static reg_errcode_t match_ctx_add_entry (re_match_context_t *cache, Idx node,
Idx str_idx, Idx from, Idx to)
internal_function;
-static Idx search_cur_bkref_entry (re_match_context_t *mctx, Idx str_idx)
+static Idx search_cur_bkref_entry (const re_match_context_t *mctx, Idx str_idx)
internal_function;
static reg_errcode_t match_ctx_add_subtop (re_match_context_t *mctx, Idx node,
Idx str_idx) internal_function;
internal_function;
static reg_errcode_t re_search_internal (const regex_t *preg,
const char *string, Idx length,
- Idx start, regoff_t range, Idx stop,
+ Idx start, Idx last_start, Idx stop,
size_t nmatch, regmatch_t pmatch[],
int eflags) internal_function;
static regoff_t re_search_2_stub (struct re_pattern_buffer *bufp,
const char *string2, Idx length2,
Idx start, regoff_t range,
struct re_registers *regs,
- Idx stop, int ret_len) internal_function;
+ Idx stop, bool ret_len) internal_function;
static regoff_t re_search_stub (struct re_pattern_buffer *bufp,
const char *string, Idx length, Idx start,
regoff_t range, Idx stop,
struct re_registers *regs,
- int ret_len) internal_function;
+ bool ret_len) internal_function;
static unsigned re_copy_regs (struct re_registers *regs, regmatch_t *pmatch,
Idx nregs, int regs_allocated) internal_function;
static reg_errcode_t prune_impossible_nodes (re_match_context_t *mctx)
internal_function;
-static Idx check_matching (re_match_context_t *mctx, int fl_longest_match,
+static Idx check_matching (re_match_context_t *mctx, bool fl_longest_match,
Idx *p_match_first)
internal_function;
static Idx check_halt_state_context (const re_match_context_t *mctx,
static reg_errcode_t set_regs (const regex_t *preg,
const re_match_context_t *mctx,
size_t nmatch, regmatch_t *pmatch,
- int fl_backtrack) internal_function;
+ bool fl_backtrack) internal_function;
static reg_errcode_t free_fail_stack_return (struct re_fail_stack_t *fs) internal_function;
#ifdef RE_ENABLE_I18N
static reg_errcode_t add_epsilon_src_nodes (re_dfa_t *dfa,
re_node_set *dest_nodes,
const re_node_set *candidates) internal_function;
-static int check_dst_limits (re_match_context_t *mctx, re_node_set *limits,
- Idx dst_node, Idx dst_idx, Idx src_node,
- Idx src_idx) internal_function;
-static int check_dst_limits_calc_pos_1 (re_match_context_t *mctx,
+static bool check_dst_limits (const re_match_context_t *mctx,
+ const re_node_set *limits,
+ Idx dst_node, Idx dst_idx, Idx src_node,
+ Idx src_idx) internal_function;
+static int check_dst_limits_calc_pos_1 (const re_match_context_t *mctx,
int boundaries, Idx subexp_idx,
Idx from_node, Idx bkref_idx) internal_function;
-static int check_dst_limits_calc_pos (re_match_context_t *mctx,
+static int check_dst_limits_calc_pos (const re_match_context_t *mctx,
Idx limit, Idx subexp_idx,
Idx node, Idx str_idx,
Idx bkref_idx) internal_function;
static reg_errcode_t expand_bkref_cache (re_match_context_t *mctx,
re_node_set *cur_nodes, Idx cur_str,
Idx subexp_num, int type) internal_function;
-static int build_trtable (re_dfa_t *dfa,
- re_dfastate_t *state) internal_function;
+static bool build_trtable (re_dfa_t *dfa,
+ re_dfastate_t *state) internal_function;
#ifdef RE_ENABLE_I18N
static int check_node_accept_bytes (re_dfa_t *dfa, Idx node_idx,
const re_string_t *input, Idx idx) internal_function;
size_t name_len) internal_function;
# endif /* _LIBC */
#endif /* RE_ENABLE_I18N */
-static Idx group_nodes_into_DFAstates (re_dfa_t *dfa,
+static Idx group_nodes_into_DFAstates (const re_dfa_t *dfa,
const re_dfastate_t *state,
re_node_set *states_node,
bitset *states_ch) internal_function;
-static int check_node_accept (const re_match_context_t *mctx,
- const re_token_t *node, Idx idx) internal_function;
+static bool check_node_accept (const re_match_context_t *mctx,
+ const re_token_t *node, Idx idx)
+ internal_function;
static reg_errcode_t extend_buffers (re_match_context_t *mctx) internal_function;
\f
/* Entry point for POSIX code. */
__libc_lock_lock (dfa->lock);
if (preg->re_no_sub)
- err = re_search_internal (preg, string, length, start, length - start,
+ err = re_search_internal (preg, string, length, start, length,
length, 0, NULL, eflags);
else
- err = re_search_internal (preg, string, length, start, length - start,
+ err = re_search_internal (preg, string, length, start, length,
length, nmatch, pmatch, eflags);
__libc_lock_unlock (dfa->lock);
return err != REG_NOERROR;
re_match (struct re_pattern_buffer *bufp, const char *string,
Idx length, Idx start, struct re_registers *regs)
{
- return re_search_stub (bufp, string, length, start, 0, length, regs, 1);
+ return re_search_stub (bufp, string, length, start, 0, length, regs, true);
}
#ifdef _LIBC
weak_alias (__re_match, re_match)
re_search (struct re_pattern_buffer *bufp, const char *string,
Idx length, Idx start, regoff_t range, struct re_registers *regs)
{
- return re_search_stub (bufp, string, length, start, range, length, regs, 0);
+ return re_search_stub (bufp, string, length, start, range, length, regs,
+ false);
}
#ifdef _LIBC
weak_alias (__re_search, re_search)
Idx start, struct re_registers *regs, Idx stop)
{
return re_search_2_stub (bufp, string1, length1, string2, length2,
- start, 0, regs, stop, 1);
+ start, 0, regs, stop, true);
}
#ifdef _LIBC
weak_alias (__re_match_2, re_match_2)
Idx start, regoff_t range, struct re_registers *regs, Idx stop)
{
return re_search_2_stub (bufp, string1, length1, string2, length2,
- start, range, regs, stop, 0);
+ start, range, regs, stop, false);
}
#ifdef _LIBC
weak_alias (__re_search_2, re_search_2)
const char *string1, Idx length1,
const char *string2, Idx length2,
Idx start, regoff_t range, struct re_registers *regs,
- Idx stop, int ret_len)
+ Idx stop, bool ret_len)
{
const char *str;
regoff_t rval;
Idx len = length1 + length2;
- int free_str = 0;
+ char *s = NULL;
- if (BE (length1 < 0 || length2 < 0 || stop < 0, 0))
+ if (BE (length1 < 0 || length2 < 0 || stop < 0 || len < length1, 0))
return -2;
/* Concatenate the strings. */
if (length2 > 0)
if (length1 > 0)
{
- char *s = re_malloc (char, len);
+ s = re_malloc (char, len);
if (BE (s == NULL, 0))
return -2;
memcpy (s, string1, length1);
memcpy (s + length1, string2, length2);
str = s;
- free_str = 1;
}
else
str = string2;
rval = re_search_stub (bufp, str, len, start, range, stop, regs,
ret_len);
- if (free_str)
- re_free ((char *) str);
+ re_free (s);
return rval;
}
/* The parameters have the same meaning as those of re_search.
Additional parameters:
- If RET_LEN is nonzero the length of the match is returned (re_match style);
+ If RET_LEN is true the length of the match is returned (re_match style);
otherwise the position of the match is returned. */
static regoff_t
re_search_stub (struct re_pattern_buffer *bufp,
const char *string, Idx length,
Idx start, regoff_t range, Idx stop, struct re_registers *regs,
- int ret_len)
+ bool ret_len)
{
reg_errcode_t result;
regmatch_t *pmatch;
#ifdef _LIBC
re_dfa_t *dfa = (re_dfa_t *) bufp->re_buffer;
#endif
+ Idx last_start = start + range;
/* Check for out-of-range. */
if (BE (start < 0 || start > length, 0))
regoff_t length_offset = length;
regoff_t start_offset = start;
if (BE (length_offset - start_offset < range, 0))
- range = length_offset - start_offset;
+ last_start = length;
else if (BE (range < - start_offset, 0))
- range = -start_offset;
+ last_start = 0;
}
else
{
- if (BE (start + range > length, 0))
- range = length - start;
- else if (BE (start + range < 0, 0))
- range = -start;
+ if (BE ((last_start < start) != (range < 0), 0))
+ {
+ /* Overflow occurred when computing last_start; substitute
+ the extreme value. */
+ last_start = range < 0 ? 0 : length;
+ }
+ else
+ {
+ if (BE (length < last_start, 0))
+ last_start = length;
+ else if (BE (last_start < 0, 0))
+ last_start = 0;
+ }
}
__libc_lock_lock (dfa->lock);
eflags |= (bufp->re_not_eol) ? REG_NOTEOL : 0;
/* Compile fastmap if we haven't yet. */
- if (range > 0 && bufp->re_fastmap != NULL && !bufp->re_fastmap_accurate)
+ if (start < last_start && bufp->re_fastmap != NULL
+ && !bufp->re_fastmap_accurate)
re_compile_fastmap (bufp);
if (BE (bufp->re_no_sub, 0))
if (regs == NULL)
nregs = 1;
else if (BE (bufp->re_regs_allocated == REG_FIXED
- && regs->rm_num_regs < bufp->re_nsub + 1, 0))
+ && regs->rm_num_regs <= bufp->re_nsub, 0))
{
nregs = regs->rm_num_regs;
if (BE (nregs < 1, 0))
}
else
nregs = bufp->re_nsub + 1;
- pmatch = re_malloc (regmatch_t, nregs);
+ pmatch = re_xmalloc (regmatch_t, nregs);
if (BE (pmatch == NULL, 0))
{
rval = -2;
goto out;
}
- result = re_search_internal (bufp, string, length, start, range, stop,
+ result = re_search_internal (bufp, string, length, start, last_start, stop,
nregs, pmatch, eflags);
rval = 0;
/* Have the register data arrays been allocated? */
if (regs_allocated == REG_UNALLOCATED)
{ /* No. So allocate them with malloc. */
- regs->rm_start = re_malloc (regoff_t, need_regs);
+ regs->rm_start = re_xmalloc (regoff_t, need_regs);
regs->rm_end = re_malloc (regoff_t, need_regs);
if (BE (regs->rm_start == NULL, 0) || BE (regs->rm_end == NULL, 0))
return REG_UNALLOCATED;
if (BE (need_regs > regs->rm_num_regs, 0))
{
regoff_t *new_start =
- re_realloc (regs->rm_start, regoff_t, need_regs);
+ re_xrealloc (regs->rm_start, regoff_t, need_regs);
regoff_t *new_end = re_realloc (regs->rm_end, regoff_t, need_regs);
if (BE (new_start == NULL, 0) || BE (new_end == NULL, 0))
return REG_UNALLOCATED;
/* Searches for a compiled pattern PREG in the string STRING, whose
length is LENGTH. NMATCH, PMATCH, and EFLAGS have the same
- mingings with regexec. START, and RANGE have the same meanings
- with re_search.
+ meaning as with regexec. LAST_START is START + RANGE, where
+ START and RANGE have the same meaning as with re_search.
Return REG_NOERROR if we find a match, and REG_NOMATCH if not,
otherwise return the error code.
Note: We assume front end functions already check ranges.
- (START + RANGE >= 0 && START + RANGE <= LENGTH) */
+ (0 <= LAST_START && LAST_START <= LENGTH) */
static reg_errcode_t
internal_function
re_search_internal (const regex_t *preg,
const char *string, Idx length,
- Idx start, regoff_t range, Idx stop,
+ Idx start, Idx last_start, Idx stop,
size_t nmatch, regmatch_t pmatch[],
int eflags)
{
re_dfa_t *dfa = (re_dfa_t *) preg->re_buffer;
Idx left_lim, right_lim;
int incr;
- int fl_longest_match, match_kind;
+ bool fl_longest_match;
+ int match_kind;
Idx match_first, match_last = REG_MISSING;
Idx extra_nmatch;
- int sb, ch;
+ bool sb;
+ int ch;
#if defined _LIBC || (defined __STDC_VERSION__ && __STDC_VERSION__ >= 199901L)
re_match_context_t mctx = { .dfa = dfa };
#else
re_match_context_t mctx;
#endif
- char *fastmap = (preg->re_fastmap != NULL && preg->re_fastmap_accurate
- && range && !preg->re_can_be_null) ? preg->re_fastmap : NULL;
+ char *fastmap = ((preg->re_fastmap != NULL && preg->re_fastmap_accurate
+ && start != last_start && !preg->re_can_be_null)
+ ? preg->re_fastmap : NULL);
unsigned REG_TRANSLATE_TYPE t =
(unsigned REG_TRANSLATE_TYPE) preg->re_translate;
#ifdef DEBUG
/* We assume front-end functions already check them. */
- assert (start + range >= 0 && start + range <= length);
+ assert (0 <= last_start && last_start <= length);
#endif
/* If initial states with non-begbuf contexts have no elements,
&& (dfa->init_state_nl->nodes.nelem == 0
|| !preg->re_newline_anchor))
{
- if (start != 0 && start + range != 0)
+ if (start != 0 && last_start != 0)
return REG_NOMATCH;
- start = range = 0;
+ start = last_start = 0;
}
/* We must check the longest matching, if nmatch > 0. */
multi character collating element. */
if (nmatch > 1 || dfa->has_mb_node)
{
- mctx.state_log = re_malloc (re_dfastate_t *, mctx.input.bufs_len + 1);
+ mctx.state_log = re_xmalloc (re_dfastate_t *, mctx.input.bufs_len + 1);
if (BE (mctx.state_log == NULL, 0))
{
err = REG_ESPACE;
: CONTEXT_NEWLINE | CONTEXT_BEGBUF;
/* Check incrementally whether of not the input string match. */
- incr = (range < 0) ? -1 : 1;
- left_lim = (range < 0) ? start + range : start;
- right_lim = (range < 0) ? start : start + range;
+ incr = (last_start < start) ? -1 : 1;
+ left_lim = (last_start < start) ? last_start : start;
+ right_lim = (last_start < start) ? start : last_start;
sb = dfa->mb_cur_max == 1;
match_kind =
(fastmap
? ((sb || !(preg->re_syntax & REG_IGNORE_CASE || t) ? 4 : 0)
- | (range >= 0 ? 2 : 0)
+ | (start <= last_start ? 2 : 0)
| (t != NULL ? 1 : 0))
: 8);
/* We assume that the matching starts from 0. */
mctx.state_log_top = mctx.nbkref_ents = mctx.max_mb_elem_len = 0;
match_last = check_matching (&mctx, fl_longest_match,
- range >= 0 ? &match_first : NULL);
+ start <= last_start ? &match_first : NULL);
if (match_last != REG_MISSING)
{
if (BE (match_last == REG_ERROR, 0))
#endif
match_last = mctx->match_last;
halt_node = mctx->last_node;
- sifted_states = re_malloc (re_dfastate_t *, match_last + 1);
+ sifted_states = re_xmalloc (re_dfastate_t *, match_last + 1);
if (BE (sifted_states == NULL, 0))
{
ret = REG_ESPACE;
}
if (dfa->nbackref)
{
- lim_states = re_malloc (re_dfastate_t *, match_last + 1);
+ lim_states = re_xmalloc (re_dfastate_t *, match_last + 1);
if (BE (lim_states == NULL, 0))
{
ret = REG_ESPACE;
static Idx
internal_function
-check_matching (re_match_context_t *mctx, int fl_longest_match,
+check_matching (re_match_context_t *mctx, bool fl_longest_match,
Idx *p_match_first)
{
re_dfa_t *const dfa = mctx->dfa;
Idx match_last = REG_MISSING;
Idx cur_str_idx = re_string_cur_idx (&mctx->input);
re_dfastate_t *cur_state;
- int at_init_state = p_match_first != NULL;
+ bool at_init_state = p_match_first != NULL;
Idx next_start_idx = cur_str_idx;
err = REG_NOERROR;
later. E.g. Processing back references. */
if (BE (dfa->nbackref, 0))
{
- at_init_state = 0;
+ at_init_state = false;
err = check_subexp_matching_top (mctx, &cur_state->nodes, 0);
if (BE (err != REG_NOERROR, 0))
return err;
if (old_state == cur_state)
next_start_idx = next_char_idx;
else
- at_init_state = 0;
+ at_init_state = false;
}
if (cur_state->halt)
/* Check NODE match the current context. */
-static int
+static bool
internal_function
check_halt_node_context (const re_dfa_t *dfa, Idx node, unsigned int context)
{
re_token_type_t type = dfa->nodes[node].type;
unsigned int constraint = dfa->nodes[node].constraint;
if (type != END_OF_RE)
- return 0;
+ return false;
if (!constraint)
- return 1;
+ return true;
if (NOT_SATISFY_NEXT_CONSTRAINT (constraint, context))
- return 0;
- return 1;
+ return false;
+ return true;
}
/* Check the halt state STATE match the current context.
{
re_dfa_t *const dfa = mctx->dfa;
Idx i;
- int err;
+ bool ok;
if (IS_EPSILON_NODE (dfa->nodes[node].type))
{
re_node_set *cur_nodes = &mctx->state_log[*pidx]->nodes;
re_node_set *edests = &dfa->edests[node];
Idx dest_node;
- err = re_node_set_insert (eps_via_nodes, node);
- if (BE (err < 0, 0))
+ ok = re_node_set_insert (eps_via_nodes, node);
+ if (BE (! ok, 0))
return REG_ERROR;
/* Pick up a valid destination, or return REG_MISSING if none
is found. */
if (naccepted == 0)
{
Idx dest_node;
- err = re_node_set_insert (eps_via_nodes, node);
- if (BE (err < 0, 0))
+ ok = re_node_set_insert (eps_via_nodes, node);
+ if (BE (! ok, 0))
return REG_ERROR;
dest_node = dfa->edests[node].elems[0];
if (re_node_set_contains (&mctx->state_log[*pidx]->nodes,
if (fs->num == fs->alloc)
{
struct re_fail_stack_ent_t *new_array =
- re_realloc (fs->stack, struct re_fail_stack_ent_t, fs->alloc * 2);
+ re_x2realloc (fs->stack, struct re_fail_stack_ent_t, &fs->alloc);
if (new_array == NULL)
return REG_ESPACE;
- fs->alloc *= 2;
fs->stack = new_array;
}
fs->stack[num].idx = str_idx;
fs->stack[num].node = dest_node;
- fs->stack[num].regs = re_malloc (regmatch_t, nregs);
+ fs->stack[num].regs = re_xmalloc (regmatch_t, nregs);
if (fs->stack[num].regs == NULL)
return REG_ESPACE;
memcpy (fs->stack[num].regs, regs, sizeof (regmatch_t) * nregs);
static reg_errcode_t
internal_function
set_regs (const regex_t *preg, const re_match_context_t *mctx,
- size_t nmatch, regmatch_t *pmatch, int fl_backtrack)
+ size_t nmatch, regmatch_t *pmatch, bool fl_backtrack)
{
re_dfa_t *dfa = (re_dfa_t *) preg->re_buffer;
Idx idx, cur_node;
struct re_fail_stack_t *fs;
struct re_fail_stack_t fs_body = { 0, 2, NULL };
regmatch_t *prev_idx_match;
- int prev_idx_match_malloced = 0;
+ bool prev_idx_match_malloced = false;
#ifdef DEBUG
assert (nmatch > 1);
if (fl_backtrack)
{
fs = &fs_body;
- fs->stack = re_malloc (struct re_fail_stack_ent_t, fs->alloc);
+ fs->stack = re_xmalloc (struct re_fail_stack_ent_t, fs->alloc);
if (fs->stack == NULL)
return REG_ESPACE;
}
cur_node = dfa->init_node;
re_node_set_init_empty (&eps_via_nodes);
+ if (re_alloc_oversized (nmatch, sizeof (regmatch_t)))
+ {
+ free_fail_stack_return (fs);
+ return REG_ESPACE;
+ }
if (__libc_use_alloca (nmatch * sizeof (regmatch_t)))
prev_idx_match = (regmatch_t *) alloca (nmatch * sizeof (regmatch_t));
else
free_fail_stack_return (fs);
return REG_ESPACE;
}
- prev_idx_match_malloced = 1;
+ prev_idx_match_malloced = true;
}
memcpy (prev_idx_match, pmatch, sizeof (regmatch_t) * nmatch);
{
Idx prev_node = cur_src->elems[i];
int naccepted = 0;
- int ret;
+ bool ok;
#ifdef DEBUG
re_token_type_t type = dfa->nodes[prev_node].type;
prev_node, str_idx))
continue;
}
- ret = re_node_set_insert (cur_dest, prev_node);
- if (BE (ret == -1, 0))
+ ok = re_node_set_insert (cur_dest, prev_node);
+ if (BE (! ok, 0))
return REG_ESPACE;
}
return REG_NOERROR;
}
-static int
+static bool
internal_function
-check_dst_limits (re_match_context_t *mctx, re_node_set *limits,
+check_dst_limits (const re_match_context_t *mctx, const re_node_set *limits,
Idx dst_node, Idx dst_idx, Idx src_node, Idx src_idx)
{
re_dfa_t *const dfa = mctx->dfa;
if (src_pos == dst_pos)
continue; /* This is unrelated limitation. */
else
- return 1;
+ return true;
}
- return 0;
+ return false;
}
static int
internal_function
-check_dst_limits_calc_pos_1 (re_match_context_t *mctx, int boundaries,
+check_dst_limits_calc_pos_1 (const re_match_context_t *mctx, int boundaries,
Idx subexp_idx, Idx from_node, Idx bkref_idx)
{
re_dfa_t *const dfa = mctx->dfa;
if (ent->node != node)
continue;
- if (subexp_idx
- < CHAR_BIT * sizeof ent->eps_reachable_subexps_map
- && !(ent->eps_reachable_subexps_map & (1u << subexp_idx)))
+ if (subexp_idx < BITSET_WORD_BITS
+ && !(ent->eps_reachable_subexps_map
+ & ((bitset_word) 1 << subexp_idx)))
continue;
/* Recurse trying to reach the OP_OPEN_SUBEXP and
if (cpos == 0 && (boundaries & 2))
return 0;
- if (subexp_idx
- < CHAR_BIT * sizeof ent->eps_reachable_subexps_map)
- ent->eps_reachable_subexps_map &= ~(1u << subexp_idx);
+ if (subexp_idx < BITSET_WORD_BITS)
+ ent->eps_reachable_subexps_map &=
+ ~ ((bitset_word) 1 << subexp_idx);
}
while (ent++->more);
}
static int
internal_function
-check_dst_limits_calc_pos (re_match_context_t *mctx, Idx limit, Idx subexp_idx,
+check_dst_limits_calc_pos (const re_match_context_t *mctx,
+ Idx limit, Idx subexp_idx,
Idx from_node, Idx str_idx, Idx bkref_idx)
{
struct re_backref_cache_entry *lim = mctx->bkref_ents + limit;
enabled_idx = first_idx;
do
{
- Idx subexp_len, to_idx, dst_node, ret;
+ bool ok;
+ Idx subexp_len, to_idx, dst_node;
re_dfastate_t *cur_state;
if (entry->node != node)
}
local_sctx.last_node = node;
local_sctx.last_str_idx = str_idx;
- ret = re_node_set_insert (&local_sctx.limits, enabled_idx);
- if (BE (ret < 0, 0))
+ ok = re_node_set_insert (&local_sctx.limits, enabled_idx);
+ if (BE (! ok, 0))
{
err = REG_ESPACE;
goto free_return;
{
Idx node = cur_nodes->elems[node_idx];
if (dfa->nodes[node].type == OP_OPEN_SUBEXP
- && dfa->nodes[node].opr.idx < CHAR_BIT * sizeof dfa->used_bkref_map
- && dfa->used_bkref_map & (1u << dfa->nodes[node].opr.idx))
+ && dfa->nodes[node].opr.idx < BITSET_WORD_BITS
+ && (dfa->used_bkref_map
+ & ((bitset_word) 1 << dfa->nodes[node].opr.idx)))
{
err = match_ctx_add_subtop (mctx, node, str_idx);
if (BE (err != REG_NOERROR, 0))
{
re_dfastate_t **new_array;
Idx old_alloc = path->alloc;
- path->alloc += last_str + mctx->max_mb_elem_len + 1;
- new_array = re_realloc (path->array, re_dfastate_t *, path->alloc);
- if (new_array == NULL)
- {
- path->alloc = old_alloc;
- return REG_ESPACE;
- }
+ Idx new_alloc = old_alloc + last_str + mctx->max_mb_elem_len + 1;
+ if (BE (new_alloc < old_alloc, 0))
+ return REG_ESPACE;
+ new_array = re_xrealloc (path->array, re_dfastate_t *, new_alloc);
+ if (BE (new_array == NULL, 0))
+ return REG_ESPACE;
path->array = new_array;
+ path->alloc = new_alloc;
memset (new_array + old_alloc, '\0',
- sizeof (re_dfastate_t *) * (path->alloc - old_alloc));
+ sizeof (re_dfastate_t *) * (new_alloc - old_alloc));
}
str_idx = path->next_idx == 0 ? top_str : path->next_idx;
re_node_set *next_nodes)
{
re_dfa_t *const dfa = mctx->dfa;
- int result;
+ bool ok;
Idx cur_idx;
reg_errcode_t err;
re_node_set union_set;
return err;
}
}
- result = re_node_set_insert (&union_set, next_node);
- if (BE (result < 0, 0))
+ ok = re_node_set_insert (&union_set, next_node);
+ if (BE (! ok, 0))
{
re_node_set_free (&union_set);
return REG_ESPACE;
if (naccepted
|| check_node_accept (mctx, dfa->nodes + cur_node, str_idx))
{
- result = re_node_set_insert (next_nodes, dfa->nexts[cur_node]);
- if (BE (result < 0, 0))
+ ok = re_node_set_insert (next_nodes, dfa->nexts[cur_node]);
+ if (BE (! ok, 0))
{
re_node_set_free (&union_set);
return REG_ESPACE;
Idx cur_node;
for (cur_node = target; !re_node_set_contains (dst_nodes, cur_node);)
{
- int err;
+ bool ok;
if (dfa->nodes[cur_node].type == type
&& dfa->nodes[cur_node].opr.idx == ex_subexp)
{
if (type == OP_CLOSE_SUBEXP)
{
- err = re_node_set_insert (dst_nodes, cur_node);
- if (BE (err == -1, 0))
+ ok = re_node_set_insert (dst_nodes, cur_node);
+ if (BE (! ok, 0))
return REG_ESPACE;
}
break;
}
- err = re_node_set_insert (dst_nodes, cur_node);
- if (BE (err == -1, 0))
+ ok = re_node_set_insert (dst_nodes, cur_node);
+ if (BE (! ok, 0))
return REG_ESPACE;
if (dfa->edests[cur_node].nelem == 0)
break;
next_node = dfa->nexts[ent->node];
if (mctx->state_log[to_idx])
{
- int ret;
+ bool ok;
if (re_node_set_contains (&mctx->state_log[to_idx]->nodes,
next_node))
continue;
err = re_node_set_init_copy (&union_set,
&mctx->state_log[to_idx]->nodes);
- ret = re_node_set_insert (&union_set, next_node);
- if (BE (err != REG_NOERROR || ret < 0, 0))
+ ok = re_node_set_insert (&union_set, next_node);
+ if (BE (err != REG_NOERROR || ! ok, 0))
{
re_node_set_free (&union_set);
err = err != REG_NOERROR ? err : REG_ESPACE;
}
/* Build transition table for the state.
- Return 1 if succeeded, otherwise return NULL. */
+ Return true if successful. */
-static int
+static bool
internal_function
build_trtable (re_dfa_t *dfa, re_dfastate_t *state)
{
reg_errcode_t err;
Idx i, j;
- int ch, need_word_trtable = 0;
- unsigned int elem, mask;
- int dests_node_malloced = 0, dest_states_malloced = 0;
+ int ch;
+ bool need_word_trtable = false;
+ bitset_word elem, mask;
+ bool dests_node_malloced = false, dest_states_malloced = false;
Idx ndests; /* Number of the destination states from `state'. */
re_dfastate_t **trtable;
re_dfastate_t **dest_states = NULL, **dest_states_word, **dest_states_nl;
bitset *dests_ch;
bitset acceptable;
+ struct dests_alloc
+ {
+ re_node_set dests_node[SBC_MAX];
+ bitset dests_ch[SBC_MAX];
+ } *dests_alloc;
+
/* We build DFA states which corresponds to the destination nodes
from `state'. `dests_node[i]' represents the nodes which i-th
destination state contains, and `dests_ch[i]' represents the
characters which i-th destination state accepts. */
- if (__libc_use_alloca ((sizeof (re_node_set) + sizeof (bitset)) * SBC_MAX))
- dests_node = (re_node_set *)
- alloca ((sizeof (re_node_set) + sizeof (bitset)) * SBC_MAX);
+ if (__libc_use_alloca (sizeof (struct dests_alloc)))
+ dests_alloc = (struct dests_alloc *) alloca (sizeof dests_alloc[0]);
else
{
- dests_node = (re_node_set *)
- malloc ((sizeof (re_node_set) + sizeof (bitset)) * SBC_MAX);
- if (BE (dests_node == NULL, 0))
- return 0;
- dests_node_malloced = 1;
+ dests_alloc = re_malloc (struct dests_alloc, 1);
+ if (BE (dests_alloc == NULL, 0))
+ return false;
+ dests_node_malloced = true;
}
- dests_ch = (bitset *) (dests_node + SBC_MAX);
+ dests_node = dests_alloc->dests_node;
+ dests_ch = dests_alloc->dests_ch;
/* Initialize transiton table. */
state->word_trtable = state->trtable = NULL;
if (BE (! REG_VALID_NONZERO_INDEX (ndests), 0))
{
if (dests_node_malloced)
- free (dests_node);
- /* Return 0 in case of an error, 1 otherwise. */
+ free (dests_alloc);
if (ndests == 0)
{
state->trtable = re_calloc (re_dfastate_t *, SBC_MAX);
- return 1;
+ return true;
}
- return 0;
+ return false;
}
err = re_node_set_alloc (&follows, ndests + 1);
if (BE (err != REG_NOERROR, 0))
goto out_free;
+ /* Avoid arithmetic overflow in size calculation. */
+ if (BE (((SIZE_MAX - (sizeof (re_node_set) + sizeof (bitset)) * SBC_MAX)
+ / (3 * sizeof (re_dfastate_t *)))
+ < ndests, 0))
+ goto out_free;
+
if (__libc_use_alloca ((sizeof (re_node_set) + sizeof (bitset)) * SBC_MAX
+ ndests * 3 * sizeof (re_dfastate_t *)))
dest_states = (re_dfastate_t **)
for (i = 0; i < ndests; ++i)
re_node_set_free (dests_node + i);
if (dests_node_malloced)
- free (dests_node);
- return 0;
+ free (dests_alloc);
+ return false;
}
- dest_states_malloced = 1;
+ dest_states_malloced = true;
}
dest_states_word = dest_states + ndests;
dest_states_nl = dest_states_word + ndests;
goto out_free;
if (dest_states[i] != dest_states_word[i] && dfa->mb_cur_max > 1)
- need_word_trtable = 1;
+ need_word_trtable = true;
dest_states_nl[i] = re_acquire_state_context (&err, dfa, &follows,
CONTEXT_NEWLINE);
goto out_free;
/* For all characters ch...: */
- for (i = 0; i < BITSET_UINTS; ++i)
- for (ch = i * UINT_BITS, elem = acceptable[i], mask = 1;
+ for (i = 0; i < BITSET_WORDS; ++i)
+ for (ch = i * BITSET_WORD_BITS, elem = acceptable[i], mask = 1;
elem;
mask <<= 1, elem >>= 1, ++ch)
if (BE (elem & 1, 0))
goto out_free;
/* For all characters ch...: */
- for (i = 0; i < BITSET_UINTS; ++i)
- for (ch = i * UINT_BITS, elem = acceptable[i], mask = 1;
+ for (i = 0; i < BITSET_WORDS; ++i)
+ for (ch = i * BITSET_WORD_BITS, elem = acceptable[i], mask = 1;
elem;
mask <<= 1, elem >>= 1, ++ch)
if (BE (elem & 1, 0))
re_node_set_free (dests_node + i);
if (dests_node_malloced)
- free (dests_node);
+ free (dests_alloc);
- return 1;
+ return true;
}
/* Group all nodes belonging to STATE into several destinations.
static Idx
internal_function
-group_nodes_into_DFAstates (re_dfa_t *dfa, const re_dfastate_t *state,
+group_nodes_into_DFAstates (const re_dfa_t *dfa, const re_dfastate_t *state,
re_node_set *dests_node, bitset *dests_ch)
{
reg_errcode_t err;
- int result;
+ bool ok;
Idx i, j, k;
Idx ndests; /* Number of the destinations from `state'. */
bitset accepts; /* Characters a node can accept. */
#ifdef RE_ENABLE_I18N
else if (type == OP_UTF8_PERIOD)
{
- memset (accepts, 255, sizeof (unsigned int) * BITSET_UINTS / 2);
+ if (SBC_MAX / 2 % BITSET_WORD_BITS == 0)
+ memset (accepts, -1, sizeof accepts / 2);
+ else
+ bitset_merge (accepts, utf8_sb_map);
if (!(dfa->syntax & REG_DOT_NEWLINE))
bitset_clear (accepts, '\n');
if (dfa->syntax & REG_DOT_NOT_NULL)
{
if (constraint & NEXT_NEWLINE_CONSTRAINT)
{
- int accepts_newline = bitset_contain (accepts, NEWLINE_CHAR);
+ bool accepts_newline = bitset_contain (accepts, NEWLINE_CHAR);
bitset_empty (accepts);
if (accepts_newline)
bitset_set (accepts, NEWLINE_CHAR);
if (constraint & NEXT_WORD_CONSTRAINT)
{
- unsigned int any_set = 0;
+ bitset_word any_set = 0;
if (type == CHARACTER && !node->word_char)
{
bitset_empty (accepts);
}
#ifdef RE_ENABLE_I18N
if (dfa->mb_cur_max > 1)
- for (j = 0; j < BITSET_UINTS; ++j)
+ for (j = 0; j < BITSET_WORDS; ++j)
any_set |= (accepts[j] &= (dfa->word_char[j] | ~dfa->sb_char[j]));
else
#endif
- for (j = 0; j < BITSET_UINTS; ++j)
+ for (j = 0; j < BITSET_WORDS; ++j)
any_set |= (accepts[j] &= dfa->word_char[j]);
if (!any_set)
continue;
}
if (constraint & NEXT_NOTWORD_CONSTRAINT)
{
- unsigned int any_set = 0;
+ bitset_word any_set = 0;
if (type == CHARACTER && node->word_char)
{
bitset_empty (accepts);
}
#ifdef RE_ENABLE_I18N
if (dfa->mb_cur_max > 1)
- for (j = 0; j < BITSET_UINTS; ++j)
+ for (j = 0; j < BITSET_WORDS; ++j)
any_set |= (accepts[j] &= ~(dfa->word_char[j] & dfa->sb_char[j]));
else
#endif
- for (j = 0; j < BITSET_UINTS; ++j)
+ for (j = 0; j < BITSET_WORDS; ++j)
any_set |= (accepts[j] &= ~dfa->word_char[j]);
if (!any_set)
continue;
bitset intersec; /* Intersection sets, see below. */
bitset remains;
/* Flags, see below. */
- int has_intersec, not_subset, not_consumed;
+ bitset_word has_intersec, not_subset, not_consumed;
/* Optimization, skip if this state doesn't accept the character. */
if (type == CHARACTER && !bitset_contain (dests_ch[j], node->opr.c))
/* Enumerate the intersection set of this state and `accepts'. */
has_intersec = 0;
- for (k = 0; k < BITSET_UINTS; ++k)
+ for (k = 0; k < BITSET_WORDS; ++k)
has_intersec |= intersec[k] = accepts[k] & dests_ch[j][k];
/* And skip if the intersection set is empty. */
if (!has_intersec)
/* Then check if this state is a subset of `accepts'. */
not_subset = not_consumed = 0;
- for (k = 0; k < BITSET_UINTS; ++k)
+ for (k = 0; k < BITSET_WORDS; ++k)
{
not_subset |= remains[k] = ~accepts[k] & dests_ch[j][k];
not_consumed |= accepts[k] = accepts[k] & ~dests_ch[j][k];
}
/* Put the position in the current group. */
- result = re_node_set_insert (&dests_node[j], cur_nodes->elems[i]);
- if (BE (result < 0, 0))
+ ok = re_node_set_insert (&dests_node[j], cur_nodes->elems[i]);
+ if (BE (! ok, 0))
goto error_return;
/* If all characters are consumed, go to next node. */
for (idx = 0; idx < extrasize;)
{
- int mbs_cnt, found = 0;
+ int mbs_cnt;
+ bool found = false;
int32_t elem_mbs_len;
/* Skip the name of collating element name. */
idx = idx + extra[idx] + 1;
break;
if (mbs_cnt == elem_mbs_len)
/* Found the entry. */
- found = 1;
+ found = true;
}
/* Skip the byte sequence of the collating element. */
idx += elem_mbs_len;
/* Check whether the node accepts the byte which is IDX-th
byte of the INPUT. */
-static int
+static bool
internal_function
check_node_accept (const re_match_context_t *mctx, const re_token_t *node,
Idx idx)
{
case CHARACTER:
if (node->opr.c != ch)
- return 0;
+ return false;
break;
case SIMPLE_BRACKET:
if (!bitset_contain (node->opr.sbcset, ch))
- return 0;
+ return false;
break;
#ifdef RE_ENABLE_I18N
case OP_UTF8_PERIOD:
if (ch >= 0x80)
- return 0;
+ return false;
/* FALLTHROUGH */
#endif
case OP_PERIOD:
if ((ch == '\n' && !(mctx->dfa->syntax & REG_DOT_NEWLINE))
|| (ch == '\0' && (mctx->dfa->syntax & REG_DOT_NOT_NULL)))
- return 0;
+ return false;
break;
default:
- return 0;
+ return false;
}
if (node->constraint)
unsigned int context = re_string_context_at (&mctx->input, idx,
mctx->eflags);
if (NOT_SATISFY_NEXT_CONSTRAINT (node->constraint, context))
- return 0;
+ return false;
}
- return 1;
+ return true;
}
/* Extend the buffers, if the buffers have run out. */
/* XXX We have no indication of the size of this buffer. If this
allocation fail we have no indication that the state_log array
does not have the right size. */
- re_dfastate_t **new_array = re_realloc (mctx->state_log, re_dfastate_t *,
- pstr->bufs_len + 1);
+ re_dfastate_t **new_array = re_xrealloc (mctx->state_log, re_dfastate_t *,
+ pstr->bufs_len + 1);
if (BE (new_array == NULL, 0))
return REG_ESPACE;
mctx->state_log = new_array;
mctx->match_last = REG_MISSING;
if (n > 0)
{
- mctx->bkref_ents = re_malloc (struct re_backref_cache_entry, n);
- mctx->sub_tops = re_malloc (re_sub_match_top_t *, n);
+ mctx->bkref_ents = re_xmalloc (struct re_backref_cache_entry, n);
+ mctx->sub_tops = re_xmalloc (re_sub_match_top_t *, n);
if (BE (mctx->bkref_ents == NULL || mctx->sub_tops == NULL, 0))
return REG_ESPACE;
}
if (mctx->nbkref_ents >= mctx->abkref_ents)
{
struct re_backref_cache_entry* new_entry;
- new_entry = re_realloc (mctx->bkref_ents, struct re_backref_cache_entry,
- mctx->abkref_ents * 2);
+ new_entry = re_x2realloc (mctx->bkref_ents, struct re_backref_cache_entry,
+ &mctx->abkref_ents);
if (BE (new_entry == NULL, 0))
{
re_free (mctx->bkref_ents);
}
mctx->bkref_ents = new_entry;
memset (mctx->bkref_ents + mctx->nbkref_ents, '\0',
- sizeof (struct re_backref_cache_entry) * mctx->abkref_ents);
- mctx->abkref_ents *= 2;
+ (sizeof (struct re_backref_cache_entry)
+ * (mctx->abkref_ents - mctx->nbkref_ents)));
}
if (mctx->nbkref_ents > 0
&& mctx->bkref_ents[mctx->nbkref_ents - 1].str_idx == str_idx)
static Idx
internal_function
-search_cur_bkref_entry (re_match_context_t *mctx, Idx str_idx)
+search_cur_bkref_entry (const re_match_context_t *mctx, Idx str_idx)
{
Idx left, right, mid, last;
last = right = mctx->nbkref_ents;
#endif
if (BE (mctx->nsub_tops == mctx->asub_tops, 0))
{
- Idx new_asub_tops = mctx->asub_tops * 2;
- re_sub_match_top_t **new_array = re_realloc (mctx->sub_tops,
- re_sub_match_top_t *,
- new_asub_tops);
+ Idx new_asub_tops = mctx->asub_tops;
+ re_sub_match_top_t **new_array = re_x2realloc (mctx->sub_tops,
+ re_sub_match_top_t *,
+ &new_asub_tops);
if (BE (new_array == NULL, 0))
return REG_ESPACE;
mctx->sub_tops = new_array;
re_sub_match_last_t *new_entry;
if (BE (subtop->nlasts == subtop->alasts, 0))
{
- Idx new_alasts = 2 * subtop->alasts + 1;
- re_sub_match_last_t **new_array = re_realloc (subtop->lasts,
- re_sub_match_last_t *,
- new_alasts);
+ Idx new_alasts = subtop->alasts;
+ re_sub_match_last_t **new_array = re_x2realloc (subtop->lasts,
+ re_sub_match_last_t *,
+ &new_alasts);
if (BE (new_array == NULL, 0))
return NULL;
subtop->lasts = new_array;
projects / gnulib.git / blobdiff