /* currtok wasn't altered by _nc_push_token() */ |
| 374 | return (retval); |
| 375 | } |
| 376 | |
| 377 | if (end_of_stream()) { |
| 378 | yyin = 0; |
| 379 | next_char(); /* frees its allocated memory */ |
| 380 | if (tok_buf != 0) { |
| 381 | if (_nc_curr_token.tk_name == tok_buf) |
| 382 | _nc_curr_token.tk_name = 0; |
| 383 | FreeAndNull(tok_buf); |
| 384 | } |
| 385 | return (EOF); |
| 386 | } |
| 387 | |
| 388 | start_token: |
| 389 | token_start = stream_pos(); |
| 390 | while ((ch = next_char()) == '\n' || iswhite(ch)) { |
| 391 | if (ch == '\n') |
| 392 | had_newline = TRUE; |
| 393 | continue; |
| 394 | } |
| 395 | |
| 396 | ch = eat_escaped_newline(ch); |
| 397 | |
| 398 | #ifdef TRACE |
| 399 | old_line = _nc_curr_line; |
| 400 | old_col = _nc_curr_col; |
| 401 | #endif |
| 402 | if (ch == EOF) |
| 403 | type = EOF; |
| 404 | else { |
| 405 | /* if this is a termcap entry, skip a leading separator */ |
| 406 | if (separator == ':' && ch == ':') |
| 407 | ch = next_char(); |
| 408 | |
| 409 | if (ch == '.' |
| 410 | #if NCURSES_EXT_FUNCS |
| 411 | && !_nc_disable_period |
| 412 | #endif |
| 413 | ) { |
| 414 | dot_flag = TRUE; |
| 415 | DEBUG(8, ("dot-flag set")); |
| 416 | |
| 417 | while ((ch = next_char()) == '.' || iswhite(ch)) |
| 418 | continue; |
| 419 | } |
| 420 | |
| 421 | if (ch == EOF) { |
| 422 | type = EOF; |
| 423 | goto end_of_token; |
| 424 | } |
| 425 | |
| 426 | /* have to make some punctuation chars legal for terminfo */ |
| 427 | if (!isalnum(UChar(ch)) |
| 428 | #if NCURSES_EXT_FUNCS |
| 429 | && !(ch == '.' && _nc_disable_period) |
| 430 | #endif |
| 431 | && !strchr(terminfo_punct, (char) ch)) { |
| 432 | if (!silent) |
| 433 | _nc_warning("Illegal character (expected alphanumeric or %s) - '%s'", |
| 434 | terminfo_punct, unctrl((chtype) ch)); |
| 435 | _nc_panic_mode(separator); |
| 436 | goto start_token; |
| 437 | } |
| 438 | |
| 439 | if (tok_buf == 0) |
| 440 | tok_buf = typeMalloc(char, TOK_BUF_SIZE); |
| 441 | |
| 442 | #ifdef TRACE |
| 443 | old_line = _nc_curr_line; |
| 444 | old_col = _nc_curr_col; |
| 445 | #endif |
| 446 | tok_ptr = tok_buf; |
| 447 | AddCh(ch); |
| 448 | |
| 449 | if (first_column) { |
| 450 | _nc_comment_start = token_start; |
| 451 | _nc_comment_end = _nc_curr_file_pos; |
| 452 | _nc_start_line = _nc_curr_line; |
| 453 | |
| 454 | _nc_syntax = ERR; |
| 455 | after_name = 0; |
| 456 | after_list = 0; |
| 457 | while ((ch = next_char()) != '\n') { |
| 458 | if (ch == EOF) { |
| 459 | _nc_err_abort(MSG_NO_INPUTS); |
| 460 | } else if (ch == '|') { |
| 461 | after_list = tok_ptr; |
| 462 | if (after_name == 0) |
| 463 | after_name = tok_ptr; |
| 464 | } else if (ch == ':' && last_char() != ',') { |
| 465 | _nc_syntax = SYN_TERMCAP; |
| 466 | separator = ':'; |
| 467 | break; |
| 468 | } else if (ch == ',') { |
| 469 | _nc_syntax = SYN_TERMINFO; |
| 470 | separator = ','; |
| 471 | /* |
| 472 | * If we did not see a '|', then we found a name with no |
| 473 | * aliases or description. |
| 474 | */ |
| 475 | if (after_name == 0) |
| 476 | break; |
| 477 | /* |
| 478 | * If we see a comma, we assume this is terminfo unless we |
| 479 | * subsequently run into a colon. But we don't stop |
| 480 | * looking for a colon until hitting a newline. This |
| 481 | * allows commas to be embedded in description fields of |
| 482 | * either syntax. |
| 483 | */ |
| 484 | } else |
| 485 | ch = eat_escaped_newline(ch); |
| 486 | |
| 487 | if (OkToAdd()) { |
| 488 | AddCh(ch); |
| 489 | } else { |
| 490 | ch = EOF; |
| 491 | break; |
| 492 | } |
| 493 | } |
| 494 | *tok_ptr = '\0'; |
| 495 | if (_nc_syntax == ERR) { |
| 496 | /* |
| 497 | * Grrr...what we ought to do here is barf, complaining that |
| 498 | * the entry is malformed. But because a couple of name fields |
| 499 | * in the 8.2 termcap file end with |\, we just have to assume |
| 500 | * it's termcap syntax. |
| 501 | */ |
| 502 | _nc_syntax = SYN_TERMCAP; |
| 503 | separator = ':'; |
| 504 | } else if (_nc_syntax == SYN_TERMINFO) { |
| 505 | /* throw away trailing /, *$/ */ |
| 506 | for (--tok_ptr; |
| 507 | iswhite(*tok_ptr) || *tok_ptr == ','; |
| 508 | tok_ptr--) |
| 509 | continue; |
| 510 | tok_ptr[1] = '\0'; |
| 511 | } |
| 512 | |
| 513 | /* |
| 514 | * This is the soonest we have the terminal name fetched. Set up |
| 515 | * for following warning messages. If there's no '|', then there |
| 516 | * is no description. |
| 517 | */ |
| 518 | if (after_name != 0) { |
| 519 | ch = *after_name; |
| 520 | *after_name = '\0'; |
| 521 | _nc_set_type(tok_buf); |
| 522 | *after_name = (char) ch; |
| 523 | } |
| 524 | |
| 525 | /* |
| 526 | * Compute the boundary between the aliases and the description |
| 527 | * field for syntax-checking purposes. |
| 528 | */ |
| 529 | if (after_list != 0) { |
| 530 | if (!silent) { |
| 531 | if (*after_list == '\0') |
| 532 | _nc_warning("empty longname field"); |
| 533 | else if (strchr(after_list, ' ') == 0) |
| 534 | _nc_warning("older tic versions may treat the description field as an alias"); |
| 535 | } |
| 536 | } else { |
| 537 | after_list = tok_buf + strlen(tok_buf); |
| 538 | DEBUG(1, ("missing description")); |
| 539 | } |
| 540 | |
| 541 | /* |
| 542 | * Whitespace in a name field other than the long name can confuse |
| 543 | * rdist and some termcap tools. Slashes are a no-no. Other |
| 544 | * special characters can be dangerous due to shell expansion. |
| 545 | */ |
| 546 | for (s = tok_buf; s < after_list; ++s) { |
| 547 | if (isspace(UChar(*s))) { |
| 548 | if (!silent) |
| 549 | _nc_warning("whitespace in name or alias field"); |
| 550 | break; |
| 551 | } else if (*s == '/') { |
| 552 | if (!silent) |
| 553 | _nc_warning("slashes aren't allowed in names or aliases"); |
| 554 | break; |
| 555 | } else if (strchr("$[]!*?", *s)) { |
| 556 | if (!silent) |
| 557 | _nc_warning("dubious character `%c' in name or alias field", *s); |
| 558 | break; |
| 559 | } |
| 560 | } |
| 561 | |
| 562 | _nc_curr_token.tk_name = tok_buf; |
| 563 | type = NAMES; |
| 564 | } else { |
| 565 | if (had_newline && _nc_syntax == SYN_TERMCAP) { |
| 566 | _nc_warning("Missing backslash before newline"); |
| 567 | had_newline = FALSE; |
| 568 | } |
| 569 | while ((ch = next_char()) != EOF) { |
| 570 | if (!isalnum(UChar(ch))) { |
| 571 | if (_nc_syntax == SYN_TERMINFO) { |
| 572 | if (ch != '_') |
| 573 | break; |
| 574 | } else { /* allow ';' for "k;" */ |
| 575 | if (ch != ';') |
| 576 | break; |
| 577 | } |
| 578 | } |
| 579 | if (OkToAdd()) { |
| 580 | AddCh(ch); |
| 581 | } else { |
| 582 | ch = EOF; |
| 583 | break; |
| 584 | } |
| 585 | } |
| 586 | |
| 587 | *tok_ptr++ = '\0'; /* separate name/value in buffer */ |
| 588 | switch (ch) { |
| 589 | case ',': |
| 590 | case ':': |
| 591 | if (ch != separator) |
| 592 | _nc_err_abort("Separator inconsistent with syntax"); |
| 593 | _nc_curr_token.tk_name = tok_buf; |
| 594 | type = BOOLEAN; |
| 595 | break; |
| 596 | case '@': |
| 597 | if ((ch = next_char()) != separator && !silent) |
| 598 | _nc_warning("Missing separator after `%s', have %s", |
| 599 | tok_buf, unctrl((chtype) ch)); |
| 600 | _nc_curr_token.tk_name = tok_buf; |
| 601 | type = CANCEL; |
| 602 | break; |
| 603 | |
| 604 | case '#': |
| 605 | found = 0; |
| 606 | while (isalnum(ch = next_char())) { |
| 607 | numbuf[found++] = (char) ch; |
| 608 | if (found >= sizeof(numbuf) - 1) |
| 609 | break; |
| 610 | } |
| 611 | numbuf[found] = '\0'; |
| 612 | number = strtol(numbuf, &numchk, 0); |
| 613 | if (!silent) { |
| 614 | if (numchk == numbuf) |
| 615 | _nc_warning("no value given for `%s'", tok_buf); |
| 616 | if ((*numchk != '\0') || (ch != separator)) |
| 617 | _nc_warning("Missing separator"); |
| 618 | } |
| 619 | _nc_curr_token.tk_name = tok_buf; |
| 620 | _nc_curr_token.tk_valnumber = number; |
| 621 | type = NUMBER; |
| 622 | break; |
| 623 | |
| 624 | case '=': |
| 625 | ch = _nc_trans_string(tok_ptr, tok_buf + TOK_BUF_SIZE); |
| 626 | if (!silent && ch != separator) |
| 627 | _nc_warning("Missing separator"); |
| 628 | _nc_curr_token.tk_name = tok_buf; |
| 629 | _nc_curr_token.tk_valstring = tok_ptr; |
| 630 | type = STRING; |
| 631 | break; |
| 632 | |
| 633 | case EOF: |
| 634 | type = EOF; |
| 635 | break; |
| 636 | default: |
| 637 | /* just to get rid of the compiler warning */ |
| 638 | type = UNDEF; |
| 639 | if (!silent) |
| 640 | _nc_warning("Illegal character - '%s'", unctrl((chtype) ch)); |
| 641 | } |
| 642 | } /* end else (first_column == FALSE) */ |
| 643 | } /* end else (ch != EOF) */ |
| 644 | |
| 645 | end_of_token: |
| 646 | |
| 647 | #ifdef TRACE |
| 648 | if (dot_flag == TRUE) |
| 649 | DEBUG(8, ("Commented out ")); |
| 650 | |
| 651 | if (_nc_tracing >= DEBUG_LEVEL(8)) { |
| 652 | _tracef("parsed %d.%d to %d.%d", |
| 653 | old_line, old_col, |
| 654 | _nc_curr_line, _nc_curr_col); |
| 655 | } |
| 656 | if (_nc_tracing >= DEBUG_LEVEL(7)) { |
| 657 | switch (type) { |
| 658 | case BOOLEAN: |
| 659 | _tracef("Token: Boolean; name='%s'", |
| 660 | _nc_curr_token.tk_name); |
| 661 | break; |
| 662 | |
| 663 | case NUMBER: |
| 664 | _tracef("Token: Number; name='%s', value=%d", |
| 665 | _nc_curr_token.tk_name, |
| 666 | _nc_curr_token.tk_valnumber); |
| 667 | break; |
| 668 | |
| 669 | case STRING: |
| 670 | _tracef("Token: String; name='%s', value=%s", |
| 671 | _nc_curr_token.tk_name, |
| 672 | _nc_visbuf(_nc_curr_token.tk_valstring)); |
| 673 | break; |
| 674 | |
| 675 | case CANCEL: |
| 676 | _tracef("Token: Cancel; name='%s'", |
| 677 | _nc_curr_token.tk_name); |
| 678 | break; |
| 679 | |
| 680 | case NAMES: |
| 681 | |
| 682 | _tracef("Token: Names; value='%s'", |
| 683 | _nc_curr_token.tk_name); |
| 684 | break; |
| 685 | |
| 686 | case EOF: |
| 687 | _tracef("Token: End of file"); |
| 688 | break; |
| 689 | |
| 690 | default: |
| 691 | _nc_warning("Bad token type"); |
| 692 | } |
| 693 | } |
| 694 | #endif |
| 695 | |
| 696 | if (dot_flag == TRUE) /* if commented out, use the next one */ |
| 697 | type = _nc_get_token(silent); |
| 698 | |
| 699 | DEBUG(3, ("token: `%s', class %d", |
| 700 | ((_nc_curr_token.tk_name != 0) |
| 701 | ? _nc_curr_token.tk_name |
| 702 | : "<null>"), |
| 703 | type)); |
| 704 | |
| 705 | return (type); |
| 706 | } |
| 707 | |
| 708 | /* |
| 709 | * char |
| 710 | * trans_string(ptr) |
| 711 | * |
| 712 | * Reads characters using next_char() until encountering a separator, nl, |
| 713 | * or end-of-file. The returned value is the character which caused |
| 714 | * reading to stop. The following translations are done on the input: |
| 715 | * |
| 716 | * ^X goes to ctrl-X (i.e. X & 037) |
| 717 | * {\E,\n,\r,\b,\t,\f} go to |
| 718 | * {ESCAPE,newline,carriage-return,backspace,tab,formfeed} |
| 719 | * {\^,\\} go to {carat,backslash} |
| 720 | * \ddd (for ddd = up to three octal digits) goes to the character ddd |
| 721 | * |
| 722 | * \e == \E |
| 723 | * \0 == \200 |
| 724 | * |
| 725 | */ |
| 726 | |
| 727 | NCURSES_EXPORT(int) |
| 728 | _nc_trans_string(char *ptr, char *last) |
| 729 | { |
| 730 | int count = 0; |
| 731 | int number = 0; |
| 732 | int i, c; |
| 733 | chtype ch, last_ch = '\0'; |
| 734 | bool ignored = FALSE; |
| 735 | bool long_warning = FALSE; |
| 736 | |
| 737 | while ((ch = c = next_char()) != (chtype) separator && c != EOF) { |
| 738 | if (ptr >= (last - 1)) { |
| 739 | if (c != EOF) { |
| 740 | while ((c = next_char()) != separator && c != EOF) { |
| 741 | ; |
| 742 | } |
| 743 | ch = c; |
| 744 | } |
| 745 | break; |
| 746 | } |
| 747 | if ((_nc_syntax == SYN_TERMCAP) && c == '\n') |
| 748 | break; |
| 749 | if (ch == '^' && last_ch != '%') { |
| 750 | ch = c = next_char(); |
| 751 | if (c == EOF) |
| 752 | _nc_err_abort(MSG_NO_INPUTS); |
| 753 | |
| 754 | if (!(is7bits(ch) && isprint(ch))) { |
| 755 | _nc_warning("Illegal ^ character - '%s'", unctrl(ch)); |
| 756 | } |
| 757 | if (ch == '?') { |
| 758 | *(ptr++) = '\177'; |
| 759 | if (_nc_tracing) |
| 760 | _nc_warning("Allow ^? as synonym for \\177"); |
| 761 | } else { |
| 762 | if ((ch &= 037) == 0) |
| 763 | ch = 128; |
| 764 | *(ptr++) = (char) (ch); |
| 765 | } |
| 766 | } else if (ch == '\\') { |
| 767 | ch = c = next_char(); |
| 768 | if (c == EOF) |
| 769 | _nc_err_abort(MSG_NO_INPUTS); |
| 770 | |
| 771 | if (ch >= '0' && ch <= '7') { |
| 772 | number = ch - '0'; |
| 773 | for (i = 0; i < 2; i++) { |
| 774 | ch = c = next_char(); |
| 775 | if (c == EOF) |
| 776 | _nc_err_abort(MSG_NO_INPUTS); |
| 777 | |
| 778 | if (c < '0' || c > '7') { |
| 779 | if (isdigit(c)) { |
| 780 | _nc_warning("Non-octal digit `%c' in \\ sequence", c); |
| 781 | /* allow the digit; it'll do less harm */ |
| 782 | } else { |
| 783 | push_back((char) c); |
| 784 | break; |
| 785 | } |
| 786 | } |
| 787 | |
| 788 | number = number * 8 + c - '0'; |
| 789 | } |
| 790 | |
| 791 | if (number == 0) |
| 792 | number = 0200; |
| 793 | *(ptr++) = (char) number; |
| 794 | } else { |
| 795 | switch (c) { |
| 796 | case 'E': |
| 797 | case 'e': |
| 798 | *(ptr++) = '\033'; |
| 799 | break; |
| 800 | |
| 801 | case 'a': |
| 802 | *(ptr++) = '\007'; |
| 803 | break; |
| 804 | |
| 805 | case 'l': |
| 806 | case 'n': |
| 807 | *(ptr++) = '\n'; |
| 808 | break; |
| 809 | |
| 810 | case 'r': |
| 811 | *(ptr++) = '\r'; |
| 812 | break; |
| 813 | |
| 814 | case 'b': |
| 815 | *(ptr++) = '\010'; |
| 816 | break; |
| 817 | |
| 818 | case 's': |
| 819 | *(ptr++) = ' '; |
| 820 | break; |
| 821 | |
| 822 | case 'f': |
| 823 | *(ptr++) = '\014'; |
| 824 | break; |
| 825 | |
| 826 | case 't': |
| 827 | *(ptr++) = '\t'; |
| 828 | break; |
| 829 | |
| 830 | case '\\': |
| 831 | *(ptr++) = '\\'; |
| 832 | break; |
| 833 | |
| 834 | case '^': |
| 835 | *(ptr++) = '^'; |
| 836 | break; |
| 837 | |
| 838 | case ',': |
| 839 | *(ptr++) = ','; |
| 840 | break; |
| 841 | |
| 842 | case ':': |
| 843 | *(ptr++) = ':'; |
| 844 | break; |
| 845 | |
| 846 | case '\n': |
| 847 | continue; |
| 848 | |
| 849 | default: |
| 850 | _nc_warning("Illegal character '%s' in \\ sequence", |
| 851 | unctrl(ch)); |
| 852 | /* FALLTHRU */ |
| 853 | case '|': |
| 854 | *(ptr++) = (char) ch; |
| 855 | } /* endswitch (ch) */ |
| 856 | } /* endelse (ch < '0' || ch > '7') */ |
| 857 | } |
| 858 | /* end else if (ch == '\\') */ |
| 859 | else if (ch == '\n' && (_nc_syntax == SYN_TERMINFO)) { |
| 860 | /* |
| 861 | * Newlines embedded in a terminfo string are ignored, provided |
| 862 | * that the next line begins with whitespace. |
| 863 | */ |
| 864 | ignored = TRUE; |
| 865 | } else { |
| 866 | *(ptr++) = (char) ch; |
| 867 | } |
| 868 | |
| 869 | if (!ignored) { |
| 870 | if (_nc_curr_col <= 1) { |
| 871 | push_back((char) ch); |
| 872 | ch = '\n'; |
| 873 | break; |
| 874 | } |
| 875 | last_ch = ch; |
| 876 | count++; |
| 877 | } |
| 878 | ignored = FALSE; |
| 879 | |
| 880 | if (count > MAXCAPLEN && !long_warning) { |
| 881 | _nc_warning("Very long string found. Missing separator?"); |
| 882 | long_warning = TRUE; |
| 883 | } |
| 884 | } /* end while */ |
| 885 | |
| 886 | *ptr = '\0'; |
| 887 | |
| 888 | return (ch); |
| 889 | } |
| 890 | |
| 891 | /* |
| 892 | * _nc_push_token() |
| 893 | * |
| 894 | * Push a token of given type so that it will be reread by the next |
| 895 | * get_token() call. |
| 896 | */ |
| 897 | |
| 898 | NCURSES_EXPORT(void) |
| 899 | _nc_push_token(int tokclass) |
| 900 | { |
| 901 | /* |
| 902 | * This implementation is kind of bogus, it will fail if we ever do more |
| 903 | * than one pushback at a time between get_token() calls. It relies on the |
| 904 | * fact that _nc_curr_token is static storage that nothing but |
| 905 | * _nc_get_token() touches. |
| 906 | */ |
| 907 | pushtype = tokclass; |
| 908 | if (pushname == 0) |
| 909 | pushname = typeMalloc(char, MAX_NAME_SIZE + 1); |
| 910 | _nc_get_type(pushname); |
| 911 | |
| 912 | DEBUG(3, ("pushing token: `%s', class %d", |
| 913 | ((_nc_curr_token.tk_name != 0) |
| 914 | ? _nc_curr_token.tk_name |
| 915 | : "<null>"), |
| 916 | pushtype)); |
| 917 | } |
| 918 | |
| 919 | /* |
| 920 | * Panic mode error recovery - skip everything until a "ch" is found. |
| 921 | */ |
| 922 | NCURSES_EXPORT(void) |
| 923 | _nc_panic_mode(char ch) |
| 924 | { |
| 925 | int c; |
| 926 | |
| 927 | for (;;) { |
| 928 | c = next_char(); |
| 929 | if (c == ch) |
| 930 | return; |
| 931 | if (c == EOF) |
| 932 | return; |
| 933 | } |
| 934 | } |
| 935 | |
| 936 | #if NO_LEAKS |
| 937 | NCURSES_EXPORT(void) |
| 938 | _nc_comp_scan_leaks(void) |
| 939 | { |
| 940 | if (pushname != 0) { |
| 941 | FreeAndNull(pushname); |
| 942 | } |
| 943 | } |
| 944 | #endif |