2 Copyright (C) 1986, 1987, 1989, 1992, 1993, 1994, 1995, 1996, 1997, 1998,
3 1999, 2000, 2001, 2002, 2003 Free Software Foundation, Inc.
4 Contributed by Per Bothner, 1994-95.
5 Based on CCCP program by Paul Rubin, June 1986
6 Adapted to ANSI C, Richard Stallman, Jan 1987
8 This program is free software; you can redistribute it and/or modify it
9 under the terms of the GNU General Public License as published by the
10 Free Software Foundation; either version 2, or (at your option) any
13 This program is distributed in the hope that it will be useful,
14 but WITHOUT ANY WARRANTY; without even the implied warranty of
15 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
16 GNU General Public License for more details.
18 You should have received a copy of the GNU General Public License
19 along with this program; if not, write to the Free Software
20 Foundation, 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */
28 static void init_library (void);
29 static void mark_named_operators (cpp_reader *);
30 static void read_original_filename (cpp_reader *);
31 static void post_options (cpp_reader *);
33 /* If we have designated initializers (GCC >2.7) these tables can be
34 initialized, constant data. Otherwise, they have to be filled in at
36 #if HAVE_DESIGNATED_INITIALIZERS
38 #define init_trigraph_map() /* Nothing. */
39 #define TRIGRAPH_MAP \
40 __extension__ const uchar _cpp_trigraph_map[UCHAR_MAX + 1] = {
43 #define s(p, v) [p] = v,
47 #define TRIGRAPH_MAP uchar _cpp_trigraph_map[UCHAR_MAX + 1] = { 0 }; \
48 static void init_trigraph_map (void) { \
49 unsigned char *x = _cpp_trigraph_map;
52 #define s(p, v) x[p] = v;
57 s('=', '#') s(')', ']') s('!', '|')
58 s('(', '[') s('\'', '^') s('>', '}')
59 s('/', '\\') s('<', '{') s('-', '~')
66 /* A set of booleans indicating what CPP features each source language
72 char extended_numbers;
74 char cplusplus_comments;
78 static const struct lang_flags lang_defaults[] =
79 { /* c99 c++ xnum std // digr */
80 /* GNUC89 */ { 0, 0, 1, 0, 1, 1 },
81 /* GNUC99 */ { 1, 0, 1, 0, 1, 1 },
82 /* STDC89 */ { 0, 0, 0, 1, 0, 0 },
83 /* STDC94 */ { 0, 0, 0, 1, 0, 1 },
84 /* STDC99 */ { 1, 0, 1, 1, 1, 1 },
85 /* GNUCXX */ { 0, 1, 1, 0, 1, 1 },
86 /* CXX98 */ { 0, 1, 1, 1, 1, 1 },
87 /* ASM */ { 0, 0, 1, 0, 1, 0 }
90 /* Sets internal flags correctly for a given language. */
92 cpp_set_lang (cpp_reader *pfile, enum c_lang lang)
94 const struct lang_flags *l = &lang_defaults[(int) lang];
96 CPP_OPTION (pfile, lang) = lang;
98 CPP_OPTION (pfile, c99) = l->c99;
99 CPP_OPTION (pfile, cplusplus) = l->cplusplus;
100 CPP_OPTION (pfile, extended_numbers) = l->extended_numbers;
101 CPP_OPTION (pfile, std) = l->std;
102 CPP_OPTION (pfile, trigraphs) = l->std;
103 CPP_OPTION (pfile, cplusplus_comments) = l->cplusplus_comments;
104 CPP_OPTION (pfile, digraphs) = l->digraphs;
107 /* Initialize library global state. */
111 static int initialized = 0;
117 /* Set up the trigraph map. This doesn't need to do anything if
118 we were compiled with a compiler that supports C99 designated
120 init_trigraph_map ();
124 /* Initialize a cpp_reader structure. */
126 cpp_create_reader (enum c_lang lang, hash_table *table)
130 /* Initialize this instance of the library if it hasn't been already. */
133 pfile = xcalloc (1, sizeof (cpp_reader));
135 cpp_set_lang (pfile, lang);
136 CPP_OPTION (pfile, warn_import) = 1;
137 CPP_OPTION (pfile, warn_multichar) = 1;
138 CPP_OPTION (pfile, discard_comments) = 1;
139 CPP_OPTION (pfile, discard_comments_in_macro_exp) = 1;
140 CPP_OPTION (pfile, show_column) = 1;
141 CPP_OPTION (pfile, tabstop) = 8;
142 CPP_OPTION (pfile, operator_names) = 1;
143 CPP_OPTION (pfile, warn_trigraphs) = 2;
144 CPP_OPTION (pfile, warn_endif_labels) = 1;
145 CPP_OPTION (pfile, warn_deprecated) = 1;
146 CPP_OPTION (pfile, warn_long_long) = !CPP_OPTION (pfile, c99);
147 CPP_OPTION (pfile, dollars_in_ident) = 1;
148 CPP_OPTION (pfile, warn_dollars) = 1;
150 /* Default CPP arithmetic to something sensible for the host for the
151 benefit of dumb users like fix-header. */
152 CPP_OPTION (pfile, precision) = CHAR_BIT * sizeof (long);
153 CPP_OPTION (pfile, char_precision) = CHAR_BIT;
154 CPP_OPTION (pfile, wchar_precision) = CHAR_BIT * sizeof (int);
155 CPP_OPTION (pfile, int_precision) = CHAR_BIT * sizeof (int);
156 CPP_OPTION (pfile, unsigned_char) = 0;
157 CPP_OPTION (pfile, unsigned_wchar) = 1;
158 CPP_OPTION (pfile, bytes_big_endian) = 1; /* does not matter */
160 /* Default to no charset conversion. */
161 CPP_OPTION (pfile, narrow_charset) = 0;
162 CPP_OPTION (pfile, wide_charset) = 0;
164 /* A fake empty "directory" used as the starting point for files
165 looked up without a search path. Name cannot be '/' because we
166 don't want to prepend anything at all to filenames using it. All
167 other entries are correct zero-initialized. */
168 pfile->no_search_path.name = (char *) "";
170 /* Initialize the line map. Start at logical line 1, so we can use
171 a line number of zero for special states. */
172 linemap_init (&pfile->line_maps);
174 /* Initialize lexer state. */
175 pfile->state.save_comments = ! CPP_OPTION (pfile, discard_comments);
177 /* Set up static tokens. */
178 pfile->avoid_paste.type = CPP_PADDING;
179 pfile->avoid_paste.val.source = NULL;
180 pfile->eof.type = CPP_EOF;
181 pfile->eof.flags = 0;
183 /* Create a token buffer for the lexer. */
184 _cpp_init_tokenrun (&pfile->base_run, 250);
185 pfile->cur_run = &pfile->base_run;
186 pfile->cur_token = pfile->base_run.base;
188 /* Initialize the base context. */
189 pfile->context = &pfile->base_context;
190 pfile->base_context.macro = 0;
191 pfile->base_context.prev = pfile->base_context.next = 0;
193 /* Aligned and unaligned storage. */
194 pfile->a_buff = _cpp_get_buff (pfile, 0);
195 pfile->u_buff = _cpp_get_buff (pfile, 0);
197 /* The expression parser stack. */
198 _cpp_expand_op_stack (pfile);
200 /* Initialize the buffer obstack. */
201 _obstack_begin (&pfile->buffer_ob, 0, 0,
202 (void *(*) (long)) xmalloc,
203 (void (*) (void *)) free);
205 _cpp_init_files (pfile);
207 _cpp_init_hashtable (pfile, table);
212 /* Free resources used by PFILE. Accessing PFILE after this function
213 returns leads to undefined behavior. Returns the error count. */
215 cpp_destroy (cpp_reader *pfile)
217 cpp_context *context, *contextn;
218 tokenrun *run, *runn;
220 free (pfile->op_stack);
222 while (CPP_BUFFER (pfile) != NULL)
223 _cpp_pop_buffer (pfile);
226 free (pfile->out.base);
228 if (pfile->macro_buffer)
230 free (pfile->macro_buffer);
231 pfile->macro_buffer = NULL;
232 pfile->macro_buffer_len = 0;
236 deps_free (pfile->deps);
237 obstack_free (&pfile->buffer_ob, 0);
239 _cpp_destroy_hashtable (pfile);
240 _cpp_cleanup_files (pfile);
241 _cpp_destroy_iconv (pfile);
243 _cpp_free_buff (pfile->a_buff);
244 _cpp_free_buff (pfile->u_buff);
245 _cpp_free_buff (pfile->free_buffs);
247 for (run = &pfile->base_run; run; run = runn)
251 if (run != &pfile->base_run)
255 for (context = pfile->base_context.next; context; context = contextn)
257 contextn = context->next;
261 linemap_free (&pfile->line_maps);
265 /* This structure defines one built-in identifier. A node will be
266 entered in the hash table under the name NAME, with value VALUE.
268 There are two tables of these. builtin_array holds all the
269 "builtin" macros: these are handled by builtin_macro() in
270 cppmacro.c. Builtin is somewhat of a misnomer -- the property of
271 interest is that these macros require special code to compute their
272 expansions. The value is a "builtin_type" enumerator.
274 operator_array holds the C++ named operators. These are keywords
275 which act as aliases for punctuators. In C++, they cannot be
276 altered through #define, and #if recognizes them as operators. In
277 C, these are not entered into the hash table at all (but see
278 <iso646.h>). The value is a token-type enumerator. */
283 unsigned short value;
286 #define B(n, t) { DSC(n), t }
287 static const struct builtin builtin_array[] =
289 B("__TIME__", BT_TIME),
290 B("__DATE__", BT_DATE),
291 B("__FILE__", BT_FILE),
292 B("__BASE_FILE__", BT_BASE_FILE),
293 B("__LINE__", BT_SPECLINE),
294 B("__INCLUDE_LEVEL__", BT_INCLUDE_LEVEL),
295 /* Keep builtins not used for -traditional-cpp at the end, and
296 update init_builtins() if any more are added. */
297 B("_Pragma", BT_PRAGMA),
298 B("__STDC__", BT_STDC),
301 static const struct builtin operator_array[] =
303 B("and", CPP_AND_AND),
304 B("and_eq", CPP_AND_EQ),
305 B("bitand", CPP_AND),
307 B("compl", CPP_COMPL),
309 B("not_eq", CPP_NOT_EQ),
311 B("or_eq", CPP_OR_EQ),
313 B("xor_eq", CPP_XOR_EQ)
317 /* Mark the C++ named operators in the hash table. */
319 mark_named_operators (cpp_reader *pfile)
321 const struct builtin *b;
323 for (b = operator_array;
324 b < (operator_array + ARRAY_SIZE (operator_array));
327 cpp_hashnode *hp = cpp_lookup (pfile, b->name, b->len);
328 hp->flags |= NODE_OPERATOR;
329 hp->is_directive = 0;
330 hp->directive_index = b->value;
334 /* Read the builtins table above and enter them, and language-specific
335 macros, into the hash table. HOSTED is true if this is a hosted
338 cpp_init_builtins (cpp_reader *pfile, int hosted)
340 const struct builtin *b;
341 size_t n = ARRAY_SIZE (builtin_array);
343 if (CPP_OPTION (pfile, traditional))
346 for(b = builtin_array; b < builtin_array + n; b++)
348 cpp_hashnode *hp = cpp_lookup (pfile, b->name, b->len);
350 hp->flags |= NODE_BUILTIN | NODE_WARN;
351 hp->value.builtin = b->value;
354 if (CPP_OPTION (pfile, cplusplus))
355 _cpp_define_builtin (pfile, "__cplusplus 1");
356 else if (CPP_OPTION (pfile, lang) == CLK_ASM)
357 _cpp_define_builtin (pfile, "__ASSEMBLER__ 1");
358 else if (CPP_OPTION (pfile, lang) == CLK_STDC94)
359 _cpp_define_builtin (pfile, "__STDC_VERSION__ 199409L");
360 else if (CPP_OPTION (pfile, c99))
361 _cpp_define_builtin (pfile, "__STDC_VERSION__ 199901L");
364 _cpp_define_builtin (pfile, "__STDC_HOSTED__ 1");
366 _cpp_define_builtin (pfile, "__STDC_HOSTED__ 0");
368 if (CPP_OPTION (pfile, objc))
369 _cpp_define_builtin (pfile, "__OBJC__ 1");
372 /* Sanity-checks are dependent on command-line options, so it is
373 called as a subroutine of cpp_read_main_file (). */
375 static void sanity_checks (cpp_reader *);
376 static void sanity_checks (cpp_reader *pfile)
379 size_t max_precision = 2 * CHAR_BIT * sizeof (cpp_num_part);
381 /* Sanity checks for assumptions about CPP arithmetic and target
382 type precisions made by cpplib. */
385 cpp_error (pfile, DL_ICE, "cppchar_t must be an unsigned type");
387 if (CPP_OPTION (pfile, precision) > max_precision)
388 cpp_error (pfile, DL_ICE,
389 "preprocessor arithmetic has maximum precision of %lu bits;"
390 " target requires %lu bits",
391 (unsigned long) max_precision,
392 (unsigned long) CPP_OPTION (pfile, precision));
394 if (CPP_OPTION (pfile, precision) < CPP_OPTION (pfile, int_precision))
395 cpp_error (pfile, DL_ICE,
396 "CPP arithmetic must be at least as precise as a target int");
398 if (CPP_OPTION (pfile, char_precision) < 8)
399 cpp_error (pfile, DL_ICE, "target char is less than 8 bits wide");
401 if (CPP_OPTION (pfile, wchar_precision) < CPP_OPTION (pfile, char_precision))
402 cpp_error (pfile, DL_ICE,
403 "target wchar_t is narrower than target char");
405 if (CPP_OPTION (pfile, int_precision) < CPP_OPTION (pfile, char_precision))
406 cpp_error (pfile, DL_ICE,
407 "target int is narrower than target char");
409 /* This is assumed in eval_token() and could be fixed if necessary. */
410 if (sizeof (cppchar_t) > sizeof (cpp_num_part))
411 cpp_error (pfile, DL_ICE, "CPP half-integer narrower than CPP character");
413 if (CPP_OPTION (pfile, wchar_precision) > BITS_PER_CPPCHAR_T)
414 cpp_error (pfile, DL_ICE,
415 "CPP on this host cannot handle wide character constants over"
416 " %lu bits, but the target requires %lu bits",
417 (unsigned long) BITS_PER_CPPCHAR_T,
418 (unsigned long) CPP_OPTION (pfile, wchar_precision));
421 # define sanity_checks(PFILE)
424 /* Add a dependency target. Can be called any number of times before
425 cpp_read_main_file(). If no targets have been added before
426 cpp_read_main_file(), then the default target is used. */
428 cpp_add_dependency_target (cpp_reader *pfile, const char *target, int quote)
431 pfile->deps = deps_init ();
433 deps_add_target (pfile->deps, target, quote);
436 /* This is called after options have been parsed, and partially
437 processed. Setup for processing input from the file named FNAME,
438 or stdin if it is the empty string. Return the original filename
439 on success (e.g. foo.i->foo.c), or NULL on failure. */
441 cpp_read_main_file (cpp_reader *pfile, const char *fname)
443 sanity_checks (pfile);
445 post_options (pfile);
447 /* Mark named operators before handling command line macros. */
448 if (CPP_OPTION (pfile, cplusplus) && CPP_OPTION (pfile, operator_names))
449 mark_named_operators (pfile);
451 if (CPP_OPTION (pfile, deps.style) != DEPS_NONE)
454 pfile->deps = deps_init ();
456 /* Set the default target (if there is none already). */
457 deps_add_default_target (pfile->deps, fname);
461 if (!cpp_stack_file (pfile, fname))
464 /* Set this here so the client can change the option if it wishes,
465 and after stacking the main file so we don't trace the main
467 pfile->line_maps.trace_includes = CPP_OPTION (pfile, print_include_names);
469 /* For foo.i, read the original filename foo.c now, for the benefit
470 of the front ends. */
471 if (CPP_OPTION (pfile, preprocessed))
472 read_original_filename (pfile);
474 return pfile->map->to_file;
477 /* For preprocessed files, if the first tokens are of the form # NUM.
478 handle the directive so we know the original file name. This will
479 generate file_change callbacks, which the front ends must handle
480 appropriately given their state of initialization. */
482 read_original_filename (cpp_reader *pfile)
484 const cpp_token *token, *token1;
486 /* Lex ahead; if the first tokens are of the form # NUM, then
487 process the directive, otherwise back up. */
488 token = _cpp_lex_direct (pfile);
489 if (token->type == CPP_HASH)
491 token1 = _cpp_lex_direct (pfile);
492 _cpp_backup_tokens (pfile, 1);
494 /* If it's a #line directive, handle it. */
495 if (token1->type == CPP_NUMBER)
497 _cpp_handle_directive (pfile, token->flags & PREV_WHITE);
502 /* Backup as if nothing happened. */
503 _cpp_backup_tokens (pfile, 1);
506 /* This is called at the end of preprocessing. It pops the last
507 buffer and writes dependency output, and returns the number of
510 Maybe it should also reset state, such that you could call
511 cpp_start_read with a new filename to restart processing. */
513 cpp_finish (cpp_reader *pfile, FILE *deps_stream)
515 /* Warn about unused macros before popping the final buffer. */
516 if (CPP_OPTION (pfile, warn_unused_macros))
517 cpp_forall_identifiers (pfile, _cpp_warn_if_unused_macro, NULL);
519 /* cpplex.c leaves the final buffer on the stack. This it so that
520 it returns an unending stream of CPP_EOFs to the client. If we
521 popped the buffer, we'd dereference a NULL buffer pointer and
522 segfault. It's nice to allow the client to do worry-free excess
523 cpp_get_token calls. */
524 while (pfile->buffer)
525 _cpp_pop_buffer (pfile);
527 /* Don't write the deps file if there are errors. */
528 if (CPP_OPTION (pfile, deps.style) != DEPS_NONE
529 && deps_stream && pfile->errors == 0)
531 deps_write (pfile->deps, deps_stream, 72);
533 if (CPP_OPTION (pfile, deps.phony_targets))
534 deps_phony_targets (pfile->deps, deps_stream);
537 /* Report on headers that could use multiple include guards. */
538 if (CPP_OPTION (pfile, print_include_names))
539 _cpp_report_missing_guards (pfile);
541 return pfile->errors;
545 post_options (cpp_reader *pfile)
547 /* -Wtraditional is not useful in C++ mode. */
548 if (CPP_OPTION (pfile, cplusplus))
549 CPP_OPTION (pfile, warn_traditional) = 0;
551 /* Permanently disable macro expansion if we are rescanning
552 preprocessed text. Read preprocesed source in ISO mode. */
553 if (CPP_OPTION (pfile, preprocessed))
555 pfile->state.prevent_expansion = 1;
556 CPP_OPTION (pfile, traditional) = 0;
559 if (CPP_OPTION (pfile, warn_trigraphs) == 2)
560 CPP_OPTION (pfile, warn_trigraphs) = !CPP_OPTION (pfile, trigraphs);
562 if (CPP_OPTION (pfile, traditional))
564 /* Traditional CPP does not accurately track column information. */
565 CPP_OPTION (pfile, show_column) = 0;
566 CPP_OPTION (pfile, trigraphs) = 0;
567 CPP_OPTION (pfile, warn_trigraphs) = 0;