/* Part of CPP library.
- Copyright (C) 1997, 1998, 1999, 2000 Free Software Foundation, Inc.
+ Copyright (C) 1997, 1998, 1999, 2000, 2001 Free Software Foundation, Inc.
This program is free software; you can redistribute it and/or modify it
under the terms of the GNU General Public License as published by the
that need to be visible across files. It's called cpphash.h for
historical reasons. */
-#ifndef __GCC_CPPHASH__
-#define __GCC_CPPHASH__
+#ifndef GCC_CPPHASH_H
+#define GCC_CPPHASH_H
+
+#include "hashtable.h"
+
+struct directive; /* Deliberately incomplete. */
/* Test if a sign is valid within a preprocessing number. */
#define VALID_SIGN(c, prevc) \
|| (((prevc) == 'p' || (prevc) == 'P') \
&& CPP_OPTION (pfile, extended_numbers))))
+#define CPP_OPTION(PFILE, OPTION) ((PFILE)->opts.OPTION)
+#define CPP_BUFFER(PFILE) ((PFILE)->buffer)
+#define CPP_BUF_COLUMN(BUF, CUR) ((CUR) - (BUF)->line_base + (BUF)->col_adjust)
+#define CPP_BUF_COL(BUF) CPP_BUF_COLUMN(BUF, (BUF)->cur)
+
+/* Maximum nesting of cpp_buffers. We use a static limit, partly for
+ efficiency, and partly to limit runaway recursion. */
+#define CPP_STACK_MAX 200
+
/* Memory pools. */
-#define ALIGN(size, align) (((size) + ((align) - 1)) & ~((align) - 1))
+#define POOL_ALIGN(size, align) (((size) + ((align) - 1)) & ~((align) - 1))
#define POOL_FRONT(p) ((p)->cur->front)
#define POOL_LIMIT(p) ((p)->cur->limit)
#define POOL_BASE(p) ((p)->cur->base)
#define POOL_SIZE(p) ((p)->cur->limit - (p)->cur->base)
#define POOL_ROOM(p) ((p)->cur->limit - (p)->cur->front)
#define POOL_USED(p) ((p)->cur->front - (p)->cur->base)
-#define POOL_COMMIT(p, len) do {((p)->cur->front += ALIGN (len, (p)->align));\
+#define POOL_COMMIT(p, len) do {\
+ ((p)->cur->front += POOL_ALIGN (len, (p)->align));\
if ((p)->cur->front > (p)->cur->limit) abort ();} while (0)
typedef struct cpp_chunk cpp_chunk;
unsigned char *base;
};
-/* List of directories to look for include files in. */
-struct file_name_list
+typedef struct cpp_pool cpp_pool;
+struct cpp_pool
+{
+ struct cpp_chunk *cur, *locked;
+ unsigned char *pos; /* Current position. */
+ unsigned int align;
+ unsigned int locks;
+};
+
+/* List of directories to look for include files in. */
+struct search_path
{
- struct file_name_list *next;
- struct file_name_list *alloc; /* for the cache of
- current directory entries */
- char *name;
- unsigned int nlen;
+ struct search_path *next;
+
+ /* NOTE: NAME may not be null terminated for the case of the current
+ file's directory! */
+ const char *name;
+ unsigned int len;
/* We use these to tell if the directory mentioned here is a duplicate
- of an earlier directory on the search path. */
+ of an earlier directory on the search path. */
ino_t ino;
dev_t dev;
- /* If the following is nonzero, it is a C-language system include
- directory. */
+ /* Non-zero if it is a system include directory. */
int sysp;
- /* Mapping of file names for this directory.
- Only used on MS-DOS and related platforms. */
+ /* Mapping of file names for this directory. Only used on MS-DOS
+ and related platforms. */
struct file_name_map *name_map;
};
-#define ABSOLUTE_PATH ((struct file_name_list *)-1)
-/* This structure is used for the table of all includes. */
-struct include_file
+/* #include types. */
+enum include_type {IT_INCLUDE, IT_INCLUDE_NEXT, IT_IMPORT, IT_CMDLINE};
+
+typedef struct toklist toklist;
+struct toklist
+{
+ cpp_token *first;
+ cpp_token *limit;
+};
+
+typedef struct cpp_context cpp_context;
+struct cpp_context
+{
+ /* Doubly-linked list. */
+ cpp_context *next, *prev;
+
+ /* Contexts other than the base context are contiguous tokens.
+ e.g. macro expansions, expanded argument tokens. */
+ struct toklist list;
+
+ /* For a macro context, these are the macro and its arguments. */
+ cpp_macro *macro;
+};
+
+struct lexer_state
+{
+ /* Nonzero if first token on line is CPP_HASH. */
+ unsigned char in_directive;
+
+ /* True if we are skipping a failed conditional group. */
+ unsigned char skipping;
+
+ /* Nonzero if in a directive that takes angle-bracketed headers. */
+ unsigned char angled_headers;
+
+ /* Nonzero to save comments. Turned off if discard_comments, and in
+ all directives apart from #define. */
+ unsigned char save_comments;
+
+ /* If nonzero the next token is at the beginning of the line. */
+ unsigned char next_bol;
+
+ /* Nonzero if we're mid-comment. */
+ unsigned char lexing_comment;
+
+ /* Nonzero if lexing __VA_ARGS__ is valid. */
+ unsigned char va_args_ok;
+
+ /* Nonzero if lexing poisoned identifiers is valid. */
+ unsigned char poisoned_ok;
+
+ /* Nonzero to prevent macro expansion. */
+ unsigned char prevent_expansion;
+
+ /* Nonzero when parsing arguments to a function-like macro. */
+ unsigned char parsing_args;
+
+ /* Nonzero when in a # NUMBER directive. */
+ unsigned char line_extension;
+};
+
+/* Special nodes - identifiers with predefined significance. */
+struct spec_nodes
+{
+ cpp_hashnode *n_L; /* L"str" */
+ cpp_hashnode *n_defined; /* defined operator */
+ cpp_hashnode *n_true; /* C++ keyword true */
+ cpp_hashnode *n_false; /* C++ keyword false */
+ cpp_hashnode *n__Pragma; /* _Pragma operator */
+ cpp_hashnode *n__STRICT_ANSI__; /* STDC_0_IN_SYSTEM_HEADERS */
+ cpp_hashnode *n__CHAR_UNSIGNED__; /* plain char is unsigned */
+ cpp_hashnode *n__VA_ARGS__; /* C99 vararg macros */
+};
+
+struct cpp_buffer
+{
+ const unsigned char *cur; /* current position */
+ const unsigned char *rlimit; /* end of valid data */
+ const unsigned char *line_base; /* start of current line */
+ cppchar_t read_ahead; /* read ahead character */
+ cppchar_t extra_char; /* extra read-ahead for long tokens. */
+
+ struct cpp_reader *pfile; /* Owns this buffer. */
+ struct cpp_buffer *prev;
+
+ const unsigned char *buf; /* entire buffer */
+
+ /* Filename specified with #line command. */
+ const char *nominal_fname;
+
+ /* Pointer into the include table. Used for include_next and
+ to record control macros. */
+ struct include_file *inc;
+
+ /* Value of if_stack at start of this file.
+ Used to prohibit unmatched #endif (etc) in an include file. */
+ struct if_stack *if_stack;
+
+ /* Token column position adjustment owing to tabs in whitespace. */
+ unsigned int col_adjust;
+
+ /* The line of the buffer that we return to after a #include.
+ Strictly this is redundant, since it can be calculated from the
+ line maps, but it is clearest to save it here. */
+ unsigned int return_to_line;
+
+ /* Contains PREV_WHITE and/or AVOID_LPASTE. */
+ unsigned char saved_flags;
+
+ /* Because of the way the lexer works, -Wtrigraphs can sometimes
+ warn twice for the same trigraph. This helps prevent that. */
+ const unsigned char *last_Wtrigraphs;
+
+ /* True if we have already warned about C++ comments in this file.
+ The warning happens only for C89 extended mode with -pedantic on,
+ or for -Wtraditional, and only once per file (otherwise it would
+ be far too noisy). */
+ unsigned char warned_cplusplus_comments;
+
+ /* True if we don't process trigraphs and escaped newlines. True
+ for preprocessed input, command line directives, and _Pragma
+ buffers. */
+ unsigned char from_stage3;
+
+ /* 1 = system header file, 2 = C system header file used for C++. */
+ unsigned char sysp;
+
+ /* Nonzero means we have printed (while error reporting) a list of
+ containing files that matches the current status. */
+ unsigned char include_stack_listed;
+
+ /* Nonzero means that the directory to start searching for ""
+ include files has been calculated and stored in "dir" below. */
+ unsigned char search_cached;
+
+ /* At EOF, a buffer is automatically popped. If RETURN_AT_EOF is
+ true, a CPP_EOF token is then returned. Otherwise, the next
+ token from the enclosing buffer is returned. */
+ bool return_at_eof;
+
+ /* Buffer type. */
+ ENUM_BITFIELD (cpp_buffer_type) type : 8;
+
+ /* The directory of the this buffer's file. Its NAME member is not
+ allocated, so we don't need to worry about freeing it. */
+ struct search_path dir;
+};
+
+/* A cpp_reader encapsulates the "state" of a pre-processor run.
+ Applying cpp_get_token repeatedly yields a stream of pre-processor
+ tokens. Usually, there is only one cpp_reader object active. */
+
+struct cpp_reader
{
- const char *name; /* actual path name of file */
- const cpp_hashnode *cmacro; /* macro, if any, preventing reinclusion. */
- const struct file_name_list *foundhere;
- /* location in search path where file was
- found, for #include_next */
- const unsigned char *buffer; /* pointer to cached file contents */
- struct stat st; /* copy of stat(2) data for file */
- int fd; /* fd open on file (short term storage only) */
- unsigned short include_count; /* number of times file has been read */
- unsigned short refcnt; /* number of stacked buffers using this file */
- unsigned char sysp; /* file is a system header */
- unsigned char mapped; /* file buffer is mmapped */
- unsigned char defined; /* cmacro prevents inclusion in this state */
+ /* Top of buffer stack. */
+ cpp_buffer *buffer;
+
+ /* Lexer state. */
+ struct lexer_state state;
+
+ /* Source line tracking. */
+ struct line_maps line_maps;
+ struct line_map *map;
+ unsigned int line;
+
+ /* The position of the last lexed token and last lexed directive. */
+ cpp_lexer_pos lexer_pos;
+ cpp_lexer_pos directive_pos;
+ unsigned int directive_line;
+
+ /* Memory pools. */
+ cpp_pool ident_pool; /* For all identifiers, and permanent
+ numbers and strings. */
+ cpp_pool macro_pool; /* For macro definitions. Permanent. */
+ cpp_pool argument_pool; /* For macro arguments. Temporary. */
+
+ /* Context stack. */
+ struct cpp_context base_context;
+ struct cpp_context *context;
+
+ /* If in_directive, the directive if known. */
+ const struct directive *directive;
+
+ /* Multiple inlcude optimisation. */
+ const cpp_hashnode *mi_cmacro;
+ const cpp_hashnode *mi_ind_cmacro;
+ bool mi_valid;
+
+ /* Token lookahead. */
+ struct cpp_lookahead *la_read; /* Read from this lookahead. */
+ struct cpp_lookahead *la_write; /* Write to this lookahead. */
+ struct cpp_lookahead *la_unused; /* Free store. */
+ struct cpp_lookahead *la_saved; /* Backup when entering directive. */
+
+ /* Error counter for exit code. */
+ unsigned int errors;
+
+ /* Line and column where a newline was first seen in a string
+ constant (multi-line strings). */
+ cpp_lexer_pos mlstring_pos;
+
+ /* Buffer to hold macro definition string. */
+ unsigned char *macro_buffer;
+ unsigned int macro_buffer_len;
+
+ /* Current depth in #include directives that use <...>. */
+ unsigned int system_include_depth;
+
+ /* Current depth of buffer stack. */
+ unsigned int buffer_stack_depth;
+
+ /* Current depth in #include directives. */
+ unsigned int include_depth;
+
+ /* Tree of other included files. See cppfiles.c. */
+ struct splay_tree_s *all_include_files;
+
+ /* Current maximum length of directory names in the search path
+ for include files. (Altered as we get more of them.) */
+ unsigned int max_include_len;
+
+ /* Date and time tokens. Calculated together if either is requested. */
+ cpp_token date;
+ cpp_token time;
+
+ /* Opaque handle to the dependencies of mkdeps.c. Used by -M etc. */
+ struct deps *deps;
+
+ /* Obstack holding all macro hash nodes. This never shrinks.
+ See cpphash.c */
+ struct obstack hash_ob;
+
+ /* Obstack holding buffer and conditional structures. This is a
+ real stack. See cpplib.c. */
+ struct obstack buffer_ob;
+
+ /* Pragma table - dynamic, because a library user can add to the
+ list of recognized pragmas. */
+ struct pragma_entry *pragmas;
+
+ /* Call backs. */
+ struct cpp_callbacks cb;
+
+ /* Identifier hash table. */
+ struct ht *hash_table;
+
+ /* User visible options. */
+ struct cpp_options opts;
+
+ /* Special nodes - identifiers with predefined significance to the
+ preprocessor. */
+ struct spec_nodes spec_nodes;
+
+ /* Whether to print our version number. Done this way so
+ we don't get it twice for -v -version. */
+ unsigned char print_version;
+
+ /* Whether cpplib owns the hashtable. */
+ unsigned char our_hashtable;
};
-/* The cmacro works like this: If it's NULL, the file is to be
- included again. If it's NEVER_REREAD, the file is never to be
- included again. Otherwise it is a macro hashnode, and the file is
- to be included again if the macro is defined or not as specified by
- DEFINED. */
-#define NEVER_REREAD ((const cpp_hashnode *)-1)
-#define DO_NOT_REREAD(inc) \
-((inc)->cmacro && ((inc)->cmacro == NEVER_REREAD \
- || ((inc)->cmacro->type == NT_MACRO) == (inc)->defined))
-
-/* Character classes.
+/* Character classes. Based on the more primitive macros in safe-ctype.h.
If the definition of `numchar' looks odd to you, please look up the
definition of a pp-number in the C standard [section 6.4.8 of C99].
In the unlikely event that characters other than \r and \n enter
the set is_vspace, the macro handle_newline() in cpplex.c must be
updated. */
-#define ISidnum 0x01 /* a-zA-Z0-9_ */
-#define ISidstart 0x02 /* _a-zA-Z */
-#define ISnumstart 0x04 /* 0-9 */
-#define IShspace 0x08 /* ' ' \t */
-#define ISvspace 0x10 /* \r \n */
-#define ISspace 0x20 /* ' ' \t \r \n \f \v \0 */
-
#define _dollar_ok(x) ((x) == '$' && CPP_OPTION (pfile, dollars_in_ident))
-#define is_idchar(x) ((_cpp_IStable[x] & ISidnum) || _dollar_ok(x))
-#define is_idstart(x) ((_cpp_IStable[x] & ISidstart) || _dollar_ok(x))
-#define is_numchar(x) (_cpp_IStable[x] & ISidnum)
-#define is_numstart(x) (_cpp_IStable[x] & ISnumstart)
-#define is_hspace(x) (_cpp_IStable[x] & IShspace)
-#define is_vspace(x) (_cpp_IStable[x] & ISvspace)
-#define is_nvspace(x) ((_cpp_IStable[x] & (ISspace | ISvspace)) == ISspace)
-#define is_space(x) (_cpp_IStable[x] & ISspace)
+#define is_idchar(x) (ISIDNUM(x) || _dollar_ok(x))
+#define is_numchar(x) ISIDNUM(x)
+#define is_idstart(x) (ISIDST(x) || _dollar_ok(x))
+#define is_numstart(x) ISDIGIT(x)
+#define is_hspace(x) ISBLANK(x)
+#define is_vspace(x) IS_VSPACE(x)
+#define is_nvspace(x) IS_NVSPACE(x)
+#define is_space(x) IS_SPACE_OR_NUL(x)
-/* These tables are constant if they can be initialized at compile time,
+/* This table is constant if it can be initialized at compile time,
which is the case if cpp was compiled with GCC >=2.7, or another
compiler that supports C99. */
#if HAVE_DESIGNATED_INITIALIZERS
-extern const unsigned char _cpp_IStable[UCHAR_MAX + 1];
extern const unsigned char _cpp_trigraph_map[UCHAR_MAX + 1];
#else
-extern unsigned char _cpp_IStable[UCHAR_MAX + 1];
extern unsigned char _cpp_trigraph_map[UCHAR_MAX + 1];
#endif
/* Macros. */
-#define CPP_PREV_BUFFER(BUFFER) ((BUFFER)->prev)
#define CPP_PRINT_DEPS(PFILE) CPP_OPTION (PFILE, print_deps)
#define CPP_IN_SYSTEM_HEADER(PFILE) \
- (CPP_BUFFER (PFILE) && CPP_BUFFER (PFILE)->inc \
- && CPP_BUFFER (PFILE)->inc->sysp)
-#define CPP_PEDANTIC(PF) \
- CPP_OPTION (PF, pedantic)
-#define CPP_WTRADITIONAL(PF) \
- CPP_OPTION (PF, warn_traditional)
-
-/* Hash step. The hash calculation is duplicated in cpp_lookup and
- parse_name. */
-#define HASHSTEP(r, c) ((r) * 67 + (c - 113));
+ (CPP_BUFFER (PFILE) && CPP_BUFFER (PFILE)->sysp)
+#define CPP_PEDANTIC(PF) CPP_OPTION (PF, pedantic)
+#define CPP_WTRADITIONAL(PF) CPP_OPTION (PF, warn_traditional)
/* In cpperror.c */
-enum error_type { WARNING = 0, PEDWARN, ERROR, FATAL, ICE };
+enum error_type { WARNING = 0, WARNING_SYSHDR, PEDWARN, ERROR, FATAL, ICE };
extern int _cpp_begin_message PARAMS ((cpp_reader *, enum error_type,
const char *, const cpp_lexer_pos *));
const cpp_lexer_pos *));
/* In cpphash.c */
-extern void _cpp_init_hashtable PARAMS ((cpp_reader *));
-extern void _cpp_cleanup_hashtable PARAMS ((cpp_reader *));
-extern cpp_hashnode *_cpp_lookup_with_hash PARAMS ((cpp_reader*, size_t,
- unsigned int));
+extern void _cpp_init_hashtable PARAMS ((cpp_reader *, hash_table *));
+extern void _cpp_destroy_hashtable PARAMS ((cpp_reader *));
/* In cppfiles.c */
-extern void _cpp_simplify_pathname PARAMS ((char *));
-extern void _cpp_execute_include PARAMS ((cpp_reader *,
- const cpp_token *, int,
- struct file_name_list *));
+extern void _cpp_fake_include PARAMS ((cpp_reader *, const char *));
+extern void _cpp_never_reread PARAMS ((struct include_file *));
+extern char *_cpp_simplify_pathname PARAMS ((char *));
+extern int _cpp_read_file PARAMS ((cpp_reader *, const char *));
+extern int _cpp_execute_include PARAMS ((cpp_reader *,
+ const cpp_token *,
+ enum include_type));
extern int _cpp_compare_file_date PARAMS ((cpp_reader *,
const cpp_token *));
extern void _cpp_report_missing_guards PARAMS ((cpp_reader *));
extern void _cpp_init_includes PARAMS ((cpp_reader *));
extern void _cpp_cleanup_includes PARAMS ((cpp_reader *));
-extern const char *_cpp_fake_include PARAMS ((cpp_reader *, const char *));
extern void _cpp_pop_file_buffer PARAMS ((cpp_reader *, cpp_buffer *));
/* In cppexp.c */
extern int _cpp_handle_directive PARAMS ((cpp_reader *, int));
extern void _cpp_define_builtin PARAMS ((cpp_reader *, const char *));
extern void _cpp_do__Pragma PARAMS ((cpp_reader *));
-extern void _cpp_init_stacks PARAMS ((cpp_reader *));
-extern void _cpp_cleanup_stacks PARAMS ((cpp_reader *));
+extern void _cpp_init_directives PARAMS ((cpp_reader *));
extern void _cpp_init_internal_pragmas PARAMS ((cpp_reader *));
+extern void _cpp_do_file_change PARAMS ((cpp_reader *, enum lc_reason,
+ unsigned int));
+extern void _cpp_pop_buffer PARAMS ((cpp_reader *));
/* Utility routines and macros. */
#define DSC(str) (const U_CHAR *)str, sizeof str - 1
#define xcnewvec(T, N) (T *) xcalloc (N, sizeof(T))
#define xobnew(O, T) (T *) obstack_alloc (O, sizeof(T))
-#endif
+/* These are inline functions instead of macros so we can get type
+ checking. */
+typedef unsigned char U_CHAR;
+#define U (const U_CHAR *) /* Intended use: U"string" */
+
+static inline int ustrcmp PARAMS ((const U_CHAR *, const U_CHAR *));
+static inline int ustrncmp PARAMS ((const U_CHAR *, const U_CHAR *,
+ size_t));
+static inline size_t ustrlen PARAMS ((const U_CHAR *));
+static inline U_CHAR *uxstrdup PARAMS ((const U_CHAR *));
+static inline U_CHAR *ustrchr PARAMS ((const U_CHAR *, int));
+static inline int ufputs PARAMS ((const U_CHAR *, FILE *));
+
+static inline int
+ustrcmp (s1, s2)
+ const U_CHAR *s1, *s2;
+{
+ return strcmp ((const char *)s1, (const char *)s2);
+}
+
+static inline int
+ustrncmp (s1, s2, n)
+ const U_CHAR *s1, *s2;
+ size_t n;
+{
+ return strncmp ((const char *)s1, (const char *)s2, n);
+}
+
+static inline size_t
+ustrlen (s1)
+ const U_CHAR *s1;
+{
+ return strlen ((const char *)s1);
+}
+
+static inline U_CHAR *
+uxstrdup (s1)
+ const U_CHAR *s1;
+{
+ return (U_CHAR *) xstrdup ((const char *)s1);
+}
+
+static inline U_CHAR *
+ustrchr (s1, c)
+ const U_CHAR *s1;
+ int c;
+{
+ return (U_CHAR *) strchr ((const char *)s1, c);
+}
+
+static inline int
+ufputs (s, f)
+ const U_CHAR *s;
+ FILE *f;
+{
+ return fputs ((const char *)s, f);
+}
+
+#endif /* ! GCC_CPPHASH_H */