630 lines
24 KiB
Diff
630 lines
24 KiB
Diff
|
commit 86a701a20479dfbc23540b3143fd5b28660a2447
|
|||
|
Author: Paul Eggert <eggert@cs.ucla.edu>
|
|||
|
Date: Tue Sep 21 07:47:45 2021 -0700
|
|||
|
|
|||
|
regex: copy back from Gnulib
|
|||
|
|
|||
|
Copy regex-related files back from Gnulib, to fix a problem with
|
|||
|
static checking of regex calls noted by Martin Sebor. This merges the
|
|||
|
following changes:
|
|||
|
|
|||
|
* New macro __attribute_nonnull__ in misc/sys/cdefs.h, for use later
|
|||
|
when copying other files back from Gnulib.
|
|||
|
|
|||
|
* Use __GNULIB_CDEFS instead of __GLIBC__ when deciding
|
|||
|
whether to include bits/wordsize.h etc.
|
|||
|
|
|||
|
* Avoid duplicate entries in epsilon closure table.
|
|||
|
|
|||
|
* New regex.h macro _REGEX_NELTS to let regexec say that its pmatch
|
|||
|
arg should contain nmatch elts. Use that for regexec, instead of
|
|||
|
__attr_access (which is incorrect).
|
|||
|
|
|||
|
* New regex.h macro _Attr_access_ which is like __attr_access except
|
|||
|
portable to non-glibc platforms.
|
|||
|
|
|||
|
* Add some DEBUG_ASSERTs to pacify gcc -fanalyzer and to catch
|
|||
|
recently-fixed performance bugs if they recur.
|
|||
|
|
|||
|
* Add Gnulib-specific stuff to port the dynarray- and lock-using parts
|
|||
|
of regex code to non-glibc platforms.
|
|||
|
|
|||
|
* Fix glibc bug 11053.
|
|||
|
|
|||
|
* Avoid some undefined behavior when popping an empty fail stack.
|
|||
|
|
|||
|
(cherry picked from commit 0b5ca7c3e551e5502f3be3b06453324fe8604e82)
|
|||
|
|
|||
|
diff --git a/include/intprops.h b/include/intprops.h
|
|||
|
index 967e32ea0cbedd56..9d10028a5966c1c6 100644
|
|||
|
--- a/include/intprops.h
|
|||
|
+++ b/include/intprops.h
|
|||
|
@@ -133,7 +133,8 @@
|
|||
|
operators might not yield numerically correct answers due to
|
|||
|
arithmetic overflow. They do not rely on undefined or
|
|||
|
implementation-defined behavior. Their implementations are simple
|
|||
|
- and straightforward, but they are a bit harder to use than the
|
|||
|
+ and straightforward, but they are harder to use and may be less
|
|||
|
+ efficient than the INT_<op>_WRAPV, INT_<op>_OK, and
|
|||
|
INT_<op>_OVERFLOW macros described below.
|
|||
|
|
|||
|
Example usage:
|
|||
|
@@ -158,6 +159,9 @@
|
|||
|
must have minimum value MIN and maximum MAX. Unsigned types should
|
|||
|
use a zero MIN of the proper type.
|
|||
|
|
|||
|
+ Because all arguments are subject to integer promotions, these
|
|||
|
+ macros typically do not work on types narrower than 'int'.
|
|||
|
+
|
|||
|
These macros are tuned for constant MIN and MAX. For commutative
|
|||
|
operations such as A + B, they are also tuned for constant B. */
|
|||
|
|
|||
|
@@ -339,9 +343,15 @@
|
|||
|
arguments should not have side effects.
|
|||
|
|
|||
|
The WRAPV macros are not constant expressions. They support only
|
|||
|
- +, binary -, and *. Because the WRAPV macros convert the result,
|
|||
|
- they report overflow in different circumstances than the OVERFLOW
|
|||
|
- macros do.
|
|||
|
+ +, binary -, and *.
|
|||
|
+
|
|||
|
+ Because the WRAPV macros convert the result, they report overflow
|
|||
|
+ in different circumstances than the OVERFLOW macros do. For
|
|||
|
+ example, in the typical case with 16-bit 'short' and 32-bit 'int',
|
|||
|
+ if A, B and R are all of type 'short' then INT_ADD_OVERFLOW (A, B)
|
|||
|
+ returns false because the addition cannot overflow after A and B
|
|||
|
+ are converted to 'int', whereas INT_ADD_WRAPV (A, B, &R) returns
|
|||
|
+ true or false depending on whether the sum fits into 'short'.
|
|||
|
|
|||
|
These macros are tuned for their last input argument being a constant.
|
|||
|
|
|||
|
diff --git a/include/regex.h b/include/regex.h
|
|||
|
index 24eca2c297bb6043..34fb67d85536bcb9 100644
|
|||
|
--- a/include/regex.h
|
|||
|
+++ b/include/regex.h
|
|||
|
@@ -37,7 +37,8 @@ extern int __regcomp (regex_t *__preg, const char *__pattern, int __cflags);
|
|||
|
libc_hidden_proto (__regcomp)
|
|||
|
|
|||
|
extern int __regexec (const regex_t *__preg, const char *__string,
|
|||
|
- size_t __nmatch, regmatch_t __pmatch[], int __eflags);
|
|||
|
+ size_t __nmatch, regmatch_t __pmatch[__nmatch],
|
|||
|
+ int __eflags);
|
|||
|
libc_hidden_proto (__regexec)
|
|||
|
|
|||
|
extern size_t __regerror (int __errcode, const regex_t *__preg,
|
|||
|
diff --git a/misc/sys/cdefs.h b/misc/sys/cdefs.h
|
|||
|
index e0ecd9147ee3ce48..b166f3d209fe361f 100644
|
|||
|
--- a/misc/sys/cdefs.h
|
|||
|
+++ b/misc/sys/cdefs.h
|
|||
|
@@ -366,16 +366,18 @@
|
|||
|
#endif
|
|||
|
|
|||
|
/* The nonnull function attribute marks pointer parameters that
|
|||
|
- must not be NULL. */
|
|||
|
-#ifndef __nonnull
|
|||
|
+ must not be NULL. This has the name __nonnull in glibc,
|
|||
|
+ and __attribute_nonnull__ in files shared with Gnulib to avoid
|
|||
|
+ collision with a different __nonnull in DragonFlyBSD 5.9. */
|
|||
|
+#ifndef __attribute_nonnull__
|
|||
|
# if __GNUC_PREREQ (3,3) || __glibc_has_attribute (__nonnull__)
|
|||
|
-# define __nonnull(params) __attribute__ ((__nonnull__ params))
|
|||
|
+# define __attribute_nonnull__(params) __attribute__ ((__nonnull__ params))
|
|||
|
# else
|
|||
|
-# define __nonnull(params)
|
|||
|
+# define __attribute_nonnull__(params)
|
|||
|
# endif
|
|||
|
-#elif !defined __GLIBC__
|
|||
|
-# undef __nonnull
|
|||
|
-# define __nonnull(params) _GL_ATTRIBUTE_NONNULL (params)
|
|||
|
+#endif
|
|||
|
+#ifndef __nonnull
|
|||
|
+# define __nonnull(params) __attribute_nonnull__ (params)
|
|||
|
#endif
|
|||
|
|
|||
|
/* The returns_nonnull function attribute marks the return type of the function
|
|||
|
@@ -541,9 +543,9 @@
|
|||
|
[!!sizeof (struct { int __error_if_negative: (expr) ? 2 : -1; })]
|
|||
|
#endif
|
|||
|
|
|||
|
-/* The #ifndef lets Gnulib avoid including these on non-glibc
|
|||
|
- platforms, where the includes typically do not exist. */
|
|||
|
-#ifdef __GLIBC__
|
|||
|
+/* Gnulib avoids including these, as they don't work on non-glibc or
|
|||
|
+ older glibc platforms. */
|
|||
|
+#ifndef __GNULIB_CDEFS
|
|||
|
# include <bits/wordsize.h>
|
|||
|
# include <bits/long-double.h>
|
|||
|
#endif
|
|||
|
diff --git a/posix/regcomp.c b/posix/regcomp.c
|
|||
|
index d93698ae78447b46..887e5b50684e22f5 100644
|
|||
|
--- a/posix/regcomp.c
|
|||
|
+++ b/posix/regcomp.c
|
|||
|
@@ -1695,12 +1695,14 @@ calc_eclosure_iter (re_node_set *new_set, re_dfa_t *dfa, Idx node, bool root)
|
|||
|
reg_errcode_t err;
|
|||
|
Idx i;
|
|||
|
re_node_set eclosure;
|
|||
|
- bool ok;
|
|||
|
bool incomplete = false;
|
|||
|
err = re_node_set_alloc (&eclosure, dfa->edests[node].nelem + 1);
|
|||
|
if (__glibc_unlikely (err != REG_NOERROR))
|
|||
|
return err;
|
|||
|
|
|||
|
+ /* An epsilon closure includes itself. */
|
|||
|
+ eclosure.elems[eclosure.nelem++] = node;
|
|||
|
+
|
|||
|
/* This indicates that we are calculating this node now.
|
|||
|
We reference this value to avoid infinite loop. */
|
|||
|
dfa->eclosures[node].nelem = -1;
|
|||
|
@@ -1753,10 +1755,6 @@ calc_eclosure_iter (re_node_set *new_set, re_dfa_t *dfa, Idx node, bool root)
|
|||
|
}
|
|||
|
}
|
|||
|
|
|||
|
- /* An epsilon closure includes itself. */
|
|||
|
- ok = re_node_set_insert (&eclosure, node);
|
|||
|
- if (__glibc_unlikely (! ok))
|
|||
|
- return REG_ESPACE;
|
|||
|
if (incomplete && !root)
|
|||
|
dfa->eclosures[node].nelem = 0;
|
|||
|
else
|
|||
|
diff --git a/posix/regex.c b/posix/regex.c
|
|||
|
index 7296be0f08da88d8..d32863972c7bcdcf 100644
|
|||
|
--- a/posix/regex.c
|
|||
|
+++ b/posix/regex.c
|
|||
|
@@ -24,6 +24,7 @@
|
|||
|
|
|||
|
# if __GNUC_PREREQ (4, 6)
|
|||
|
# pragma GCC diagnostic ignored "-Wsuggest-attribute=pure"
|
|||
|
+# pragma GCC diagnostic ignored "-Wvla"
|
|||
|
# endif
|
|||
|
# if __GNUC_PREREQ (4, 3)
|
|||
|
# pragma GCC diagnostic ignored "-Wold-style-definition"
|
|||
|
diff --git a/posix/regex.h b/posix/regex.h
|
|||
|
index 14fb1d8364a11d29..adb69768ee520554 100644
|
|||
|
--- a/posix/regex.h
|
|||
|
+++ b/posix/regex.h
|
|||
|
@@ -522,6 +522,30 @@ typedef struct
|
|||
|
|
|||
|
/* Declarations for routines. */
|
|||
|
|
|||
|
+#ifndef _REGEX_NELTS
|
|||
|
+# if (defined __STDC_VERSION__ && 199901L <= __STDC_VERSION__ \
|
|||
|
+ && !defined __STDC_NO_VLA__)
|
|||
|
+# define _REGEX_NELTS(n) n
|
|||
|
+# else
|
|||
|
+# define _REGEX_NELTS(n)
|
|||
|
+# endif
|
|||
|
+#endif
|
|||
|
+
|
|||
|
+#if defined __GNUC__ && 4 < __GNUC__ + (6 <= __GNUC_MINOR__)
|
|||
|
+# pragma GCC diagnostic push
|
|||
|
+# pragma GCC diagnostic ignored "-Wvla"
|
|||
|
+#endif
|
|||
|
+
|
|||
|
+#ifndef _Attr_access_
|
|||
|
+# ifdef __attr_access
|
|||
|
+# define _Attr_access_(arg) __attr_access (arg)
|
|||
|
+# elif defined __GNUC__ && 10 <= __GNUC__
|
|||
|
+# define _Attr_access_(x) __attribute__ ((__access__ x))
|
|||
|
+# else
|
|||
|
+# define _Attr_access_(x)
|
|||
|
+# endif
|
|||
|
+#endif
|
|||
|
+
|
|||
|
#ifdef __USE_GNU
|
|||
|
/* Sets the current default syntax to SYNTAX, and return the old syntax.
|
|||
|
You can also simply assign to the 're_syntax_options' variable. */
|
|||
|
@@ -537,7 +561,7 @@ extern reg_syntax_t re_set_syntax (reg_syntax_t __syntax);
|
|||
|
'regfree'. */
|
|||
|
extern const char *re_compile_pattern (const char *__pattern, size_t __length,
|
|||
|
struct re_pattern_buffer *__buffer)
|
|||
|
- __attr_access ((__read_only__, 1, 2));
|
|||
|
+ _Attr_access_ ((__read_only__, 1, 2));
|
|||
|
|
|||
|
|
|||
|
/* Compile a fastmap for the compiled pattern in BUFFER; used to
|
|||
|
@@ -555,7 +579,7 @@ extern regoff_t re_search (struct re_pattern_buffer *__buffer,
|
|||
|
const char *__String, regoff_t __length,
|
|||
|
regoff_t __start, regoff_t __range,
|
|||
|
struct re_registers *__regs)
|
|||
|
- __attr_access ((__read_only__, 2, 3));
|
|||
|
+ _Attr_access_ ((__read_only__, 2, 3));
|
|||
|
|
|||
|
|
|||
|
/* Like 're_search', but search in the concatenation of STRING1 and
|
|||
|
@@ -566,8 +590,8 @@ extern regoff_t re_search_2 (struct re_pattern_buffer *__buffer,
|
|||
|
regoff_t __start, regoff_t __range,
|
|||
|
struct re_registers *__regs,
|
|||
|
regoff_t __stop)
|
|||
|
- __attr_access ((__read_only__, 2, 3))
|
|||
|
- __attr_access ((__read_only__, 4, 5));
|
|||
|
+ _Attr_access_ ((__read_only__, 2, 3))
|
|||
|
+ _Attr_access_ ((__read_only__, 4, 5));
|
|||
|
|
|||
|
|
|||
|
/* Like 're_search', but return how many characters in STRING the regexp
|
|||
|
@@ -575,7 +599,7 @@ extern regoff_t re_search_2 (struct re_pattern_buffer *__buffer,
|
|||
|
extern regoff_t re_match (struct re_pattern_buffer *__buffer,
|
|||
|
const char *__String, regoff_t __length,
|
|||
|
regoff_t __start, struct re_registers *__regs)
|
|||
|
- __attr_access ((__read_only__, 2, 3));
|
|||
|
+ _Attr_access_ ((__read_only__, 2, 3));
|
|||
|
|
|||
|
|
|||
|
/* Relates to 're_match' as 're_search_2' relates to 're_search'. */
|
|||
|
@@ -584,8 +608,8 @@ extern regoff_t re_match_2 (struct re_pattern_buffer *__buffer,
|
|||
|
const char *__string2, regoff_t __length2,
|
|||
|
regoff_t __start, struct re_registers *__regs,
|
|||
|
regoff_t __stop)
|
|||
|
- __attr_access ((__read_only__, 2, 3))
|
|||
|
- __attr_access ((__read_only__, 4, 5));
|
|||
|
+ _Attr_access_ ((__read_only__, 2, 3))
|
|||
|
+ _Attr_access_ ((__read_only__, 4, 5));
|
|||
|
|
|||
|
|
|||
|
/* Set REGS to hold NUM_REGS registers, storing them in STARTS and
|
|||
|
@@ -654,16 +678,19 @@ extern int regcomp (regex_t *_Restrict_ __preg,
|
|||
|
|
|||
|
extern int regexec (const regex_t *_Restrict_ __preg,
|
|||
|
const char *_Restrict_ __String, size_t __nmatch,
|
|||
|
- regmatch_t __pmatch[_Restrict_arr_],
|
|||
|
- int __eflags)
|
|||
|
- __attr_access ((__write_only__, 4, 3));
|
|||
|
+ regmatch_t __pmatch[_Restrict_arr_
|
|||
|
+ _REGEX_NELTS (__nmatch)],
|
|||
|
+ int __eflags);
|
|||
|
|
|||
|
extern size_t regerror (int __errcode, const regex_t *_Restrict_ __preg,
|
|||
|
char *_Restrict_ __errbuf, size_t __errbuf_size)
|
|||
|
- __attr_access ((__write_only__, 3, 4));
|
|||
|
+ _Attr_access_ ((__write_only__, 3, 4));
|
|||
|
|
|||
|
extern void regfree (regex_t *__preg);
|
|||
|
|
|||
|
+#if defined __GNUC__ && 4 < __GNUC__ + (6 <= __GNUC_MINOR__)
|
|||
|
+# pragma GCC diagnostic pop
|
|||
|
+#endif
|
|||
|
|
|||
|
#ifdef __cplusplus
|
|||
|
}
|
|||
|
diff --git a/posix/regex_internal.c b/posix/regex_internal.c
|
|||
|
index 9dd387ef85d64e62..aefcfa2f52e68c6a 100644
|
|||
|
--- a/posix/regex_internal.c
|
|||
|
+++ b/posix/regex_internal.c
|
|||
|
@@ -1211,6 +1211,10 @@ re_node_set_merge (re_node_set *dest, const re_node_set *src)
|
|||
|
|
|||
|
if (__glibc_unlikely (dest->nelem == 0))
|
|||
|
{
|
|||
|
+ /* Although we already guaranteed above that dest->alloc != 0 and
|
|||
|
+ therefore dest->elems != NULL, add a debug assertion to pacify
|
|||
|
+ GCC 11.2.1's -fanalyzer. */
|
|||
|
+ DEBUG_ASSERT (dest->elems);
|
|||
|
dest->nelem = src->nelem;
|
|||
|
memcpy (dest->elems, src->elems, src->nelem * sizeof (Idx));
|
|||
|
return REG_NOERROR;
|
|||
|
@@ -1286,7 +1290,10 @@ re_node_set_insert (re_node_set *set, Idx elem)
|
|||
|
|
|||
|
if (__glibc_unlikely (set->nelem) == 0)
|
|||
|
{
|
|||
|
- /* We already guaranteed above that set->alloc != 0. */
|
|||
|
+ /* Although we already guaranteed above that set->alloc != 0 and
|
|||
|
+ therefore set->elems != NULL, add a debug assertion to pacify
|
|||
|
+ GCC 11.2 -fanalyzer. */
|
|||
|
+ DEBUG_ASSERT (set->elems);
|
|||
|
set->elems[0] = elem;
|
|||
|
++set->nelem;
|
|||
|
return true;
|
|||
|
@@ -1314,6 +1321,7 @@ re_node_set_insert (re_node_set *set, Idx elem)
|
|||
|
{
|
|||
|
for (idx = set->nelem; set->elems[idx - 1] > elem; idx--)
|
|||
|
set->elems[idx] = set->elems[idx - 1];
|
|||
|
+ DEBUG_ASSERT (set->elems[idx - 1] < elem);
|
|||
|
}
|
|||
|
|
|||
|
/* Insert the new element. */
|
|||
|
diff --git a/posix/regex_internal.h b/posix/regex_internal.h
|
|||
|
index edcdc07e999694ac..1245e782ffc69086 100644
|
|||
|
--- a/posix/regex_internal.h
|
|||
|
+++ b/posix/regex_internal.h
|
|||
|
@@ -32,6 +32,10 @@
|
|||
|
#include <stdbool.h>
|
|||
|
#include <stdint.h>
|
|||
|
|
|||
|
+#ifndef _LIBC
|
|||
|
+# include <dynarray.h>
|
|||
|
+#endif
|
|||
|
+
|
|||
|
#include <intprops.h>
|
|||
|
#include <verify.h>
|
|||
|
|
|||
|
@@ -49,14 +53,14 @@
|
|||
|
# 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
|
|||
|
+#elif defined GNULIB_LOCK && !defined GNULIB_REGEX_SINGLE_THREAD
|
|||
|
# include "glthread/lock.h"
|
|||
|
# define lock_define(name) gl_lock_define (, name)
|
|||
|
# 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
|
|||
|
+#elif defined GNULIB_PTHREAD && !defined GNULIB_REGEX_SINGLE_THREAD
|
|||
|
# include <pthread.h>
|
|||
|
# define lock_define(name) pthread_mutex_t name;
|
|||
|
# define lock_init(lock) pthread_mutex_init (&(lock), 0)
|
|||
|
diff --git a/posix/regexec.c b/posix/regexec.c
|
|||
|
index f7b4f9cfc3f030df..83e9aaf8cad956a2 100644
|
|||
|
--- a/posix/regexec.c
|
|||
|
+++ b/posix/regexec.c
|
|||
|
@@ -59,7 +59,7 @@ static void update_regs (const re_dfa_t *dfa, regmatch_t *pmatch,
|
|||
|
Idx cur_idx, Idx nmatch);
|
|||
|
static reg_errcode_t push_fail_stack (struct re_fail_stack_t *fs,
|
|||
|
Idx str_idx, Idx dest_node, Idx nregs,
|
|||
|
- regmatch_t *regs,
|
|||
|
+ regmatch_t *regs, regmatch_t *prevregs,
|
|||
|
re_node_set *eps_via_nodes);
|
|||
|
static reg_errcode_t set_regs (const regex_t *preg,
|
|||
|
const re_match_context_t *mctx,
|
|||
|
@@ -186,11 +186,12 @@ static reg_errcode_t extend_buffers (re_match_context_t *mctx, int min_len);
|
|||
|
REG_NOTBOL is set, then ^ does not match at the beginning of the
|
|||
|
string; if REG_NOTEOL is set, then $ does not match at the end.
|
|||
|
|
|||
|
- We return 0 if we find a match and REG_NOMATCH if not. */
|
|||
|
+ Return 0 if a match is found, REG_NOMATCH if not, REG_BADPAT if
|
|||
|
+ EFLAGS is invalid. */
|
|||
|
|
|||
|
int
|
|||
|
regexec (const regex_t *__restrict preg, const char *__restrict string,
|
|||
|
- size_t nmatch, regmatch_t pmatch[], int eflags)
|
|||
|
+ size_t nmatch, regmatch_t pmatch[_REGEX_NELTS (nmatch)], int eflags)
|
|||
|
{
|
|||
|
reg_errcode_t err;
|
|||
|
Idx start, length;
|
|||
|
@@ -234,7 +235,7 @@ int
|
|||
|
attribute_compat_text_section
|
|||
|
__compat_regexec (const regex_t *__restrict preg,
|
|||
|
const char *__restrict string, size_t nmatch,
|
|||
|
- regmatch_t pmatch[], int eflags)
|
|||
|
+ regmatch_t pmatch[_REGEX_NELTS (nmatch)], int eflags)
|
|||
|
{
|
|||
|
return regexec (preg, string, nmatch, pmatch,
|
|||
|
eflags & (REG_NOTBOL | REG_NOTEOL));
|
|||
|
@@ -269,8 +270,8 @@ compat_symbol (libc, __compat_regexec, regexec, GLIBC_2_0);
|
|||
|
strings.)
|
|||
|
|
|||
|
On success, re_match* functions return the length of the match, re_search*
|
|||
|
- return the position of the start of the match. Return value -1 means no
|
|||
|
- match was found and -2 indicates an internal error. */
|
|||
|
+ return the position of the start of the match. They return -1 on
|
|||
|
+ match failure, -2 on error. */
|
|||
|
|
|||
|
regoff_t
|
|||
|
re_match (struct re_pattern_buffer *bufp, const char *string, Idx length,
|
|||
|
@@ -1206,27 +1207,30 @@ check_halt_state_context (const re_match_context_t *mctx,
|
|||
|
/* Compute the next node to which "NFA" transit from NODE("NFA" is a NFA
|
|||
|
corresponding to the DFA).
|
|||
|
Return the destination node, and update EPS_VIA_NODES;
|
|||
|
- return -1 in case of errors. */
|
|||
|
+ return -1 on match failure, -2 on error. */
|
|||
|
|
|||
|
static Idx
|
|||
|
proceed_next_node (const re_match_context_t *mctx, Idx nregs, regmatch_t *regs,
|
|||
|
+ regmatch_t *prevregs,
|
|||
|
Idx *pidx, Idx node, re_node_set *eps_via_nodes,
|
|||
|
struct re_fail_stack_t *fs)
|
|||
|
{
|
|||
|
const re_dfa_t *const dfa = mctx->dfa;
|
|||
|
- Idx i;
|
|||
|
- 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;
|
|||
|
- ok = re_node_set_insert (eps_via_nodes, node);
|
|||
|
- if (__glibc_unlikely (! ok))
|
|||
|
- return -2;
|
|||
|
- /* Pick up a valid destination, or return -1 if none
|
|||
|
- is found. */
|
|||
|
- for (dest_node = -1, i = 0; i < edests->nelem; ++i)
|
|||
|
+
|
|||
|
+ if (! re_node_set_contains (eps_via_nodes, node))
|
|||
|
+ {
|
|||
|
+ bool ok = re_node_set_insert (eps_via_nodes, node);
|
|||
|
+ if (__glibc_unlikely (! ok))
|
|||
|
+ return -2;
|
|||
|
+ }
|
|||
|
+
|
|||
|
+ /* Pick a valid destination, or return -1 if none is found. */
|
|||
|
+ Idx dest_node = -1;
|
|||
|
+ for (Idx i = 0; i < edests->nelem; i++)
|
|||
|
{
|
|||
|
Idx candidate = edests->elems[i];
|
|||
|
if (!re_node_set_contains (cur_nodes, candidate))
|
|||
|
@@ -1244,7 +1248,7 @@ proceed_next_node (const re_match_context_t *mctx, Idx nregs, regmatch_t *regs,
|
|||
|
/* Otherwise, push the second epsilon-transition on the fail stack. */
|
|||
|
else if (fs != NULL
|
|||
|
&& push_fail_stack (fs, *pidx, candidate, nregs, regs,
|
|||
|
- eps_via_nodes))
|
|||
|
+ prevregs, eps_via_nodes))
|
|||
|
return -2;
|
|||
|
|
|||
|
/* We know we are going to exit. */
|
|||
|
@@ -1288,7 +1292,7 @@ proceed_next_node (const re_match_context_t *mctx, Idx nregs, regmatch_t *regs,
|
|||
|
if (naccepted == 0)
|
|||
|
{
|
|||
|
Idx dest_node;
|
|||
|
- ok = re_node_set_insert (eps_via_nodes, node);
|
|||
|
+ bool ok = re_node_set_insert (eps_via_nodes, node);
|
|||
|
if (__glibc_unlikely (! ok))
|
|||
|
return -2;
|
|||
|
dest_node = dfa->edests[node].elems[0];
|
|||
|
@@ -1317,7 +1321,8 @@ proceed_next_node (const re_match_context_t *mctx, Idx nregs, regmatch_t *regs,
|
|||
|
static reg_errcode_t
|
|||
|
__attribute_warn_unused_result__
|
|||
|
push_fail_stack (struct re_fail_stack_t *fs, Idx str_idx, Idx dest_node,
|
|||
|
- Idx nregs, regmatch_t *regs, re_node_set *eps_via_nodes)
|
|||
|
+ Idx nregs, regmatch_t *regs, regmatch_t *prevregs,
|
|||
|
+ re_node_set *eps_via_nodes)
|
|||
|
{
|
|||
|
reg_errcode_t err;
|
|||
|
Idx num = fs->num++;
|
|||
|
@@ -1333,25 +1338,30 @@ push_fail_stack (struct re_fail_stack_t *fs, Idx str_idx, Idx dest_node,
|
|||
|
}
|
|||
|
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_malloc (regmatch_t, 2 * nregs);
|
|||
|
if (fs->stack[num].regs == NULL)
|
|||
|
return REG_ESPACE;
|
|||
|
memcpy (fs->stack[num].regs, regs, sizeof (regmatch_t) * nregs);
|
|||
|
+ memcpy (fs->stack[num].regs + nregs, prevregs, sizeof (regmatch_t) * nregs);
|
|||
|
err = re_node_set_init_copy (&fs->stack[num].eps_via_nodes, eps_via_nodes);
|
|||
|
return err;
|
|||
|
}
|
|||
|
|
|||
|
static Idx
|
|||
|
pop_fail_stack (struct re_fail_stack_t *fs, Idx *pidx, Idx nregs,
|
|||
|
- regmatch_t *regs, re_node_set *eps_via_nodes)
|
|||
|
+ regmatch_t *regs, regmatch_t *prevregs,
|
|||
|
+ re_node_set *eps_via_nodes)
|
|||
|
{
|
|||
|
+ if (fs == NULL || fs->num == 0)
|
|||
|
+ return -1;
|
|||
|
Idx num = --fs->num;
|
|||
|
- DEBUG_ASSERT (num >= 0);
|
|||
|
*pidx = fs->stack[num].idx;
|
|||
|
memcpy (regs, fs->stack[num].regs, sizeof (regmatch_t) * nregs);
|
|||
|
+ memcpy (prevregs, fs->stack[num].regs + nregs, sizeof (regmatch_t) * nregs);
|
|||
|
re_node_set_free (eps_via_nodes);
|
|||
|
re_free (fs->stack[num].regs);
|
|||
|
*eps_via_nodes = fs->stack[num].eps_via_nodes;
|
|||
|
+ DEBUG_ASSERT (0 <= fs->stack[num].node);
|
|||
|
return fs->stack[num].node;
|
|||
|
}
|
|||
|
|
|||
|
@@ -1407,33 +1417,32 @@ set_regs (const regex_t *preg, const re_match_context_t *mctx, size_t nmatch,
|
|||
|
{
|
|||
|
update_regs (dfa, pmatch, prev_idx_match, cur_node, idx, nmatch);
|
|||
|
|
|||
|
- if (idx == pmatch[0].rm_eo && cur_node == mctx->last_node)
|
|||
|
+ if ((idx == pmatch[0].rm_eo && cur_node == mctx->last_node)
|
|||
|
+ || (fs && re_node_set_contains (&eps_via_nodes, cur_node)))
|
|||
|
{
|
|||
|
Idx reg_idx;
|
|||
|
+ cur_node = -1;
|
|||
|
if (fs)
|
|||
|
{
|
|||
|
for (reg_idx = 0; reg_idx < nmatch; ++reg_idx)
|
|||
|
if (pmatch[reg_idx].rm_so > -1 && pmatch[reg_idx].rm_eo == -1)
|
|||
|
- break;
|
|||
|
- if (reg_idx == nmatch)
|
|||
|
- {
|
|||
|
- re_node_set_free (&eps_via_nodes);
|
|||
|
- regmatch_list_free (&prev_match);
|
|||
|
- return free_fail_stack_return (fs);
|
|||
|
- }
|
|||
|
- cur_node = pop_fail_stack (fs, &idx, nmatch, pmatch,
|
|||
|
- &eps_via_nodes);
|
|||
|
+ {
|
|||
|
+ cur_node = pop_fail_stack (fs, &idx, nmatch, pmatch,
|
|||
|
+ prev_idx_match, &eps_via_nodes);
|
|||
|
+ break;
|
|||
|
+ }
|
|||
|
}
|
|||
|
- else
|
|||
|
+ if (cur_node < 0)
|
|||
|
{
|
|||
|
re_node_set_free (&eps_via_nodes);
|
|||
|
regmatch_list_free (&prev_match);
|
|||
|
- return REG_NOERROR;
|
|||
|
+ return free_fail_stack_return (fs);
|
|||
|
}
|
|||
|
}
|
|||
|
|
|||
|
/* Proceed to next node. */
|
|||
|
- cur_node = proceed_next_node (mctx, nmatch, pmatch, &idx, cur_node,
|
|||
|
+ cur_node = proceed_next_node (mctx, nmatch, pmatch, prev_idx_match,
|
|||
|
+ &idx, cur_node,
|
|||
|
&eps_via_nodes, fs);
|
|||
|
|
|||
|
if (__glibc_unlikely (cur_node < 0))
|
|||
|
@@ -1445,13 +1454,13 @@ set_regs (const regex_t *preg, const re_match_context_t *mctx, size_t nmatch,
|
|||
|
free_fail_stack_return (fs);
|
|||
|
return REG_ESPACE;
|
|||
|
}
|
|||
|
- if (fs)
|
|||
|
- cur_node = pop_fail_stack (fs, &idx, nmatch, pmatch,
|
|||
|
- &eps_via_nodes);
|
|||
|
- else
|
|||
|
+ cur_node = pop_fail_stack (fs, &idx, nmatch, pmatch,
|
|||
|
+ prev_idx_match, &eps_via_nodes);
|
|||
|
+ if (cur_node < 0)
|
|||
|
{
|
|||
|
re_node_set_free (&eps_via_nodes);
|
|||
|
regmatch_list_free (&prev_match);
|
|||
|
+ free_fail_stack_return (fs);
|
|||
|
return REG_NOMATCH;
|
|||
|
}
|
|||
|
}
|
|||
|
@@ -1495,10 +1504,10 @@ update_regs (const re_dfa_t *dfa, regmatch_t *pmatch,
|
|||
|
}
|
|||
|
else if (type == OP_CLOSE_SUBEXP)
|
|||
|
{
|
|||
|
+ /* We are at the last node of this sub expression. */
|
|||
|
Idx reg_num = dfa->nodes[cur_node].opr.idx + 1;
|
|||
|
if (reg_num < nmatch)
|
|||
|
{
|
|||
|
- /* We are at the last node of this sub expression. */
|
|||
|
if (pmatch[reg_num].rm_so < cur_idx)
|
|||
|
{
|
|||
|
pmatch[reg_num].rm_eo = cur_idx;
|
|||
|
@@ -2195,6 +2204,7 @@ sift_states_iter_mb (const re_match_context_t *mctx, re_sift_context_t *sctx,
|
|||
|
|
|||
|
/* Return the next state to which the current state STATE will transit by
|
|||
|
accepting the current input byte, and update STATE_LOG if necessary.
|
|||
|
+ Return NULL on failure.
|
|||
|
If STATE can accept a multibyte char/collating element/back reference
|
|||
|
update the destination of STATE_LOG. */
|
|||
|
|
|||
|
@@ -2395,7 +2405,7 @@ check_subexp_matching_top (re_match_context_t *mctx, re_node_set *cur_nodes,
|
|||
|
|
|||
|
#if 0
|
|||
|
/* Return the next state to which the current state STATE will transit by
|
|||
|
- accepting the current input byte. */
|
|||
|
+ accepting the current input byte. Return NULL on failure. */
|
|||
|
|
|||
|
static re_dfastate_t *
|
|||
|
transit_state_sb (reg_errcode_t *err, re_match_context_t *mctx,
|
|||
|
@@ -2817,7 +2827,8 @@ find_subexp_node (const re_dfa_t *dfa, const re_node_set *nodes,
|
|||
|
/* Check whether the node TOP_NODE at TOP_STR can arrive to the node
|
|||
|
LAST_NODE at LAST_STR. We record the path onto PATH since it will be
|
|||
|
heavily reused.
|
|||
|
- Return REG_NOERROR if it can arrive, or REG_NOMATCH otherwise. */
|
|||
|
+ Return REG_NOERROR if it can arrive, REG_NOMATCH if it cannot,
|
|||
|
+ REG_ESPACE if memory is exhausted. */
|
|||
|
|
|||
|
static reg_errcode_t
|
|||
|
__attribute_warn_unused_result__
|
|||
|
@@ -3433,7 +3444,8 @@ build_trtable (const re_dfa_t *dfa, re_dfastate_t *state)
|
|||
|
/* Group all nodes belonging to STATE into several destinations.
|
|||
|
Then for all destinations, set the nodes belonging to the destination
|
|||
|
to DESTS_NODE[i] and set the characters accepted by the destination
|
|||
|
- to DEST_CH[i]. This function return the number of destinations. */
|
|||
|
+ to DEST_CH[i]. Return the number of destinations if successful,
|
|||
|
+ -1 on internal error. */
|
|||
|
|
|||
|
static Idx
|
|||
|
group_nodes_into_DFAstates (const re_dfa_t *dfa, const re_dfastate_t *state,
|
|||
|
@@ -4211,7 +4223,8 @@ match_ctx_add_subtop (re_match_context_t *mctx, Idx node, Idx str_idx)
|
|||
|
}
|
|||
|
|
|||
|
/* Register the node NODE, whose type is OP_CLOSE_SUBEXP, and which matches
|
|||
|
- at STR_IDX, whose corresponding OP_OPEN_SUBEXP is SUB_TOP. */
|
|||
|
+ at STR_IDX, whose corresponding OP_OPEN_SUBEXP is SUB_TOP.
|
|||
|
+ Return the new entry if successful, NULL if memory is exhausted. */
|
|||
|
|
|||
|
static re_sub_match_last_t *
|
|||
|
match_ctx_add_sublast (re_sub_match_top_t *subtop, Idx node, Idx str_idx)
|