diff options
author | Adhemerval Zanella <adhemerval.zanella@linaro.org> | 2017-12-20 09:47:44 -0200 |
---|---|---|
committer | Adhemerval Zanella <adhemerval.zanella@linaro.org> | 2018-07-04 09:54:45 -0300 |
commit | eb04c21373e2a2885f3d52ff192b0499afe3c672 (patch) | |
tree | 040d92dd6f342295f296df31de00ed9ace1b7d67 /posix/regex_internal.h | |
parent | Bug 23308: Update to Unicode 11.0.0 (diff) | |
download | glibc-eb04c21373e2a2885f3d52ff192b0499afe3c672.tar.gz glibc-eb04c21373e2a2885f3d52ff192b0499afe3c672.tar.bz2 glibc-eb04c21373e2a2885f3d52ff192b0499afe3c672.zip |
posix: Sync gnulib regex implementation
This patch syncs the regex implementation with gnulib (commit 0ee5212).
Only two changes in GLIBC regex testing are required:
1. posix/bug-regex28.c: as previously discussed [1] the change of
expected results on the pattern should be safe.
2. posix/PCRE.tests: the ERE (a)|\1 is malformed (in the sense that
the \1 doesn't mean anything) and although current GLIBC accepts
it has undefined behavior. This patch removes the specific test.
This sync contains some patches from thread 'Regex: Make libc regex
more usable outside GLIBC.' [2] which have been pushed upstream in
gnulib. This patches also fixes some regex issues (BZ #23233,
BZ #21163, BZ #18986, BZ #13762) and I did not add testcases for
both #23233 and #13762 because I couldn't think a simple way to
trigger the expected failure path to trigger them.
Checked on x86_64-linux-gnu and i686-linux-gnu.
[BZ #23233]
[BZ #21163]
[BZ #18986]
[BZ #13762]
* posix/Makefile (tests): Add bug-regex37 and bug-regex38.
* posix/PCRE.tests: Remove invalid test.
* posix/bug-regex28.c: Fix expected values for used syntax.
* posix/bug-regex37.c: New file.
* posix/bug-regex38.c: Likewise.
* posix/regcomp.c: Sync with gnulib.
* posix/regex.c: Likewise.
* posix/regex.h: Likewise.
* posix/regex_internal.c: Likewise.
* posix/regex_internal.h: Likewise.
* posix/regexec.c: Likewise.
[1] https://sourceware.org/ml/libc-alpha/2017-12/msg00807.html
[2] https://sourceware.org/ml/libc-alpha/2017-12/msg00237.html
Diffstat (limited to 'posix/regex_internal.h')
-rw-r--r-- | posix/regex_internal.h | 442 |
1 files changed, 299 insertions, 143 deletions
diff --git a/posix/regex_internal.h b/posix/regex_internal.h index 41bf2d371c..3b836ed206 100644 --- a/posix/regex_internal.h +++ b/posix/regex_internal.h @@ -15,7 +15,7 @@ You should have received a copy of the GNU Lesser General Public License along with the GNU C Library; if not, see - <http://www.gnu.org/licenses/>. */ + <https://www.gnu.org/licenses/>. */ #ifndef _REGEX_INTERNAL_H #define _REGEX_INTERNAL_H 1 @@ -26,35 +26,78 @@ #include <stdlib.h> #include <string.h> -#if defined HAVE_LANGINFO_H || defined HAVE_LANGINFO_CODESET || defined _LIBC -# include <langinfo.h> -#endif -#if defined HAVE_LOCALE_H || defined _LIBC -# include <locale.h> +#include <langinfo.h> +#include <locale.h> +#include <wchar.h> +#include <wctype.h> +#include <stdbool.h> +#include <stdint.h> + +/* Properties of integers. Although Gnulib has intprops.h, glibc does + without for now. */ +#ifndef _LIBC +# include "intprops.h" +#else +/* True if the real type T is signed. */ +# define TYPE_SIGNED(t) (! ((t) 0 < (t) -1)) + +/* True if adding the nonnegative Idx values A and B would overflow. + If false, set *R to A + B. A, B, and R may be evaluated more than + once, or zero times. Although this is not a full implementation of + Gnulib INT_ADD_WRAPV, it is good enough for glibc regex code. + FIXME: This implementation is a fragile stopgap, and this file would + be simpler and more robust if intprops.h were migrated into glibc. */ +# define INT_ADD_WRAPV(a, b, r) \ + (IDX_MAX - (a) < (b) ? true : (*(r) = (a) + (b), false)) #endif -#if defined HAVE_WCHAR_H || defined _LIBC -# include <wchar.h> -#endif /* HAVE_WCHAR_H || _LIBC */ -#if defined HAVE_WCTYPE_H || defined _LIBC -# include <wctype.h> -#endif /* HAVE_WCTYPE_H || _LIBC */ -#if defined HAVE_STDBOOL_H || defined _LIBC -# include <stdbool.h> -#endif /* HAVE_STDBOOL_H || _LIBC */ -#if defined HAVE_STDINT_H || defined _LIBC -# include <stdint.h> -#endif /* HAVE_STDINT_H || _LIBC */ -#if defined _LIBC + +#ifdef _LIBC # include <libc-lock.h> +# define lock_define(name) __libc_lock_define (, name) +# define lock_init(lock) (__libc_lock_init (lock), 0) +# define lock_fini(lock) ((void) 0) +# define lock_lock(lock) __libc_lock_lock (lock) +# define lock_unlock(lock) __libc_lock_unlock (lock) +#elif defined GNULIB_LOCK && !defined USE_UNLOCKED_IO +# include "glthread/lock.h" + /* Use gl_lock_define if empty macro arguments are known to work. + Otherwise, fall back on less-portable substitutes. */ +# if ((defined __GNUC__ && !defined __STRICT_ANSI__) \ + || (defined __STDC_VERSION__ && 199901L <= __STDC_VERSION__)) +# define lock_define(name) gl_lock_define (, name) +# elif USE_POSIX_THREADS +# define lock_define(name) pthread_mutex_t name; +# elif USE_PTH_THREADS +# define lock_define(name) pth_mutex_t name; +# elif USE_SOLARIS_THREADS +# define lock_define(name) mutex_t name; +# elif USE_WINDOWS_THREADS +# define lock_define(name) gl_lock_t name; +# else +# define lock_define(name) +# endif +# define lock_init(lock) glthread_lock_init (&(lock)) +# define lock_fini(lock) glthread_lock_destroy (&(lock)) +# define lock_lock(lock) glthread_lock_lock (&(lock)) +# define lock_unlock(lock) glthread_lock_unlock (&(lock)) +#elif defined GNULIB_PTHREAD && !defined USE_UNLOCKED_IO +# include <pthread.h> +# define lock_define(name) pthread_mutex_t name; +# define lock_init(lock) pthread_mutex_init (&(lock), 0) +# define lock_fini(lock) pthread_mutex_destroy (&(lock)) +# define lock_lock(lock) pthread_mutex_lock (&(lock)) +# define lock_unlock(lock) pthread_mutex_unlock (&(lock)) #else -# define __libc_lock_define(CLASS,NAME) -# define __libc_lock_init(NAME) do { } while (0) -# define __libc_lock_lock(NAME) do { } while (0) -# define __libc_lock_unlock(NAME) do { } while (0) +# define lock_define(name) +# define lock_init(lock) 0 +# define lock_fini(lock) ((void) 0) + /* The 'dfa' avoids an "unused variable 'dfa'" warning from GCC. */ +# define lock_lock(lock) ((void) dfa) +# define lock_unlock(lock) ((void) 0) #endif /* In case that the system doesn't have isblank(). */ -#if !defined _LIBC && !defined HAVE_ISBLANK && !defined isblank +#if !defined _LIBC && ! (defined isblank || (HAVE_ISBLANK && HAVE_DECL_ISBLANK)) # define isblank(ch) ((ch) == ' ' || (ch) == '\t') #endif @@ -75,6 +118,7 @@ __dcgettext (_libc_intl_domainname, msgid, LC_MESSAGES) # endif #else +# undef gettext # define gettext(msgid) (msgid) #endif @@ -84,23 +128,17 @@ # define gettext_noop(String) String #endif -/* For loser systems without the definition. */ -#ifndef SIZE_MAX -# define SIZE_MAX ((size_t) -1) -#endif - #if (defined MB_CUR_MAX && HAVE_WCTYPE_H && HAVE_ISWCTYPE) || _LIBC # define RE_ENABLE_I18N #endif -#if __GNUC__ >= 3 -# define BE(expr, val) __builtin_expect (expr, val) -#else -# define BE(expr, val) (expr) -#endif +#define BE(expr, val) __builtin_expect (expr, val) -/* Number of single byte character. */ -#define SBC_MAX 256 +/* Number of ASCII characters. */ +#define ASCII_CHARS 0x80 + +/* Number of single byte characters. */ +#define SBC_MAX (UCHAR_MAX + 1) #define COLL_ELEM_LEN_MAX 8 @@ -110,11 +148,15 @@ /* Rename to standard API for using out of glibc. */ #ifndef _LIBC +# undef __wctype +# undef __iswctype # define __wctype wctype +# define __iswalnum iswalnum # define __iswctype iswctype +# define __towlower towlower +# define __towupper towupper # define __btowc btowc # define __mbrtowc mbrtowc -# define __mempcpy mempcpy # define __wcrtomb wcrtomb # define __regfree regfree # define attribute_hidden @@ -124,32 +166,70 @@ # define __attribute__(arg) #endif -extern const char __re_error_msgid[] attribute_hidden; -extern const size_t __re_error_msgid_idx[] attribute_hidden; +#ifndef SSIZE_MAX +# define SSIZE_MAX ((ssize_t) (SIZE_MAX / 2)) +#endif + +/* The type of indexes into strings. This is signed, not size_t, + since the API requires indexes to fit in regoff_t anyway, and using + signed integers makes the code a bit smaller and presumably faster. + The traditional GNU regex implementation uses int for indexes. + The POSIX-compatible implementation uses a possibly-wider type. + The name 'Idx' is three letters to minimize the hassle of + reindenting a lot of regex code that formerly used 'int'. */ +typedef regoff_t Idx; +#ifdef _REGEX_LARGE_OFFSETS +# define IDX_MAX SSIZE_MAX +#else +# define IDX_MAX INT_MAX +#endif + +/* A hash value, suitable for computing hash tables. */ +typedef __re_size_t re_hashval_t; /* An integer used to represent a set of bits. It must be unsigned, and must be at least as wide as unsigned int. */ typedef unsigned long int bitset_word_t; /* All bits set in a bitset_word_t. */ #define BITSET_WORD_MAX ULONG_MAX -/* Number of bits in a bitset_word_t. */ -#define BITSET_WORD_BITS (sizeof (bitset_word_t) * CHAR_BIT) -/* Number of bitset_word_t in a bit_set. */ -#define BITSET_WORDS (SBC_MAX / BITSET_WORD_BITS) + +/* Number of bits in a bitset_word_t. For portability to hosts with + padding bits, do not use '(sizeof (bitset_word_t) * CHAR_BIT)'; + instead, deduce it directly from BITSET_WORD_MAX. Avoid + greater-than-32-bit integers and unconditional shifts by more than + 31 bits, as they're not portable. */ +#if BITSET_WORD_MAX == 0xffffffffUL +# define BITSET_WORD_BITS 32 +#elif BITSET_WORD_MAX >> 31 >> 4 == 1 +# define BITSET_WORD_BITS 36 +#elif BITSET_WORD_MAX >> 31 >> 16 == 1 +# define BITSET_WORD_BITS 48 +#elif BITSET_WORD_MAX >> 31 >> 28 == 1 +# define BITSET_WORD_BITS 60 +#elif BITSET_WORD_MAX >> 31 >> 31 >> 1 == 1 +# define BITSET_WORD_BITS 64 +#elif BITSET_WORD_MAX >> 31 >> 31 >> 9 == 1 +# define BITSET_WORD_BITS 72 +#elif BITSET_WORD_MAX >> 31 >> 31 >> 31 >> 31 >> 3 == 1 +# define BITSET_WORD_BITS 128 +#elif BITSET_WORD_MAX >> 31 >> 31 >> 31 >> 31 >> 31 >> 31 >> 31 >> 31 >> 7 == 1 +# define BITSET_WORD_BITS 256 +#elif BITSET_WORD_MAX >> 31 >> 31 >> 31 >> 31 >> 31 >> 31 >> 31 >> 31 >> 7 > 1 +# define BITSET_WORD_BITS 257 /* any value > SBC_MAX will do here */ +# if BITSET_WORD_BITS <= SBC_MAX +# error "Invalid SBC_MAX" +# endif +#else +# error "Add case for new bitset_word_t size" +#endif + +/* Number of bitset_word_t values in a bitset_t. */ +#define BITSET_WORDS ((SBC_MAX + BITSET_WORD_BITS - 1) / BITSET_WORD_BITS) + typedef bitset_word_t bitset_t[BITSET_WORDS]; typedef bitset_word_t *re_bitset_ptr_t; typedef const bitset_word_t *re_const_bitset_ptr_t; -#define bitset_set(set,i) \ - (set[i / BITSET_WORD_BITS] |= (bitset_word_t) 1 << i % BITSET_WORD_BITS) -#define bitset_clear(set,i) \ - (set[i / BITSET_WORD_BITS] &= ~((bitset_word_t) 1 << i % BITSET_WORD_BITS)) -#define bitset_contain(set,i) \ - (set[i / BITSET_WORD_BITS] & ((bitset_word_t) 1 << i % BITSET_WORD_BITS)) -#define bitset_empty(set) memset (set, '\0', sizeof (bitset_t)) -#define bitset_set_all(set) memset (set, '\xff', sizeof (bitset_t)) -#define bitset_copy(dest,src) memcpy (dest, src, sizeof (bitset_t)) - #define PREV_WORD_CONSTRAINT 0x0001 #define PREV_NOTWORD_CONSTRAINT 0x0002 #define NEXT_WORD_CONSTRAINT 0x0004 @@ -177,9 +257,9 @@ typedef enum typedef struct { - int alloc; - int nelem; - int *elems; + Idx alloc; + Idx nelem; + Idx *elems; } re_node_set; typedef enum @@ -265,19 +345,19 @@ typedef struct unsigned int non_match : 1; /* # of multibyte characters. */ - int nmbchars; + Idx nmbchars; /* # of collating symbols. */ - int ncoll_syms; + Idx ncoll_syms; /* # of equivalence classes. */ - int nequiv_classes; + Idx nequiv_classes; /* # of range expressions. */ - int nranges; + Idx nranges; /* # of character classes. */ - int nchar_classes; + Idx nchar_classes; } re_charset_t; #endif /* RE_ENABLE_I18N */ @@ -290,10 +370,10 @@ typedef struct #ifdef RE_ENABLE_I18N re_charset_t *mbcset; /* for COMPLEX_BRACKET */ #endif /* RE_ENABLE_I18N */ - int idx; /* for BACK_REF */ + Idx idx; /* for BACK_REF */ re_context_type ctx_type; /* for ANCHOR */ } opr; -#if __GNUC__ >= 2 +#if __GNUC__ >= 2 && !defined __STRICT_ANSI__ re_token_type_t type : 8; #else re_token_type_t type; @@ -324,30 +404,30 @@ struct re_string_t #ifdef RE_ENABLE_I18N /* Store the wide character string which is corresponding to MBS. */ wint_t *wcs; - int *offsets; + Idx *offsets; mbstate_t cur_state; #endif /* Index in RAW_MBS. Each character mbs[i] corresponds to raw_mbs[raw_mbs_idx + i]. */ - int raw_mbs_idx; + Idx raw_mbs_idx; /* The length of the valid characters in the buffers. */ - int valid_len; + Idx valid_len; /* The corresponding number of bytes in raw_mbs array. */ - int valid_raw_len; + Idx valid_raw_len; /* The length of the buffers MBS and WCS. */ - int bufs_len; + Idx bufs_len; /* The index in MBS, which is updated by re_string_fetch_byte. */ - int cur_idx; + Idx cur_idx; /* length of RAW_MBS array. */ - int raw_len; + Idx raw_len; /* This is RAW_LEN - RAW_MBS_IDX + VALID_LEN - VALID_RAW_LEN. */ - int len; + Idx len; /* End of the buffer may be shorter than its length in the cases such as re_match_2, re_search_2. Then, we use STOP for end of the buffer instead of LEN. */ - int raw_stop; + Idx raw_stop; /* This is RAW_STOP - RAW_MBS_IDX adjusted through OFFSETS. */ - int stop; + Idx stop; /* The context of mbs[0]. We store the context independently, since the context of mbs[0] may be different from raw_mbs[0], which is @@ -357,7 +437,7 @@ struct re_string_t RE_TRANSLATE_TYPE trans; /* Copy of re_dfa_t's word_char. */ re_const_bitset_ptr_t word_char; - /* 1 if REG_ICASE. */ + /* true if REG_ICASE. */ unsigned char icase; unsigned char is_utf8; unsigned char map_notascii; @@ -373,18 +453,10 @@ typedef struct re_string_t re_string_t; struct re_dfa_t; typedef struct re_dfa_t re_dfa_t; -#if IS_IN (libc) -static reg_errcode_t re_string_realloc_buffers (re_string_t *pstr, - int new_buf_len); -# ifdef RE_ENABLE_I18N -static void build_wcs_buffer (re_string_t *pstr); -static reg_errcode_t build_wcs_upper_buffer (re_string_t *pstr); -# endif /* RE_ENABLE_I18N */ -static void build_upper_buffer (re_string_t *pstr); -static void re_string_translate_buffer (re_string_t *pstr); -static unsigned int re_string_context_at (const re_string_t *input, int idx, - int eflags) __attribute__ ((pure)); +#ifndef _LIBC +# define IS_IN(libc) false #endif + #define re_string_peek_byte(pstr, offset) \ ((pstr)->mbs[(pstr)->cur_idx + offset]) #define re_string_fetch_byte(pstr) \ @@ -402,7 +474,9 @@ static unsigned int re_string_context_at (const re_string_t *input, int idx, #define re_string_skip_bytes(pstr,idx) ((pstr)->cur_idx += (idx)) #define re_string_set_index(pstr,idx) ((pstr)->cur_idx = (idx)) -#include <alloca.h> +#if defined _LIBC || HAVE_ALLOCA +# include <alloca.h> +#endif #ifndef _LIBC # if HAVE_ALLOCA @@ -414,9 +488,24 @@ static unsigned int re_string_context_at (const re_string_t *input, int idx, # else /* alloca is implemented with malloc, so just use malloc. */ # define __libc_use_alloca(n) 0 +# undef alloca +# define alloca(n) malloc (n) # endif #endif +#ifdef _LIBC +# define MALLOC_0_IS_NONNULL 1 +#elif !defined MALLOC_0_IS_NONNULL +# define MALLOC_0_IS_NONNULL 0 +#endif + +#ifndef MAX +# define MAX(a,b) ((a) < (b) ? (b) : (a)) +#endif +#ifndef MIN +# define MIN(a,b) ((a) < (b) ? (a) : (b)) +#endif + #define re_malloc(t,n) ((t *) malloc ((n) * sizeof (t))) #define re_realloc(p,t,n) ((t *) realloc (p, (n) * sizeof (t))) #define re_free(p) free (p) @@ -431,9 +520,9 @@ struct bin_tree_t re_token_t token; - /* `node_idx' is the index in dfa->nodes, if `type' == 0. - Otherwise `type' indicate the type of this node. */ - int node_idx; + /* 'node_idx' is the index in dfa->nodes, if 'type' == 0. + Otherwise 'type' indicate the type of this node. */ + Idx node_idx; }; typedef struct bin_tree_t bin_tree_t; @@ -477,7 +566,7 @@ typedef struct bin_tree_storage_t bin_tree_storage_t; struct re_dfastate_t { - unsigned int hash; + re_hashval_t hash; re_node_set nodes; re_node_set non_eps_nodes; re_node_set inveclosure; @@ -485,9 +574,9 @@ struct re_dfastate_t struct re_dfastate_t **trtable, **word_trtable; unsigned int context : 4; unsigned int halt : 1; - /* If this state can accept `multi byte'. + /* If this state can accept "multi byte". Note that we refer to multibyte characters, and multi character - collating elements as `multi byte'. */ + collating elements as "multi byte". */ unsigned int accept_mb : 1; /* If this state has backreference node(s). */ unsigned int has_backref : 1; @@ -497,8 +586,8 @@ typedef struct re_dfastate_t re_dfastate_t; struct re_state_table_entry { - int num; - int alloc; + Idx num; + Idx alloc; re_dfastate_t **array; }; @@ -506,8 +595,8 @@ struct re_state_table_entry typedef struct { - int next_idx; - int alloc; + Idx next_idx; + Idx alloc; re_dfastate_t **array; } state_array_t; @@ -515,8 +604,8 @@ typedef struct typedef struct { - int node; - int str_idx; /* The position NODE match at. */ + Idx node; + Idx str_idx; /* The position NODE match at. */ state_array_t path; } re_sub_match_last_t; @@ -526,20 +615,20 @@ typedef struct typedef struct { - int str_idx; - int node; + Idx str_idx; + Idx node; state_array_t *path; - int alasts; /* Allocation size of LASTS. */ - int nlasts; /* The number of LASTS. */ + Idx alasts; /* Allocation size of LASTS. */ + Idx nlasts; /* The number of LASTS. */ re_sub_match_last_t **lasts; } re_sub_match_top_t; struct re_backref_cache_entry { - int node; - int str_idx; - int subexp_from; - int subexp_to; + Idx node; + Idx str_idx; + Idx subexp_from; + Idx subexp_to; char more; char unused; unsigned short int eps_reachable_subexps_map; @@ -557,18 +646,18 @@ typedef struct /* EFLAGS of the argument of regexec. */ int eflags; /* Where the matching ends. */ - int match_last; - int last_node; + Idx match_last; + Idx last_node; /* The state log used by the matcher. */ re_dfastate_t **state_log; - int state_log_top; + Idx state_log_top; /* Back reference cache. */ - int nbkref_ents; - int abkref_ents; + Idx nbkref_ents; + Idx abkref_ents; struct re_backref_cache_entry *bkref_ents; int max_mb_elem_len; - int nsub_tops; - int asub_tops; + Idx nsub_tops; + Idx asub_tops; re_sub_match_top_t **sub_tops; } re_match_context_t; @@ -576,23 +665,23 @@ typedef struct { re_dfastate_t **sifted_states; re_dfastate_t **limited_states; - int last_node; - int last_str_idx; + Idx last_node; + Idx last_str_idx; re_node_set limits; } re_sift_context_t; struct re_fail_stack_ent_t { - int idx; - int node; + Idx idx; + Idx node; regmatch_t *regs; re_node_set eps_via_nodes; }; struct re_fail_stack_t { - int num; - int alloc; + Idx num; + Idx alloc; struct re_fail_stack_ent_t *stack; }; @@ -601,8 +690,8 @@ struct re_dfa_t re_token_t *nodes; size_t nodes_alloc; size_t nodes_len; - int *nexts; - int *org_indices; + Idx *nexts; + Idx *org_indices; re_node_set *edests; re_node_set *eclosures; re_node_set *inveclosures; @@ -616,10 +705,10 @@ struct re_dfa_t re_bitset_ptr_t sb_char; int str_tree_storage_idx; - /* number of subexpressions `re_nsub' is in regex_t. */ - unsigned int state_hash_mask; - int init_node; - int nbackref; /* The number of backreference in this dfa. */ + /* number of subexpressions 're_nsub' is in regex_t. */ + re_hashval_t state_hash_mask; + Idx init_node; + Idx nbackref; /* The number of backreference in this dfa. */ /* Bitmap expressing which backreference is used. */ bitset_word_t used_bkref_map; @@ -636,11 +725,11 @@ struct re_dfa_t int mb_cur_max; bitset_t word_char; reg_syntax_t syntax; - int *subexp_map; + Idx *subexp_map; #ifdef DEBUG char* re_str; #endif - __libc_lock_define (, lock) + lock_define (lock) }; #define re_node_set_init_empty(set) memset (set, '\0', sizeof (re_node_set)) @@ -671,16 +760,60 @@ typedef struct } bracket_elem_t; -/* Inline functions for bitset operation. */ -static void __attribute__ ((unused)) +/* Functions for bitset_t operation. */ + +static inline void +bitset_set (bitset_t set, Idx i) +{ + set[i / BITSET_WORD_BITS] |= (bitset_word_t) 1 << i % BITSET_WORD_BITS; +} + +static inline void +bitset_clear (bitset_t set, Idx i) +{ + set[i / BITSET_WORD_BITS] &= ~ ((bitset_word_t) 1 << i % BITSET_WORD_BITS); +} + +static inline bool +bitset_contain (const bitset_t set, Idx i) +{ + return (set[i / BITSET_WORD_BITS] >> i % BITSET_WORD_BITS) & 1; +} + +static inline void +bitset_empty (bitset_t set) +{ + memset (set, '\0', sizeof (bitset_t)); +} + +static inline void +bitset_set_all (bitset_t set) +{ + memset (set, -1, sizeof (bitset_word_t) * (SBC_MAX / BITSET_WORD_BITS)); + if (SBC_MAX % BITSET_WORD_BITS != 0) + set[BITSET_WORDS - 1] = + ((bitset_word_t) 1 << SBC_MAX % BITSET_WORD_BITS) - 1; +} + +static inline void +bitset_copy (bitset_t dest, const bitset_t src) +{ + memcpy (dest, src, sizeof (bitset_t)); +} + +static inline void bitset_not (bitset_t set) { int bitset_i; - for (bitset_i = 0; bitset_i < BITSET_WORDS; ++bitset_i) + for (bitset_i = 0; bitset_i < SBC_MAX / BITSET_WORD_BITS; ++bitset_i) set[bitset_i] = ~set[bitset_i]; + if (SBC_MAX % BITSET_WORD_BITS != 0) + set[BITSET_WORDS - 1] = + ((((bitset_word_t) 1 << SBC_MAX % BITSET_WORD_BITS) - 1) + & ~set[BITSET_WORDS - 1]); } -static void __attribute__ ((unused)) +static inline void bitset_merge (bitset_t dest, const bitset_t src) { int bitset_i; @@ -688,7 +821,7 @@ bitset_merge (bitset_t dest, const bitset_t src) dest[bitset_i] |= src[bitset_i]; } -static void __attribute__ ((unused)) +static inline void bitset_mask (bitset_t dest, const bitset_t src) { int bitset_i; @@ -697,10 +830,10 @@ bitset_mask (bitset_t dest, const bitset_t src) } #ifdef RE_ENABLE_I18N -/* Inline functions for re_string. */ +/* Functions for re_string. */ static int __attribute__ ((pure, unused)) -re_string_char_size_at (const re_string_t *pstr, int idx) +re_string_char_size_at (const re_string_t *pstr, Idx idx) { int byte_idx; if (pstr->mb_cur_max == 1) @@ -713,23 +846,22 @@ re_string_char_size_at (const re_string_t *pstr, int idx) static wint_t __attribute__ ((pure, unused)) -re_string_wchar_at (const re_string_t *pstr, int idx) +re_string_wchar_at (const re_string_t *pstr, Idx idx) { if (pstr->mb_cur_max == 1) return (wint_t) pstr->mbs[idx]; return (wint_t) pstr->wcs[idx]; } -# if IS_IN (libc) -# ifdef _LIBC -# include <locale/weight.h> -# endif +# ifdef _LIBC +# include <locale/weight.h> +# endif static int __attribute__ ((pure, unused)) -re_string_elem_size_at (const re_string_t *pstr, int idx) +re_string_elem_size_at (const re_string_t *pstr, Idx idx) { -# ifdef _LIBC +# ifdef _LIBC const unsigned char *p, *extra; const int32_t *table, *indirect; uint_fast32_t nrules = _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES); @@ -746,10 +878,34 @@ re_string_elem_size_at (const re_string_t *pstr, int idx) return p - pstr->mbs - idx; } else -# endif /* _LIBC */ +# endif /* _LIBC */ return 1; } -# endif #endif /* RE_ENABLE_I18N */ +#ifndef __GNUC_PREREQ +# if defined __GNUC__ && defined __GNUC_MINOR__ +# define __GNUC_PREREQ(maj, min) \ + ((__GNUC__ << 16) + __GNUC_MINOR__ >= ((maj) << 16) + (min)) +# else +# define __GNUC_PREREQ(maj, min) 0 +# endif +#endif + +#if __GNUC_PREREQ (3,4) +# undef __attribute_warn_unused_result__ +# define __attribute_warn_unused_result__ \ + __attribute__ ((__warn_unused_result__)) +#else +# define __attribute_warn_unused_result__ /* empty */ +#endif + +#ifndef FALLTHROUGH +# if __GNUC__ < 7 +# define FALLTHROUGH ((void) 0) +# else +# define FALLTHROUGH __attribute__ ((__fallthrough__)) +# endif +#endif + #endif /* _REGEX_INTERNAL_H */ |