* cppcharset.c (one_utf8_to_cppchar, one_cppchar_to_utf8,
[official-gcc.git] / gcc / cpphash.h
blob0e1bb3922ce88c27421eed615fd7c00de2f965de
1 /* Part of CPP library.
2 Copyright (C) 1997, 1998, 1999, 2000, 2001, 2002, 2003
3 Free Software Foundation, Inc.
5 This program is free software; you can redistribute it and/or modify it
6 under the terms of the GNU General Public License as published by the
7 Free Software Foundation; either version 2, or (at your option) any
8 later version.
10 This program is distributed in the hope that it will be useful,
11 but WITHOUT ANY WARRANTY; without even the implied warranty of
12 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13 GNU General Public License for more details.
15 You should have received a copy of the GNU General Public License
16 along with this program; if not, write to the Free Software
17 Foundation, 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */
19 /* This header defines all the internal data structures and functions
20 that need to be visible across files. It's called cpphash.h for
21 historical reasons. */
23 #ifndef GCC_CPPHASH_H
24 #define GCC_CPPHASH_H
26 #include "hashtable.h"
28 #ifdef HAVE_ICONV
29 #include <iconv.h>
30 #else
31 #define HAVE_ICONV 0
32 typedef int iconv_t; /* dummy */
33 #endif
35 struct directive; /* Deliberately incomplete. */
36 struct pending_option;
37 struct op;
38 struct strbuf;
40 typedef bool (*convert_f) (iconv_t, const unsigned char *, size_t,
41 struct strbuf *);
42 struct cset_converter
44 convert_f func;
45 iconv_t cd;
48 #ifndef HAVE_UCHAR
49 typedef unsigned char uchar;
50 #endif
51 #define U (const uchar *) /* Intended use: U"string" */
53 #define BITS_PER_CPPCHAR_T (CHAR_BIT * sizeof (cppchar_t))
55 /* Test if a sign is valid within a preprocessing number. */
56 #define VALID_SIGN(c, prevc) \
57 (((c) == '+' || (c) == '-') && \
58 ((prevc) == 'e' || (prevc) == 'E' \
59 || (((prevc) == 'p' || (prevc) == 'P') \
60 && CPP_OPTION (pfile, extended_numbers))))
62 #define CPP_OPTION(PFILE, OPTION) ((PFILE)->opts.OPTION)
63 #define CPP_BUFFER(PFILE) ((PFILE)->buffer)
64 #define CPP_BUF_COLUMN(BUF, CUR) ((CUR) - (BUF)->line_base)
65 #define CPP_BUF_COL(BUF) CPP_BUF_COLUMN(BUF, (BUF)->cur)
67 /* Maximum nesting of cpp_buffers. We use a static limit, partly for
68 efficiency, and partly to limit runaway recursion. */
69 #define CPP_STACK_MAX 200
71 /* Host alignment handling. */
72 struct dummy
74 char c;
75 union
77 double d;
78 int *p;
79 } u;
82 #define DEFAULT_ALIGNMENT offsetof (struct dummy, u)
83 #define CPP_ALIGN2(size, align) (((size) + ((align) - 1)) & ~((align) - 1))
84 #define CPP_ALIGN(size) CPP_ALIGN2 (size, DEFAULT_ALIGNMENT)
86 /* Each macro definition is recorded in a cpp_macro structure.
87 Variadic macros cannot occur with traditional cpp. */
88 struct cpp_macro
90 /* Parameters, if any. */
91 cpp_hashnode **params;
93 /* Replacement tokens (ISO) or replacement text (traditional). See
94 comment at top of cpptrad.c for how traditional function-like
95 macros are encoded. */
96 union
98 cpp_token *tokens;
99 const uchar *text;
100 } exp;
102 /* Definition line number. */
103 unsigned int line;
105 /* Number of tokens in expansion, or bytes for traditional macros. */
106 unsigned int count;
108 /* Number of parameters. */
109 unsigned short paramc;
111 /* If a function-like macro. */
112 unsigned int fun_like : 1;
114 /* If a variadic macro. */
115 unsigned int variadic : 1;
117 /* If macro defined in system header. */
118 unsigned int syshdr : 1;
120 /* Nonzero if it has been expanded or had its existence tested. */
121 unsigned int used : 1;
124 #define _cpp_mark_macro_used(NODE) do { \
125 if ((NODE)->type == NT_MACRO && !((NODE)->flags & NODE_BUILTIN)) \
126 (NODE)->value.macro->used = 1; } while (0)
128 /* A generic memory buffer, and operations on it. */
129 typedef struct _cpp_buff _cpp_buff;
130 struct _cpp_buff
132 struct _cpp_buff *next;
133 unsigned char *base, *cur, *limit;
136 extern _cpp_buff *_cpp_get_buff (cpp_reader *, size_t);
137 extern void _cpp_release_buff (cpp_reader *, _cpp_buff *);
138 extern void _cpp_extend_buff (cpp_reader *, _cpp_buff **, size_t);
139 extern _cpp_buff *_cpp_append_extend_buff (cpp_reader *, _cpp_buff *, size_t);
140 extern void _cpp_free_buff (_cpp_buff *);
141 extern unsigned char *_cpp_aligned_alloc (cpp_reader *, size_t);
142 extern unsigned char *_cpp_unaligned_alloc (cpp_reader *, size_t);
144 #define BUFF_ROOM(BUFF) (size_t) ((BUFF)->limit - (BUFF)->cur)
145 #define BUFF_FRONT(BUFF) ((BUFF)->cur)
146 #define BUFF_LIMIT(BUFF) ((BUFF)->limit)
148 /* #include types. */
149 enum include_type {IT_INCLUDE, IT_INCLUDE_NEXT, IT_IMPORT, IT_CMDLINE};
151 union utoken
153 const cpp_token *token;
154 const cpp_token **ptoken;
157 /* A "run" of tokens; part of a chain of runs. */
158 typedef struct tokenrun tokenrun;
159 struct tokenrun
161 tokenrun *next, *prev;
162 cpp_token *base, *limit;
165 /* Accessor macros for struct cpp_context. */
166 #define FIRST(c) ((c)->u.iso.first)
167 #define LAST(c) ((c)->u.iso.last)
168 #define CUR(c) ((c)->u.trad.cur)
169 #define RLIMIT(c) ((c)->u.trad.rlimit)
171 typedef struct cpp_context cpp_context;
172 struct cpp_context
174 /* Doubly-linked list. */
175 cpp_context *next, *prev;
177 union
179 /* For ISO macro expansion. Contexts other than the base context
180 are contiguous tokens. e.g. macro expansions, expanded
181 argument tokens. */
182 struct
184 union utoken first;
185 union utoken last;
186 } iso;
188 /* For traditional macro expansion. */
189 struct
191 const uchar *cur;
192 const uchar *rlimit;
193 } trad;
194 } u;
196 /* If non-NULL, a buffer used for storage related to this context.
197 When the context is popped, the buffer is released. */
198 _cpp_buff *buff;
200 /* For a macro context, the macro node, otherwise NULL. */
201 cpp_hashnode *macro;
203 /* True if utoken element is token, else ptoken. */
204 bool direct_p;
207 struct lexer_state
209 /* Nonzero if first token on line is CPP_HASH. */
210 unsigned char in_directive;
212 /* Nonzero if in a directive that will handle padding tokens itself.
213 #include needs this to avoid problems with computed include and
214 spacing between tokens. */
215 unsigned char directive_wants_padding;
217 /* True if we are skipping a failed conditional group. */
218 unsigned char skipping;
220 /* Nonzero if in a directive that takes angle-bracketed headers. */
221 unsigned char angled_headers;
223 /* Nonzero if in a #if or #elif directive. */
224 unsigned char in_expression;
226 /* Nonzero to save comments. Turned off if discard_comments, and in
227 all directives apart from #define. */
228 unsigned char save_comments;
230 /* Nonzero if lexing __VA_ARGS__ is valid. */
231 unsigned char va_args_ok;
233 /* Nonzero if lexing poisoned identifiers is valid. */
234 unsigned char poisoned_ok;
236 /* Nonzero to prevent macro expansion. */
237 unsigned char prevent_expansion;
239 /* Nonzero when parsing arguments to a function-like macro. */
240 unsigned char parsing_args;
242 /* Nonzero to skip evaluating part of an expression. */
243 unsigned int skip_eval;
246 /* Special nodes - identifiers with predefined significance. */
247 struct spec_nodes
249 cpp_hashnode *n_defined; /* defined operator */
250 cpp_hashnode *n_true; /* C++ keyword true */
251 cpp_hashnode *n_false; /* C++ keyword false */
252 cpp_hashnode *n__VA_ARGS__; /* C99 vararg macros */
255 typedef struct _cpp_line_note _cpp_line_note;
256 struct _cpp_line_note
258 /* Location in the clean line the note refers to. */
259 const uchar *pos;
261 /* Type of note. The 9 'from' trigraph characters represent those
262 trigraphs, '\\' an escaped newline, ' ' an escaped newline with
263 intervening space, and anything else is invalid. */
264 unsigned int type;
267 /* Represents the contents of a file cpplib has read in. */
268 struct cpp_buffer
270 const uchar *cur; /* Current location. */
271 const uchar *line_base; /* Start of current physical line. */
272 const uchar *next_line; /* Start of to-be-cleaned logical line. */
274 const uchar *buf; /* Entire character buffer. */
275 const uchar *rlimit; /* Writable byte at end of file. */
277 _cpp_line_note *notes; /* Array of notes. */
278 unsigned int cur_note; /* Next note to process. */
279 unsigned int notes_used; /* Number of notes. */
280 unsigned int notes_cap; /* Size of allocated array. */
282 struct cpp_buffer *prev;
284 /* Pointer into the include table; non-NULL if this is a file
285 buffer. Used for include_next and to record control macros. */
286 struct include_file *inc;
288 /* Value of if_stack at start of this file.
289 Used to prohibit unmatched #endif (etc) in an include file. */
290 struct if_stack *if_stack;
292 /* True if we need to get the next clean line. */
293 bool need_line;
295 /* True if we have already warned about C++ comments in this file.
296 The warning happens only for C89 extended mode with -pedantic on,
297 or for -Wtraditional, and only once per file (otherwise it would
298 be far too noisy). */
299 unsigned char warned_cplusplus_comments;
301 /* True if we don't process trigraphs and escaped newlines. True
302 for preprocessed input, command line directives, and _Pragma
303 buffers. */
304 unsigned char from_stage3;
306 /* Nonzero means that the directory to start searching for ""
307 include files has been calculated and stored in "dir" below. */
308 unsigned char search_cached;
310 /* At EOF, a buffer is automatically popped. If RETURN_AT_EOF is
311 true, a CPP_EOF token is then returned. Otherwise, the next
312 token from the enclosing buffer is returned. */
313 bool return_at_eof;
315 /* The directory of the this buffer's file. Its NAME member is not
316 allocated, so we don't need to worry about freeing it. */
317 struct cpp_path dir;
319 /* Used for buffer overlays by cpptrad.c. */
320 const uchar *saved_cur, *saved_rlimit;
323 /* A cpp_reader encapsulates the "state" of a pre-processor run.
324 Applying cpp_get_token repeatedly yields a stream of pre-processor
325 tokens. Usually, there is only one cpp_reader object active. */
326 struct cpp_reader
328 /* Top of buffer stack. */
329 cpp_buffer *buffer;
331 /* Overlaid buffer (can be different after processing #include). */
332 cpp_buffer *overlaid_buffer;
334 /* Lexer state. */
335 struct lexer_state state;
337 /* Source line tracking. */
338 struct line_maps line_maps;
339 const struct line_map *map;
340 unsigned int line;
342 /* The line of the '#' of the current directive. */
343 unsigned int directive_line;
345 /* Memory buffers. */
346 _cpp_buff *a_buff; /* Aligned permanent storage. */
347 _cpp_buff *u_buff; /* Unaligned permanent storage. */
348 _cpp_buff *free_buffs; /* Free buffer chain. */
350 /* Context stack. */
351 struct cpp_context base_context;
352 struct cpp_context *context;
354 /* If in_directive, the directive if known. */
355 const struct directive *directive;
357 /* Search paths for include files. */
358 struct cpp_path *quote_include; /* "" */
359 struct cpp_path *bracket_include; /* <> */
361 /* Multiple include optimization. */
362 const cpp_hashnode *mi_cmacro;
363 const cpp_hashnode *mi_ind_cmacro;
364 bool mi_valid;
366 /* Lexing. */
367 cpp_token *cur_token;
368 tokenrun base_run, *cur_run;
369 unsigned int lookaheads;
371 /* Nonzero prevents the lexer from re-using the token runs. */
372 unsigned int keep_tokens;
374 /* Error counter for exit code. */
375 unsigned int errors;
377 /* Buffer to hold macro definition string. */
378 unsigned char *macro_buffer;
379 unsigned int macro_buffer_len;
381 /* Descriptor for converting from the source character set to the
382 execution character set. */
383 struct cset_converter narrow_cset_desc;
385 /* Descriptor for converting from the source character set to the
386 wide execution character set. */
387 struct cset_converter wide_cset_desc;
389 /* Tree of other included files. See cppfiles.c. */
390 struct splay_tree_s *all_include_files;
392 /* Current maximum length of directory names in the search path
393 for include files. (Altered as we get more of them.) */
394 unsigned int max_include_len;
396 /* Date and time text. Calculated together if either is requested. */
397 const uchar *date;
398 const uchar *time;
400 /* EOF token, and a token forcing paste avoidance. */
401 cpp_token avoid_paste;
402 cpp_token eof;
404 /* Opaque handle to the dependencies of mkdeps.c. */
405 struct deps *deps;
407 /* Obstack holding all macro hash nodes. This never shrinks.
408 See cpphash.c */
409 struct obstack hash_ob;
411 /* Obstack holding buffer and conditional structures. This is a
412 real stack. See cpplib.c. */
413 struct obstack buffer_ob;
415 /* Pragma table - dynamic, because a library user can add to the
416 list of recognized pragmas. */
417 struct pragma_entry *pragmas;
419 /* Call backs to cpplib client. */
420 struct cpp_callbacks cb;
422 /* Identifier hash table. */
423 struct ht *hash_table;
425 /* Expression parser stack. */
426 struct op *op_stack, *op_limit;
428 /* User visible options. */
429 struct cpp_options opts;
431 /* Special nodes - identifiers with predefined significance to the
432 preprocessor. */
433 struct spec_nodes spec_nodes;
435 /* Nonzero means don't look for #include "foo" the source-file
436 directory. */
437 unsigned char quote_ignores_source_dir;
439 /* Whether cpplib owns the hashtable. */
440 unsigned char our_hashtable;
442 /* Traditional preprocessing output buffer (a logical line). */
443 struct
445 uchar *base;
446 uchar *limit;
447 uchar *cur;
448 unsigned int first_line;
449 } out;
451 /* Used to save the original line number during traditional
452 preprocessing. */
453 unsigned int saved_line;
455 /* A saved list of the defined macros, for dependency checking
456 of precompiled headers. */
457 struct cpp_savedstate *savedstate;
460 /* Character classes. Based on the more primitive macros in safe-ctype.h.
461 If the definition of `numchar' looks odd to you, please look up the
462 definition of a pp-number in the C standard [section 6.4.8 of C99].
464 In the unlikely event that characters other than \r and \n enter
465 the set is_vspace, the macro handle_newline() in cpplex.c must be
466 updated. */
467 #define _dollar_ok(x) ((x) == '$' && CPP_OPTION (pfile, dollars_in_ident))
469 #define is_idchar(x) (ISIDNUM(x) || _dollar_ok(x))
470 #define is_numchar(x) ISIDNUM(x)
471 #define is_idstart(x) (ISIDST(x) || _dollar_ok(x))
472 #define is_numstart(x) ISDIGIT(x)
473 #define is_hspace(x) ISBLANK(x)
474 #define is_vspace(x) IS_VSPACE(x)
475 #define is_nvspace(x) IS_NVSPACE(x)
476 #define is_space(x) IS_SPACE_OR_NUL(x)
478 /* This table is constant if it can be initialized at compile time,
479 which is the case if cpp was compiled with GCC >=2.7, or another
480 compiler that supports C99. */
481 #if HAVE_DESIGNATED_INITIALIZERS
482 extern const unsigned char _cpp_trigraph_map[UCHAR_MAX + 1];
483 #else
484 extern unsigned char _cpp_trigraph_map[UCHAR_MAX + 1];
485 #endif
487 /* Macros. */
489 #define CPP_IN_SYSTEM_HEADER(PFILE) ((PFILE)->map && (PFILE)->map->sysp)
490 #define CPP_PEDANTIC(PF) CPP_OPTION (PF, pedantic)
491 #define CPP_WTRADITIONAL(PF) CPP_OPTION (PF, warn_traditional)
493 /* In cpperror.c */
494 extern int _cpp_begin_message (cpp_reader *, int, unsigned int, unsigned int);
496 /* In cppmacro.c */
497 extern void _cpp_free_definition (cpp_hashnode *);
498 extern bool _cpp_create_definition (cpp_reader *, cpp_hashnode *);
499 extern void _cpp_pop_context (cpp_reader *);
500 extern void _cpp_push_text_context (cpp_reader *, cpp_hashnode *,
501 const uchar *, size_t);
502 extern bool _cpp_save_parameter (cpp_reader *, cpp_macro *, cpp_hashnode *);
503 extern bool _cpp_arguments_ok (cpp_reader *, cpp_macro *, const cpp_hashnode *,
504 unsigned int);
505 extern const uchar *_cpp_builtin_macro_text (cpp_reader *, cpp_hashnode *);
506 int _cpp_warn_if_unused_macro (cpp_reader *, cpp_hashnode *, void *);
507 /* In cpphash.c */
508 extern void _cpp_init_hashtable (cpp_reader *, hash_table *);
509 extern void _cpp_destroy_hashtable (cpp_reader *);
511 /* In cppfiles.c */
512 extern void _cpp_fake_include (cpp_reader *, const char *);
513 extern void _cpp_never_reread (struct include_file *);
514 extern bool _cpp_read_file (cpp_reader *, const char *);
515 extern bool _cpp_execute_include (cpp_reader *, const char *, int,
516 enum include_type);
517 extern int _cpp_compare_file_date (cpp_reader *, const char *, int);
518 extern void _cpp_report_missing_guards (cpp_reader *);
519 extern void _cpp_init_includes (cpp_reader *);
520 extern void _cpp_cleanup_includes (cpp_reader *);
521 extern void _cpp_pop_file_buffer (cpp_reader *, struct include_file *);
523 /* In cppexp.c */
524 extern bool _cpp_parse_expr (cpp_reader *);
525 extern struct op *_cpp_expand_op_stack (cpp_reader *);
527 /* In cpplex.c */
528 extern void _cpp_process_line_notes (cpp_reader *, int);
529 extern void _cpp_clean_line (cpp_reader *);
530 extern bool _cpp_get_fresh_line (cpp_reader *);
531 extern bool _cpp_skip_block_comment (cpp_reader *);
532 extern cpp_token *_cpp_temp_token (cpp_reader *);
533 extern const cpp_token *_cpp_lex_token (cpp_reader *);
534 extern cpp_token *_cpp_lex_direct (cpp_reader *);
535 extern int _cpp_equiv_tokens (const cpp_token *, const cpp_token *);
536 extern void _cpp_init_tokenrun (tokenrun *, unsigned int);
538 /* In cppinit.c. */
539 extern void _cpp_maybe_push_include_file (cpp_reader *);
541 /* In cpplib.c */
542 extern int _cpp_test_assertion (cpp_reader *, unsigned int *);
543 extern int _cpp_handle_directive (cpp_reader *, int);
544 extern void _cpp_define_builtin (cpp_reader *, const char *);
545 extern char ** _cpp_save_pragma_names (cpp_reader *);
546 extern void _cpp_restore_pragma_names (cpp_reader *, char **);
547 extern void _cpp_do__Pragma (cpp_reader *);
548 extern void _cpp_init_directives (cpp_reader *);
549 extern void _cpp_init_internal_pragmas (cpp_reader *);
550 extern void _cpp_do_file_change (cpp_reader *, enum lc_reason, const char *,
551 unsigned int, unsigned int);
552 extern void _cpp_pop_buffer (cpp_reader *);
554 /* In cpptrad.c. */
555 extern bool scan_out_logical_line (cpp_reader *, cpp_macro *);
556 extern bool _cpp_read_logical_line_trad (cpp_reader *);
557 extern void _cpp_overlay_buffer (cpp_reader *pfile, const uchar *, size_t);
558 extern void _cpp_remove_overlay (cpp_reader *);
559 extern bool _cpp_create_trad_definition (cpp_reader *, cpp_macro *);
560 extern bool _cpp_expansions_different_trad (const cpp_macro *,
561 const cpp_macro *);
562 extern uchar *_cpp_copy_replacement_text (const cpp_macro *, uchar *);
563 extern size_t _cpp_replacement_text_len (const cpp_macro *);
565 /* In cppcharset.c. */
566 extern cppchar_t _cpp_valid_ucn (cpp_reader *, const uchar **,
567 const uchar *, int);
568 extern void _cpp_destroy_iconv (cpp_reader *);
569 extern bool _cpp_interpret_string_notranslate (cpp_reader *, const cpp_string *,
570 cpp_string *);
572 /* Utility routines and macros. */
573 #define DSC(str) (const uchar *)str, sizeof str - 1
574 #define xnew(T) (T *) xmalloc (sizeof(T))
575 #define xcnew(T) (T *) xcalloc (1, sizeof(T))
576 #define xnewvec(T, N) (T *) xmalloc (sizeof(T) * (N))
577 #define xcnewvec(T, N) (T *) xcalloc (N, sizeof(T))
578 #define xobnew(O, T) (T *) obstack_alloc (O, sizeof(T))
580 /* These are inline functions instead of macros so we can get type
581 checking. */
582 static inline int ustrcmp (const uchar *, const uchar *);
583 static inline int ustrncmp (const uchar *, const uchar *, size_t);
584 static inline size_t ustrlen (const uchar *);
585 static inline uchar *uxstrdup (const uchar *);
586 static inline uchar *ustrchr (const uchar *, int);
587 static inline int ufputs (const uchar *, FILE *);
589 static inline int
590 ustrcmp (const uchar *s1, const uchar *s2)
592 return strcmp ((const char *)s1, (const char *)s2);
595 static inline int
596 ustrncmp (const uchar *s1, const uchar *s2, size_t n)
598 return strncmp ((const char *)s1, (const char *)s2, n);
601 static inline size_t
602 ustrlen (const uchar *s1)
604 return strlen ((const char *)s1);
607 static inline uchar *
608 uxstrdup (const uchar *s1)
610 return (uchar *) xstrdup ((const char *)s1);
613 static inline uchar *
614 ustrchr (const uchar *s1, int c)
616 return (uchar *) strchr ((const char *)s1, c);
619 static inline int
620 ufputs (const uchar *s, FILE *f)
622 return fputs ((const char *)s, f);
625 #endif /* ! GCC_CPPHASH_H */