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. */
24 #include "coretypes.h"
30 static void init_library PARAMS ((void));
31 static void mark_named_operators PARAMS ((cpp_reader *));
32 static void read_original_filename PARAMS ((cpp_reader *));
33 static void post_options PARAMS ((cpp_reader *));
35 /* If we have designated initializers (GCC >2.7) these tables can be
36 initialized, constant data. Otherwise, they have to be filled in at
38 #if HAVE_DESIGNATED_INITIALIZERS
40 #define init_trigraph_map() /* Nothing. */
41 #define TRIGRAPH_MAP \
42 __extension__ const uchar _cpp_trigraph_map[UCHAR_MAX + 1] = {
45 #define s(p, v) [p] = v,
49 #define TRIGRAPH_MAP uchar _cpp_trigraph_map[UCHAR_MAX + 1] = { 0 }; \
50 static void init_trigraph_map PARAMS ((void)) { \
51 unsigned char *x = _cpp_trigraph_map;
54 #define s(p, v) x[p] = v;
59 s('=', '#') s(')', ']') s('!', '|')
60 s('(', '[') s('\'', '^') s('>', '}')
61 s('/', '\\') s('<', '{') s('-', '~')
68 /* A set of booleans indicating what CPP features each source language
74 char extended_numbers;
76 char dollars_in_ident;
77 char cplusplus_comments;
81 /* ??? Enable $ in identifiers in assembly? */
82 static const struct lang_flags lang_defaults[] =
83 { /* c99 c++ xnum std dollar c++comm digr */
84 /* GNUC89 */ { 0, 0, 1, 0, 1, 1, 1 },
85 /* GNUC99 */ { 1, 0, 1, 0, 1, 1, 1 },
86 /* STDC89 */ { 0, 0, 0, 1, 0, 0, 0 },
87 /* STDC94 */ { 0, 0, 0, 1, 0, 0, 1 },
88 /* STDC99 */ { 1, 0, 1, 1, 0, 1, 1 },
89 /* GNUCXX */ { 0, 1, 1, 0, 1, 1, 1 },
90 /* CXX98 */ { 0, 1, 1, 1, 0, 1, 1 },
91 /* ASM */ { 0, 0, 1, 0, 0, 1, 0 }
94 /* Sets internal flags correctly for a given language. */
96 cpp_set_lang (pfile, lang)
100 const struct lang_flags *l = &lang_defaults[(int) lang];
102 CPP_OPTION (pfile, lang) = lang;
104 CPP_OPTION (pfile, c99) = l->c99;
105 CPP_OPTION (pfile, cplusplus) = l->cplusplus;
106 CPP_OPTION (pfile, extended_numbers) = l->extended_numbers;
107 CPP_OPTION (pfile, std) = l->std;
108 CPP_OPTION (pfile, trigraphs) = l->std;
109 CPP_OPTION (pfile, dollars_in_ident) = l->dollars_in_ident;
110 CPP_OPTION (pfile, cplusplus_comments) = l->cplusplus_comments;
111 CPP_OPTION (pfile, digraphs) = l->digraphs;
114 /* Initialize library global state. */
118 static int initialized = 0;
124 /* Set up the trigraph map. This doesn't need to do anything if
125 we were compiled with a compiler that supports C99 designated
127 init_trigraph_map ();
131 /* Initialize a cpp_reader structure. */
133 cpp_create_reader (lang, table)
139 /* Initialize this instance of the library if it hasn't been already. */
142 pfile = (cpp_reader *) xcalloc (1, sizeof (cpp_reader));
144 cpp_set_lang (pfile, lang);
145 CPP_OPTION (pfile, warn_import) = 1;
146 CPP_OPTION (pfile, warn_multichar) = 1;
147 CPP_OPTION (pfile, discard_comments) = 1;
148 CPP_OPTION (pfile, discard_comments_in_macro_exp) = 1;
149 CPP_OPTION (pfile, show_column) = 1;
150 CPP_OPTION (pfile, tabstop) = 8;
151 CPP_OPTION (pfile, operator_names) = 1;
152 CPP_OPTION (pfile, warn_trigraphs) = 2;
153 CPP_OPTION (pfile, warn_endif_labels) = 1;
154 CPP_OPTION (pfile, warn_deprecated) = 1;
155 CPP_OPTION (pfile, warn_long_long) = !CPP_OPTION (pfile, c99);
157 /* Default CPP arithmetic to something sensible for the host for the
158 benefit of dumb users like fix-header. */
159 CPP_OPTION (pfile, precision) = CHAR_BIT * sizeof (long);
160 CPP_OPTION (pfile, char_precision) = CHAR_BIT;
161 CPP_OPTION (pfile, wchar_precision) = CHAR_BIT * sizeof (int);
162 CPP_OPTION (pfile, int_precision) = CHAR_BIT * sizeof (int);
163 CPP_OPTION (pfile, unsigned_char) = 0;
164 CPP_OPTION (pfile, unsigned_wchar) = 1;
166 /* Initialize the line map. Start at logical line 1, so we can use
167 a line number of zero for special states. */
168 init_line_maps (&pfile->line_maps);
170 /* Initialize lexer state. */
171 pfile->state.save_comments = ! CPP_OPTION (pfile, discard_comments);
173 /* Set up static tokens. */
174 pfile->avoid_paste.type = CPP_PADDING;
175 pfile->avoid_paste.val.source = NULL;
176 pfile->eof.type = CPP_EOF;
177 pfile->eof.flags = 0;
179 /* Create a token buffer for the lexer. */
180 _cpp_init_tokenrun (&pfile->base_run, 250);
181 pfile->cur_run = &pfile->base_run;
182 pfile->cur_token = pfile->base_run.base;
184 /* Initialize the base context. */
185 pfile->context = &pfile->base_context;
186 pfile->base_context.macro = 0;
187 pfile->base_context.prev = pfile->base_context.next = 0;
189 /* Aligned and unaligned storage. */
190 pfile->a_buff = _cpp_get_buff (pfile, 0);
191 pfile->u_buff = _cpp_get_buff (pfile, 0);
193 /* The expression parser stack. */
194 _cpp_expand_op_stack (pfile);
196 /* Initialize the buffer obstack. */
197 gcc_obstack_init (&pfile->buffer_ob);
199 _cpp_init_includes (pfile);
201 _cpp_init_hashtable (pfile, table);
206 /* Free resources used by PFILE. Accessing PFILE after this function
207 returns leads to undefined behavior. Returns the error count. */
212 cpp_context *context, *contextn;
213 tokenrun *run, *runn;
215 free (pfile->op_stack);
217 while (CPP_BUFFER (pfile) != NULL)
218 _cpp_pop_buffer (pfile);
221 free (pfile->out.base);
223 if (pfile->macro_buffer)
225 free ((PTR) pfile->macro_buffer);
226 pfile->macro_buffer = NULL;
227 pfile->macro_buffer_len = 0;
231 deps_free (pfile->deps);
232 obstack_free (&pfile->buffer_ob, 0);
234 _cpp_destroy_hashtable (pfile);
235 _cpp_cleanup_includes (pfile);
237 _cpp_free_buff (pfile->a_buff);
238 _cpp_free_buff (pfile->u_buff);
239 _cpp_free_buff (pfile->free_buffs);
241 for (run = &pfile->base_run; run; run = runn)
245 if (run != &pfile->base_run)
249 for (context = pfile->base_context.next; context; context = contextn)
251 contextn = context->next;
255 free_line_maps (&pfile->line_maps);
259 /* This structure defines one built-in identifier. A node will be
260 entered in the hash table under the name NAME, with value VALUE.
262 There are two tables of these. builtin_array holds all the
263 "builtin" macros: these are handled by builtin_macro() in
264 cppmacro.c. Builtin is somewhat of a misnomer -- the property of
265 interest is that these macros require special code to compute their
266 expansions. The value is a "builtin_type" enumerator.
268 operator_array holds the C++ named operators. These are keywords
269 which act as aliases for punctuators. In C++, they cannot be
270 altered through #define, and #if recognizes them as operators. In
271 C, these are not entered into the hash table at all (but see
272 <iso646.h>). The value is a token-type enumerator. */
277 unsigned short value;
280 #define B(n, t) { DSC(n), t }
281 static const struct builtin builtin_array[] =
283 B("__TIME__", BT_TIME),
284 B("__DATE__", BT_DATE),
285 B("__FILE__", BT_FILE),
286 B("__BASE_FILE__", BT_BASE_FILE),
287 B("__LINE__", BT_SPECLINE),
288 B("__INCLUDE_LEVEL__", BT_INCLUDE_LEVEL),
289 /* Keep builtins not used for -traditional-cpp at the end, and
290 update init_builtins() if any more are added. */
291 B("_Pragma", BT_PRAGMA),
292 B("__STDC__", BT_STDC),
295 static const struct builtin operator_array[] =
297 B("and", CPP_AND_AND),
298 B("and_eq", CPP_AND_EQ),
299 B("bitand", CPP_AND),
301 B("compl", CPP_COMPL),
303 B("not_eq", CPP_NOT_EQ),
305 B("or_eq", CPP_OR_EQ),
307 B("xor_eq", CPP_XOR_EQ)
311 /* Mark the C++ named operators in the hash table. */
313 mark_named_operators (pfile)
316 const struct builtin *b;
318 for (b = operator_array;
319 b < (operator_array + ARRAY_SIZE (operator_array));
322 cpp_hashnode *hp = cpp_lookup (pfile, b->name, b->len);
323 hp->flags |= NODE_OPERATOR;
324 hp->is_directive = 0;
325 hp->directive_index = b->value;
329 /* Read the builtins table above and enter them, and language-specific
330 macros, into the hash table. */
332 cpp_init_builtins (pfile)
335 const struct builtin *b;
336 size_t n = ARRAY_SIZE (builtin_array);
338 if (CPP_OPTION (pfile, traditional))
341 for(b = builtin_array; b < builtin_array + n; b++)
343 cpp_hashnode *hp = cpp_lookup (pfile, b->name, b->len);
345 hp->flags |= NODE_BUILTIN | NODE_WARN;
346 hp->value.builtin = b->value;
349 if (CPP_OPTION (pfile, cplusplus))
350 _cpp_define_builtin (pfile, "__cplusplus 1");
351 else if (CPP_OPTION (pfile, lang) == CLK_ASM)
352 _cpp_define_builtin (pfile, "__ASSEMBLER__ 1");
353 else if (CPP_OPTION (pfile, lang) == CLK_STDC94)
354 _cpp_define_builtin (pfile, "__STDC_VERSION__ 199409L");
355 else if (CPP_OPTION (pfile, c99))
356 _cpp_define_builtin (pfile, "__STDC_VERSION__ 199901L");
358 if (CPP_OPTION (pfile, objc))
359 _cpp_define_builtin (pfile, "__OBJC__ 1");
362 /* Sanity-checks are dependent on command-line options, so it is
363 called as a subroutine of cpp_read_main_file (). */
365 static void sanity_checks PARAMS ((cpp_reader *));
366 static void sanity_checks (pfile)
370 size_t max_precision = 2 * CHAR_BIT * sizeof (cpp_num_part);
372 /* Sanity checks for assumptions about CPP arithmetic and target
373 type precisions made by cpplib. */
376 cpp_error (pfile, DL_ICE, "cppchar_t must be an unsigned type");
378 if (CPP_OPTION (pfile, precision) > max_precision)
379 cpp_error (pfile, DL_ICE,
380 "preprocessor arithmetic has maximum precision of %lu bits; target requires %lu bits",
381 (unsigned long) max_precision,
382 (unsigned long) CPP_OPTION (pfile, precision));
384 if (CPP_OPTION (pfile, precision) < CPP_OPTION (pfile, int_precision))
385 cpp_error (pfile, DL_ICE,
386 "CPP arithmetic must be at least as precise as a target int");
388 if (CPP_OPTION (pfile, char_precision) < 8)
389 cpp_error (pfile, DL_ICE, "target char is less than 8 bits wide");
391 if (CPP_OPTION (pfile, wchar_precision) < CPP_OPTION (pfile, char_precision))
392 cpp_error (pfile, DL_ICE,
393 "target wchar_t is narrower than target char");
395 if (CPP_OPTION (pfile, int_precision) < CPP_OPTION (pfile, char_precision))
396 cpp_error (pfile, DL_ICE,
397 "target int is narrower than target char");
399 /* This is assumed in eval_token() and could be fixed if necessary. */
400 if (sizeof (cppchar_t) > sizeof (cpp_num_part))
401 cpp_error (pfile, DL_ICE, "CPP half-integer narrower than CPP character");
403 if (CPP_OPTION (pfile, wchar_precision) > BITS_PER_CPPCHAR_T)
404 cpp_error (pfile, DL_ICE,
405 "CPP on this host cannot handle wide character constants over %lu bits, but the target requires %lu bits",
406 (unsigned long) BITS_PER_CPPCHAR_T,
407 (unsigned long) CPP_OPTION (pfile, wchar_precision));
410 # define sanity_checks(PFILE)
413 /* Add a dependency target. Can be called any number of times before
414 cpp_read_main_file(). If no targets have been added before
415 cpp_read_main_file(), then the default target is used. */
417 cpp_add_dependency_target (pfile, target, quote)
423 pfile->deps = deps_init ();
425 deps_add_target (pfile->deps, target, quote);
428 /* This is called after options have been parsed, and partially
429 processed. Setup for processing input from the file named FNAME,
430 or stdin if it is the empty string. Return the original filename
431 on success (e.g. foo.i->foo.c), or NULL on failure. */
433 cpp_read_main_file (pfile, fname)
437 sanity_checks (pfile);
439 post_options (pfile);
441 /* Mark named operators before handling command line macros. */
442 if (CPP_OPTION (pfile, cplusplus) && CPP_OPTION (pfile, operator_names))
443 mark_named_operators (pfile);
445 if (CPP_OPTION (pfile, deps.style) != DEPS_NONE)
448 pfile->deps = deps_init ();
450 /* Set the default target (if there is none already). */
451 deps_add_default_target (pfile->deps, fname);
454 /* Open the main input file. */
456 if (!_cpp_read_file (pfile, fname))
459 /* Set this here so the client can change the option if it wishes,
460 and after stacking the main file so we don't trace the main
462 pfile->line_maps.trace_includes = CPP_OPTION (pfile, print_include_names);
464 /* For foo.i, read the original filename foo.c now, for the benefit
465 of the front ends. */
466 if (CPP_OPTION (pfile, preprocessed))
467 read_original_filename (pfile);
469 return pfile->map->to_file;
472 /* For preprocessed files, if the first tokens are of the form # NUM.
473 handle the directive so we know the original file name. This will
474 generate file_change callbacks, which the front ends must handle
475 appropriately given their state of initialization. */
477 read_original_filename (pfile)
480 const cpp_token *token, *token1;
482 /* Lex ahead; if the first tokens are of the form # NUM, then
483 process the directive, otherwise back up. */
484 token = _cpp_lex_direct (pfile);
485 if (token->type == CPP_HASH)
487 token1 = _cpp_lex_direct (pfile);
488 _cpp_backup_tokens (pfile, 1);
490 /* If it's a #line directive, handle it. */
491 if (token1->type == CPP_NUMBER)
493 _cpp_handle_directive (pfile, token->flags & PREV_WHITE);
498 /* Backup as if nothing happened. */
499 _cpp_backup_tokens (pfile, 1);
502 /* This is called at the end of preprocessing. It pops the last
503 buffer and writes dependency output, and returns the number of
506 Maybe it should also reset state, such that you could call
507 cpp_start_read with a new filename to restart processing. */
509 cpp_finish (pfile, deps_stream)
513 /* Warn about unused macros before popping the final buffer. */
514 if (CPP_OPTION (pfile, warn_unused_macros))
515 cpp_forall_identifiers (pfile, _cpp_warn_if_unused_macro, NULL);
517 /* cpplex.c leaves the final buffer on the stack. This it so that
518 it returns an unending stream of CPP_EOFs to the client. If we
519 popped the buffer, we'd dereference a NULL buffer pointer and
520 segfault. It's nice to allow the client to do worry-free excess
521 cpp_get_token calls. */
522 while (pfile->buffer)
523 _cpp_pop_buffer (pfile);
525 /* Don't write the deps file if there are errors. */
526 if (CPP_OPTION (pfile, deps.style) != DEPS_NONE
527 && deps_stream && pfile->errors == 0)
529 deps_write (pfile->deps, deps_stream, 72);
531 if (CPP_OPTION (pfile, deps.phony_targets))
532 deps_phony_targets (pfile->deps, deps_stream);
535 /* Report on headers that could use multiple include guards. */
536 if (CPP_OPTION (pfile, print_include_names))
537 _cpp_report_missing_guards (pfile);
539 return pfile->errors;
546 /* -Wtraditional is not useful in C++ mode. */
547 if (CPP_OPTION (pfile, cplusplus))
548 CPP_OPTION (pfile, warn_traditional) = 0;
550 /* Permanently disable macro expansion if we are rescanning
551 preprocessed text. Read preprocesed source in ISO mode. */
552 if (CPP_OPTION (pfile, preprocessed))
554 pfile->state.prevent_expansion = 1;
555 CPP_OPTION (pfile, traditional) = 0;
558 if (CPP_OPTION (pfile, warn_trigraphs) == 2)
559 CPP_OPTION (pfile, warn_trigraphs) = !CPP_OPTION (pfile, trigraphs);
561 if (CPP_OPTION (pfile, traditional))
563 /* Traditional CPP does not accurately track column information. */
564 CPP_OPTION (pfile, show_column) = 0;
565 CPP_OPTION (pfile, trigraphs) = 0;
566 CPP_OPTION (pfile, warn_trigraphs) = 0;