#include <assert.h>
#include <ctype.h>
+#include <stdbool.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
+#ifndef _LIBC
+# include "strcase.h"
+#endif
+
#if defined HAVE_LANGINFO_H || defined HAVE_LANGINFO_CODESET || defined _LIBC
# include <langinfo.h>
#endif
# define RE_ENABLE_I18N
#endif
+#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, 1)
+# define always_inline
+#endif
+
#if __GNUC__ >= 3
# define BE(expr, val) __builtin_expect (expr, val)
#else
# define BE(expr, val) (expr)
# define inline
+# define pure
#endif
-/* Number of bits in a byte. */
-#define BYTE_BITS 8
/* Number of single byte character. */
#define SBC_MAX 256
# define attribute_hidden
#endif /* not _LIBC */
-#ifdef __GNUC__
+#if __GNUC__ >= 3
# define __attribute(arg) __attribute__ (arg)
#else
# define __attribute(arg)
extern const char __re_error_msgid[] attribute_hidden;
extern const size_t __re_error_msgid_idx[] attribute_hidden;
-/* Number of bits in an unsinged int. */
-#define UINT_BITS (sizeof (unsigned int) * BYTE_BITS)
-/* Number of unsigned int in an bit_set. */
-#define BITSET_UINTS ((SBC_MAX + UINT_BITS - 1) / UINT_BITS)
-typedef unsigned int bitset[BITSET_UINTS];
-typedef unsigned int *re_bitset_ptr_t;
-typedef const unsigned int *re_const_bitset_ptr_t;
-
-#define bitset_set(set,i) (set[i / UINT_BITS] |= 1 << i % UINT_BITS)
-#define bitset_clear(set,i) (set[i / UINT_BITS] &= ~(1 << i % UINT_BITS))
-#define bitset_contain(set,i) (set[i / UINT_BITS] & (1 << i % UINT_BITS))
-#define bitset_empty(set) memset (set, 0, sizeof (unsigned int) * BITSET_UINTS)
-#define bitset_set_all(set) \
- memset (set, 255, sizeof (unsigned int) * BITSET_UINTS)
-#define bitset_copy(dest,src) \
- memcpy (dest, src, sizeof (unsigned int) * BITSET_UINTS)
-static inline void bitset_not (bitset set);
-static inline void bitset_merge (bitset dest, const bitset src);
-static inline void bitset_not_merge (bitset dest, const bitset src);
-static inline void bitset_mask (bitset dest, const bitset src);
+typedef __re_idx_t Idx;
+
+/* Special return value for failure to match. */
+#define REG_MISSING ((Idx) -1)
+
+/* Special return value for internal error. */
+#define REG_ERROR ((Idx) -2)
+
+/* Test whether N is a valid index, and is not one of the above. */
+#ifdef _REGEX_LARGE_OFFSETS
+# define REG_VALID_INDEX(n) ((Idx) (n) < REG_ERROR)
+#else
+# define REG_VALID_INDEX(n) (0 <= (n))
+#endif
+
+/* Test whether N is a valid nonzero index. */
+#ifdef _REGEX_LARGE_OFFSETS
+# define REG_VALID_NONZERO_INDEX(n) ((Idx) ((n) - 1) < (Idx) (REG_ERROR - 1))
+#else
+# define REG_VALID_NONZERO_INDEX(n) (0 < (n))
+#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;
+
+/* Maximum value of a bitset word. It must be useful in preprocessor
+ contexts, and must be consistent with bitset_word. */
+#define BITSET_WORD_MAX ULONG_MAX
+
+/* Number of bits in a bitset word. Avoid greater-than-32-bit
+ integers and unconditional shifts by more than 31 bits, as they're
+ not portable. */
+#if BITSET_WORD_MAX == 0xffffffff
+# define BITSET_WORD_BITS 32
+#elif BITSET_WORD_MAX >> 31 >> 5 == 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 size"
+#endif
+
+/* Number of bitset words in a bitset. */
+#define BITSET_WORDS ((SBC_MAX + BITSET_WORD_BITS - 1) / BITSET_WORD_BITS)
+
+typedef bitset_word bitset[BITSET_WORDS];
+typedef bitset_word *re_bitset_ptr_t;
+typedef const bitset_word *re_const_bitset_ptr_t;
#define PREV_WORD_CONSTRAINT 0x0001
#define PREV_NOTWORD_CONSTRAINT 0x0002
typedef struct
{
- int alloc;
- int nelem;
- int *elems;
+ Idx alloc;
+ Idx nelem;
+ Idx *elems;
} re_node_set;
typedef enum
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 */
#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
#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
unsigned REG_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;
# endif
#endif
-static reg_errcode_t re_string_allocate (re_string_t *pstr, const char *str,
- int len, int init_len,
- REG_TRANSLATE_TYPE trans, int icase,
- const re_dfa_t *dfa)
- internal_function;
-static reg_errcode_t re_string_construct (re_string_t *pstr, const char *str,
- int len, REG_TRANSLATE_TYPE trans,
- int icase, const re_dfa_t *dfa)
- internal_function;
-static reg_errcode_t re_string_reconstruct (re_string_t *pstr, int idx,
- int eflags) internal_function;
static reg_errcode_t re_string_realloc_buffers (re_string_t *pstr,
- int new_buf_len)
+ Idx new_buf_len)
internal_function;
#ifdef RE_ENABLE_I18N
static void build_wcs_buffer (re_string_t *pstr) internal_function;
#endif /* RE_ENABLE_I18N */
static void build_upper_buffer (re_string_t *pstr) internal_function;
static void re_string_translate_buffer (re_string_t *pstr) internal_function;
-static void re_string_destruct (re_string_t *pstr) internal_function;
-#ifdef RE_ENABLE_I18N
-static int re_string_elem_size_at (const re_string_t *pstr, int idx)
- internal_function __attribute ((pure));
-static inline int re_string_char_size_at (const re_string_t *pstr, int idx)
- internal_function __attribute ((pure));
-static inline wint_t re_string_wchar_at (const re_string_t *pstr, int idx)
- internal_function __attribute ((pure));
-#endif /* RE_ENABLE_I18N */
-static unsigned int re_string_context_at (const re_string_t *input, int idx,
- int eflags)
- internal_function __attribute ((pure));
-static unsigned char re_string_peek_byte_case (const re_string_t *pstr,
- int idx)
- internal_function __attribute ((pure));
-static unsigned char re_string_fetch_byte_case (re_string_t *pstr)
+static unsigned int re_string_context_at (const re_string_t *input,
+ Idx idx, int eflags)
internal_function __attribute ((pure));
#define re_string_peek_byte(pstr, offset) \
#include <alloca.h>
-#ifndef __LIBC
+#ifndef _LIBC
# if HAVE_ALLOCA
/* The OS usually guarantees only one guard page at the bottom of the stack,
and a page size can be as small as 4096 bytes. So we cannot safely
#endif
#define re_malloc(t,n) ((t *) malloc ((n) * sizeof (t)))
+#define re_xmalloc(t,n) ((t *) re_xnmalloc (n, sizeof (t)))
#define re_calloc(t,n) ((t *) calloc (n, sizeof (t)))
#define re_realloc(p,t,n) ((t *) realloc (p, (n) * sizeof (t)))
+#define re_xrealloc(p,t,n) ((t *) re_xnrealloc (p, n, sizeof (t)))
+#define re_x2realloc(p,t,pn) ((t *) re_x2nrealloc (p, pn, sizeof (t)))
#define re_free(p) free (p)
+#ifndef SIZE_MAX
+# define SIZE_MAX ((size_t) -1)
+#endif
+
+/* Return true if an array of N objects, each of size S, cannot exist
+ due to size arithmetic overflow. S must be nonzero. */
+static inline bool
+re_alloc_oversized (size_t n, size_t s)
+{
+ return BE (SIZE_MAX / s < n, 0);
+}
+
+/* Return true if an array of (2 * N + 1) objects, each of size S,
+ cannot exist due to size arithmetic overflow. S must be nonzero. */
+static inline bool
+re_x2alloc_oversized (size_t n, size_t s)
+{
+ return BE ((SIZE_MAX / s - 1) / 2 < n, 0);
+}
+
+/* Allocate an array of N objects, each with S bytes of memory,
+ dynamically, with error checking. S must be nonzero. */
+static inline void *
+re_xnmalloc (size_t n, size_t s)
+{
+ return re_alloc_oversized (n, s) ? NULL : malloc (n * s);
+}
+
+/* Change the size of an allocated block of memory P to an array of N
+ objects each of S bytes, with error checking. S must be nonzero. */
+static inline void *
+re_xnrealloc (void *p, size_t n, size_t s)
+{
+ return re_alloc_oversized (n, s) ? NULL : realloc (p, n * s);
+}
+
+/* Reallocate a block of memory P to an array of (2 * (*PN) + 1)
+ objects each of S bytes, with error checking. S must be nonzero.
+ If the allocation is successful, set *PN to the new allocation
+ count and return the resulting pointer. Otherwise, return
+ NULL. */
+static inline void *
+re_x2nrealloc (void *p, size_t *pn, size_t s)
+{
+ if (re_x2alloc_oversized (*pn, s))
+ return NULL;
+ else
+ {
+ /* Add 1 in case *PN is zero. */
+ size_t n1 = 2 * *pn + 1;
+ p = realloc (p, n1 * s);
+ if (BE (p != NULL, 1))
+ *pn = n1;
+ return p;
+ }
+}
+
struct bin_tree_t
{
struct bin_tree_t *parent;
/* `node_idx' is the index in dfa->nodes, if `type' == 0.
Otherwise `type' indicate the type of this node. */
- int node_idx;
+ Idx node_idx;
};
typedef struct bin_tree_t bin_tree_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;
struct re_state_table_entry
{
- int num;
- int alloc;
+ Idx num;
+ Idx alloc;
re_dfastate_t **array;
};
typedef struct
{
- int next_idx;
- int alloc;
+ Idx next_idx;
+ Idx alloc;
re_dfastate_t **array;
} state_array_t;
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;
typedef struct
{
- int str_idx;
- int node;
- int next_last_offset;
+ 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;
/* 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;
{
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;
};
struct re_dfa_t
{
re_token_t *nodes;
- int nodes_alloc;
- int nodes_len;
- int *nexts;
- int *org_indices;
+ Idx nodes_alloc;
+ Idx nodes_len;
+ Idx *nexts;
+ Idx *org_indices;
re_node_set *edests;
re_node_set *eclosures;
re_node_set *inveclosures;
int str_tree_storage_idx;
/* number of subexpressions `re_nsub' is in regex_t. */
- unsigned int state_hash_mask;
- int states_alloc;
- int init_node;
- int nbackref; /* The number of backreference in this dfa. */
+ re_hashval_t state_hash_mask;
+ Idx init_node;
+ Idx nbackref; /* The number of backreference in this dfa. */
/* Bitmap expressing which backreference is used. */
- unsigned int used_bkref_map;
- unsigned int completed_bkref_map;
+ bitset_word used_bkref_map;
+ bitset_word completed_bkref_map;
unsigned int has_plural_match : 1;
/* If this dfa has "multibyte node", which is a backreference or
int mb_cur_max;
bitset word_char;
reg_syntax_t syntax;
- int *subexp_map;
+ Idx *subexp_map;
#ifdef DEBUG
char* re_str;
#endif
__libc_lock_define (, lock)
};
-#ifndef RE_NO_INTERNAL_PROTOTYPES
-static reg_errcode_t re_node_set_alloc (re_node_set *set, int size) internal_function;
-static reg_errcode_t re_node_set_init_1 (re_node_set *set, int elem) internal_function;
-static reg_errcode_t re_node_set_init_2 (re_node_set *set, int elem1,
- int elem2) internal_function;
#define re_node_set_init_empty(set) memset (set, '\0', sizeof (re_node_set))
-static reg_errcode_t re_node_set_init_copy (re_node_set *dest,
- const re_node_set *src) internal_function;
-static reg_errcode_t re_node_set_add_intersect (re_node_set *dest,
- const re_node_set *src1,
- const re_node_set *src2) internal_function;
-static reg_errcode_t re_node_set_init_union (re_node_set *dest,
- const re_node_set *src1,
- const re_node_set *src2) internal_function;
-static reg_errcode_t re_node_set_merge (re_node_set *dest,
- const re_node_set *src) internal_function;
-static int re_node_set_insert (re_node_set *set, int elem) internal_function;
-static int re_node_set_insert_last (re_node_set *set,
- int elem) internal_function;
-static int re_node_set_compare (const re_node_set *set1,
- const re_node_set *set2)
- internal_function __attribute ((pure));
-static int re_node_set_contains (const re_node_set *set, int elem)
- internal_function __attribute ((pure));
-static void re_node_set_remove_at (re_node_set *set, int idx) internal_function;
#define re_node_set_remove(set,id) \
(re_node_set_remove_at (set, re_node_set_contains (set, id) - 1))
#define re_node_set_empty(p) ((p)->nelem = 0)
#define re_node_set_free(set) re_free ((set)->elems)
-static int re_dfa_add_node (re_dfa_t *dfa, re_token_t token) internal_function;
-static re_dfastate_t *re_acquire_state (reg_errcode_t *err, re_dfa_t *dfa,
- const re_node_set *nodes) internal_function;
-static re_dfastate_t *re_acquire_state_context (reg_errcode_t *err,
- re_dfa_t *dfa,
- const re_node_set *nodes,
- unsigned int context) internal_function;
+
static void free_state (re_dfastate_t *state) internal_function;
-#endif
\f
typedef enum
/* Inline functions for bitset operation. */
+
static inline void
-bitset_not (bitset set)
+bitset_set (bitset set, Idx i)
{
- int bitset_i;
- for (bitset_i = 0; bitset_i < BITSET_UINTS; ++bitset_i)
- set[bitset_i] = ~set[bitset_i];
+ set[i / BITSET_WORD_BITS] |= (bitset_word) 1 << i % BITSET_WORD_BITS;
}
static inline void
-bitset_merge (bitset dest, const bitset src)
+bitset_clear (bitset set, Idx i)
+{
+ set[i / BITSET_WORD_BITS] &= ~ ((bitset_word) 1 << i % BITSET_WORD_BITS);
+}
+
+static inline bool
+bitset_contain (const bitset set, Idx i)
+{
+ return (set[i / BITSET_WORD_BITS] >> i % BITSET_WORD_BITS) & 1;
+}
+
+static inline void
+bitset_empty (bitset set)
{
- int bitset_i;
- for (bitset_i = 0; bitset_i < BITSET_UINTS; ++bitset_i)
- dest[bitset_i] |= src[bitset_i];
+ memset (set, 0, sizeof (bitset));
}
static inline void
-bitset_not_merge (bitset dest, const bitset src)
+bitset_set_all (bitset set)
+{
+ memset (set, -1, sizeof (bitset_word) * (SBC_MAX / BITSET_WORD_BITS));
+ if (SBC_MAX % BITSET_WORD_BITS != 0)
+ set[BITSET_WORDS - 1] =
+ ((bitset_word) 1 << SBC_MAX % BITSET_WORD_BITS) - 1;
+}
+
+static inline void
+bitset_copy (bitset dest, const bitset src)
+{
+ memcpy (dest, src, sizeof (bitset));
+}
+
+static inline void
+bitset_not (bitset set)
+{
+ int i;
+ for (i = 0; i < SBC_MAX / BITSET_WORD_BITS; ++i)
+ set[i] = ~set[i];
+ if (SBC_MAX % BITSET_WORD_BITS != 0)
+ set[BITSET_WORDS - 1] =
+ ((((bitset_word) 1 << SBC_MAX % BITSET_WORD_BITS) - 1)
+ & ~set[BITSET_WORDS - 1]);
+}
+
+static inline void
+bitset_merge (bitset dest, const bitset src)
{
int i;
- for (i = 0; i < BITSET_UINTS; ++i)
- dest[i] |= ~src[i];
+ for (i = 0; i < BITSET_WORDS; ++i)
+ dest[i] |= src[i];
}
static inline void
bitset_mask (bitset dest, const bitset src)
{
- int bitset_i;
- for (bitset_i = 0; bitset_i < BITSET_UINTS; ++bitset_i)
- dest[bitset_i] &= src[bitset_i];
+ int i;
+ for (i = 0; i < BITSET_WORDS; ++i)
+ dest[i] &= src[i];
}
-#if defined RE_ENABLE_I18N && !defined RE_NO_INTERNAL_PROTOTYPES
+#if defined RE_ENABLE_I18N
/* Inline functions for re_string. */
static inline int
-internal_function
-re_string_char_size_at (const re_string_t *pstr, int idx)
+internal_function __attribute ((pure))
+re_string_char_size_at (const re_string_t *pstr, Idx idx)
{
int byte_idx;
if (pstr->mb_cur_max == 1)
}
static inline wint_t
-internal_function
-re_string_wchar_at (const re_string_t *pstr, int idx)
+internal_function __attribute ((pure))
+re_string_wchar_at (const re_string_t *pstr, Idx idx)
{
if (pstr->mb_cur_max == 1)
return (wint_t) pstr->mbs[idx];
}
static int
-internal_function
-re_string_elem_size_at (const re_string_t *pstr, int idx)
+internal_function __attribute ((pure))
+re_string_elem_size_at (const re_string_t *pstr, Idx idx)
{
#ifdef _LIBC
const unsigned char *p, *extra;
projects / gnulib.git / blobdiff