user: implement mlibc as the libc, finally.

It's finally done..

Signed-off-by: kaguya <vpshinomiya@protonmail.com>
This commit is contained in:
kaguya
2026-05-02 03:31:49 -04:00
parent 2fa39ad85a
commit 9a9b91c940
2387 changed files with 152741 additions and 315 deletions
@@ -0,0 +1,321 @@
/*
* An implementation of what I call the "Sea of Stars" algorithm for
* POSIX fnmatch(). The basic idea is that we factor the pattern into
* a head component (which we match first and can reject without ever
* measuring the length of the string), an optional tail component
* (which only exists if the pattern contains at least one star), and
* an optional "sea of stars", a set of star-separated components
* between the head and tail. After the head and tail matches have
* been removed from the input string, the components in the "sea of
* stars" are matched sequentially by searching for their first
* occurrence past the end of the previous match.
*
* - Rich Felker, April 2012
*/
#include <string.h>
#include <fnmatch.h>
#include <stdlib.h>
#include <wchar.h>
#include <wctype.h>
// #include "locale_impl.h"
#define END 0
#define UNMATCHABLE -2
#define BRACKET -3
#define QUESTION -4
#define STAR -5
static int str_next(const char *str, size_t n, size_t *step)
{
if (!n) {
*step = 0;
return 0;
}
if (str[0] >= 128U) {
wchar_t wc;
int k = mbtowc(&wc, str, n);
if (k<0) {
*step = 1;
return -1;
}
*step = k;
return wc;
}
*step = 1;
return str[0];
}
static int pat_next(const char *pat, size_t m, size_t *step, int flags)
{
int esc = 0;
if (!m || !*pat) {
*step = 0;
return END;
}
*step = 1;
if (pat[0]=='\\' && pat[1] && !(flags & FNM_NOESCAPE)) {
*step = 2;
pat++;
esc = 1;
goto escaped;
}
if (pat[0]=='[') {
size_t k = 1;
if (k<m) if (pat[k] == '^' || pat[k] == '!') k++;
if (k<m) if (pat[k] == ']') k++;
for (; k<m && pat[k] && pat[k]!=']'; k++) {
if (k+1<m && pat[k+1] && pat[k]=='[' && (pat[k+1]==':' || pat[k+1]=='.' || pat[k+1]=='=')) {
int z = pat[k+1];
k+=2;
if (k<m && pat[k]) k++;
while (k<m && pat[k] && (pat[k-1]!=z || pat[k]!=']')) k++;
if (k==m || !pat[k]) break;
}
}
if (k==m || !pat[k]) {
*step = 1;
return '[';
}
*step = k+1;
return BRACKET;
}
if (pat[0] == '*')
return STAR;
if (pat[0] == '?')
return QUESTION;
escaped:
if (pat[0] >= 128U) {
wchar_t wc;
int k = mbtowc(&wc, pat, m);
if (k<0) {
*step = 0;
return UNMATCHABLE;
}
*step = k + esc;
return wc;
}
return pat[0];
}
static int casefold(int k)
{
int c = towupper(k);
return c == k ? towlower(k) : c;
}
static int match_bracket(const char *p, int k, int kfold)
{
wchar_t wc;
int inv = 0;
p++;
if (*p=='^' || *p=='!') {
inv = 1;
p++;
}
if (*p==']') {
if (k==']') return !inv;
p++;
} else if (*p=='-') {
if (k=='-') return !inv;
p++;
}
wc = p[-1];
for (; *p != ']'; p++) {
if (p[0]=='-' && p[1]!=']') {
wchar_t wc2;
int l = mbtowc(&wc2, p+1, 4);
if (l < 0) return 0;
if (wc <= wc2)
if ((unsigned)k-wc <= wc2-wc ||
(unsigned)kfold-wc <= wc2-wc)
return !inv;
p += l-1;
continue;
}
if (p[0]=='[' && (p[1]==':' || p[1]=='.' || p[1]=='=')) {
const char *p0 = p+2;
int z = p[1];
p+=3;
while (p[-1]!=z || p[0]!=']') p++;
if (z == ':' && p-1-p0 < 16) {
char buf[16];
memcpy(buf, p0, p-1-p0);
buf[p-1-p0] = 0;
if (iswctype(k, wctype(buf)) ||
iswctype(kfold, wctype(buf)))
return !inv;
}
continue;
}
if (*p < 128U) {
wc = (unsigned char)*p;
} else {
int l = mbtowc(&wc, p, 4);
if (l < 0) return 0;
p += l-1;
}
if (wc==(wchar_t)k || wc==(wchar_t)kfold) return !inv;
}
return inv;
}
static int fnmatch_internal(const char *pat, size_t m, const char *str, size_t n, int flags)
{
const char *p, *ptail, *endpat;
const char *s, *stail, *endstr;
size_t pinc, sinc, tailcnt=0;
int c, k, kfold;
if (flags & FNM_PERIOD) {
if (*str == '.' && *pat != '.')
return FNM_NOMATCH;
}
for (;;) {
switch ((c = pat_next(pat, m, &pinc, flags))) {
case UNMATCHABLE:
return FNM_NOMATCH;
case STAR:
pat++;
m--;
break;
default:
k = str_next(str, n, &sinc);
if (k <= 0)
return (c==END) ? 0 : FNM_NOMATCH;
str += sinc;
n -= sinc;
kfold = flags & FNM_CASEFOLD ? casefold(k) : k;
if (c == BRACKET) {
if (!match_bracket(pat, k, kfold))
return FNM_NOMATCH;
} else if (c != QUESTION && k != c && kfold != c) {
return FNM_NOMATCH;
}
pat+=pinc;
m-=pinc;
continue;
}
break;
}
/* Compute real pat length if it was initially unknown/-1 */
m = strnlen(pat, m);
endpat = pat + m;
/* Find the last * in pat and count chars needed after it */
for (p=ptail=pat; p<endpat; p+=pinc) {
switch (pat_next(p, endpat-p, &pinc, flags)) {
case UNMATCHABLE:
return FNM_NOMATCH;
case STAR:
tailcnt=0;
ptail = p+1;
break;
default:
tailcnt++;
break;
}
}
/* Past this point we need not check for UNMATCHABLE in pat,
* because all of pat has already been parsed once. */
/* Compute real str length if it was initially unknown/-1 */
n = strnlen(str, n);
endstr = str + n;
if (n < tailcnt) return FNM_NOMATCH;
/* Find the final tailcnt chars of str, accounting for UTF-8.
* On illegal sequences we may get it wrong, but in that case
* we necessarily have a matching failure anyway. */
for (s=endstr; s>str && tailcnt; tailcnt--) {
if (s[-1] < 128U || MB_CUR_MAX==1) s--;
else while ((unsigned char)*--s-0x80U<0x40 && s>str);
}
if (tailcnt) return FNM_NOMATCH;
stail = s;
/* Check that the pat and str tails match */
p = ptail;
for (;;) {
c = pat_next(p, endpat-p, &pinc, flags);
p += pinc;
if ((k = str_next(s, endstr-s, &sinc)) <= 0) {
if (c != END) return FNM_NOMATCH;
break;
}
s += sinc;
kfold = flags & FNM_CASEFOLD ? casefold(k) : k;
if (c == BRACKET) {
if (!match_bracket(p-pinc, k, kfold))
return FNM_NOMATCH;
} else if (c != QUESTION && k != c && kfold != c) {
return FNM_NOMATCH;
}
}
/* We're all done with the tails now, so throw them out */
endstr = stail;
endpat = ptail;
/* Match pattern components until there are none left */
while (pat<endpat) {
p = pat;
s = str;
for (;;) {
c = pat_next(p, endpat-p, &pinc, flags);
p += pinc;
/* Encountering * completes/commits a component */
if (c == STAR) {
pat = p;
str = s;
break;
}
k = str_next(s, endstr-s, &sinc);
if (!k)
return FNM_NOMATCH;
kfold = flags & FNM_CASEFOLD ? casefold(k) : k;
if (c == BRACKET) {
if (!match_bracket(p-pinc, k, kfold))
break;
} else if (c != QUESTION && k != c && kfold != c) {
break;
}
s += sinc;
}
if (c == STAR) continue;
/* If we failed, advance str, by 1 char if it's a valid
* char, or past all invalid bytes otherwise. */
k = str_next(str, endstr-str, &sinc);
if (k > 0) str += sinc;
else for (str++; str_next(str, endstr-str, &sinc)<0; str++);
}
return 0;
}
int fnmatch(const char *pat, const char *str, int flags)
{
const char *s, *p;
size_t inc;
int c;
if (flags & FNM_PATHNAME) for (;;) {
for (s=str; *s && *s!='/'; s++);
for (p=pat; (c=pat_next(p, -1, &inc, flags))!=END && c!='/'; p+=inc);
if (c!=*s && (!*s || !(flags & FNM_LEADING_DIR)))
return FNM_NOMATCH;
if (fnmatch_internal(pat, p-pat, str, s-str, flags))
return FNM_NOMATCH;
if (!c) return 0;
str = s+1;
pat = p+inc;
} else if (flags & FNM_LEADING_DIR) {
for (s=str; *s; s++) {
if (*s != '/') continue;
if (!fnmatch_internal(pat, -1, str, s-str, flags))
return 0;
}
}
return fnmatch_internal(pat, -1, str, -1, flags);
}
@@ -0,0 +1,311 @@
#define _BSD_SOURCE
#include <glob.h>
#include <fnmatch.h>
#include <sys/stat.h>
#include <dirent.h>
#include <limits.h>
#include <string.h>
#include <stdlib.h>
#include <errno.h>
#include <stddef.h>
#include <unistd.h>
#include <pwd.h>
struct match
{
struct match *next;
char name[];
};
static int append(struct match **tail, const char *name, size_t len, int mark)
{
struct match *new = malloc(sizeof(struct match) + len + 2);
if (!new) return -1;
(*tail)->next = new;
new->next = NULL;
memcpy(new->name, name, len+1);
if (mark && len && name[len-1]!='/') {
new->name[len] = '/';
new->name[len+1] = 0;
}
*tail = new;
return 0;
}
static int do_glob(char *buf, size_t pos, int type, char *pat, int flags, int (*errfunc)(const char *path, int err), struct match **tail)
{
/* If GLOB_MARK is unused, we don't care about type. */
if (!type && !(flags & GLOB_MARK)) type = DT_REG;
/* Special-case the remaining pattern being all slashes, in
* which case we can use caller-passed type if it's a dir. */
if (*pat && type!=DT_DIR) type = 0;
while (pos+1 < PATH_MAX && *pat=='/') buf[pos++] = *pat++;
/* Consume maximal [escaped-]literal prefix of pattern, copying
* and un-escaping it to the running buffer as we go. */
ptrdiff_t i=0, j=0;
int in_bracket = 0, overflow = 0;
for (; pat[i]!='*' && pat[i]!='?' && (!in_bracket || pat[i]!=']'); i++) {
if (!pat[i]) {
if (overflow) return 0;
pat += i;
pos += j;
i = j = 0;
break;
} else if (pat[i] == '[') {
in_bracket = 1;
} else if (pat[i] == '\\' && !(flags & GLOB_NOESCAPE)) {
/* Backslashes inside a bracket are (at least by
* our interpretation) non-special, so if next
* char is ']' we have a complete expression. */
if (in_bracket && pat[i+1]==']') break;
/* Unpaired final backslash never matches. */
if (!pat[i+1]) return 0;
i++;
}
if (pat[i] == '/') {
if (overflow) return 0;
in_bracket = 0;
pat += i+1;
i = -1;
pos += j+1;
j = -1;
}
/* Only store a character if it fits in the buffer, but if
* a potential bracket expression is open, the overflow
* must be remembered and handled later only if the bracket
* is unterminated (and thereby a literal), so as not to
* disallow long bracket expressions with short matches. */
if (pos+(j+1) < PATH_MAX) {
buf[pos+j++] = pat[i];
} else if (in_bracket) {
overflow = 1;
} else {
return 0;
}
/* If we consume any new components, the caller-passed type
* or dummy type from above is no longer valid. */
type = 0;
}
buf[pos] = 0;
if (!*pat) {
/* If we consumed any components above, or if GLOB_MARK is
* requested and we don't yet know if the match is a dir,
* we must confirm the file exists and/or determine its type.
*
* If marking dirs, symlink type is inconclusive; we need the
* type for the symlink target, and therefore must try stat
* first unless type is known not to be a symlink. Otherwise,
* or if that fails, use lstat for determining existence to
* avoid false negatives in the case of broken symlinks. */
struct stat st;
if ((flags & GLOB_MARK) && (!type||type==DT_LNK) && !stat(buf, &st)) {
if (S_ISDIR(st.st_mode)) type = DT_DIR;
else type = DT_REG;
}
if (!type && lstat(buf, &st)) {
if (errno!=ENOENT && (errfunc(buf, errno) || (flags & GLOB_ERR)))
return GLOB_ABORTED;
return 0;
}
if (append(tail, buf, pos, (flags & GLOB_MARK) && type==DT_DIR))
return GLOB_NOSPACE;
return 0;
}
char *p2 = strchr(pat, '/'), saved_sep = '/';
/* Check if the '/' was escaped and, if so, remove the escape char
* so that it will not be unpaired when passed to fnmatch. */
if (p2 && !(flags & GLOB_NOESCAPE)) {
char *p;
for (p=p2; p>pat && p[-1]=='\\'; p--);
if ((p2-p)%2) {
p2--;
saved_sep = '\\';
}
}
DIR *dir = opendir(pos ? buf : ".");
if (!dir) {
if (errfunc(buf, errno) || (flags & GLOB_ERR))
return GLOB_ABORTED;
return 0;
}
int old_errno = errno;
struct dirent *de;
while (errno=0, de=readdir(dir)) {
/* Quickly skip non-directories when there's pattern left. */
if (p2 && de->d_type && de->d_type!=DT_DIR && de->d_type!=DT_LNK)
continue;
size_t l = strlen(de->d_name);
if (l >= PATH_MAX-pos) continue;
if (p2) *p2 = 0;
int fnm_flags= ((flags & GLOB_NOESCAPE) ? FNM_NOESCAPE : 0)
| ((!(flags & GLOB_PERIOD)) ? FNM_PERIOD : 0);
if (fnmatch(pat, de->d_name, fnm_flags))
continue;
/* With GLOB_PERIOD, don't allow matching . or .. unless
* fnmatch would match them with FNM_PERIOD rules in effect. */
if (p2 && (flags & GLOB_PERIOD) && de->d_name[0]=='.'
&& (!de->d_name[1] || de->d_name[1]=='.' && !de->d_name[2])
&& fnmatch(pat, de->d_name, fnm_flags | FNM_PERIOD))
continue;
memcpy(buf+pos, de->d_name, l+1);
if (p2) *p2 = saved_sep;
int r = do_glob(buf, pos+l, de->d_type, p2 ? p2 : "", flags, errfunc, tail);
if (r) {
closedir(dir);
return r;
}
}
int readerr = errno;
if (p2) *p2 = saved_sep;
closedir(dir);
if (readerr && (errfunc(buf, errno) || (flags & GLOB_ERR)))
return GLOB_ABORTED;
errno = old_errno;
return 0;
}
static int ignore_err(const char *path, int err)
{
return 0;
}
static void freelist(struct match *head)
{
struct match *match, *next;
for (match=head->next; match; match=next) {
next = match->next;
free(match);
}
}
static int sort(const void *a, const void *b)
{
return strcmp(*(const char **)a, *(const char **)b);
}
static int expand_tilde(char **pat, char *buf, size_t *pos)
{
char *p = *pat + 1;
size_t i = 0;
char delim, *name_end = strchrnul(p, '/');
if ((delim = *name_end)) *name_end++ = 0;
*pat = name_end;
char *home = *p ? NULL : getenv("HOME");
if (!home) {
struct passwd pw, *res;
switch (*p ? getpwnam_r(p, &pw, buf, PATH_MAX, &res)
: getpwuid_r(getuid(), &pw, buf, PATH_MAX, &res)) {
case ENOMEM:
return GLOB_NOSPACE;
case 0:
if (!res)
default:
return GLOB_NOMATCH;
}
home = pw.pw_dir;
}
while (i < PATH_MAX - 2 && *home)
buf[i++] = *home++;
if (*home)
return GLOB_NOMATCH;
if ((buf[i] = delim))
buf[++i] = 0;
*pos = i;
return 0;
}
int glob(const char *restrict pat, int flags, int (*errfunc)(const char *path, int err), glob_t *restrict g)
{
struct match head = { .next = NULL }, *tail = &head;
size_t cnt, i;
size_t offs = (flags & GLOB_DOOFFS) ? g->gl_offs : 0;
int error = 0;
char buf[PATH_MAX];
if (!errfunc) errfunc = ignore_err;
if (!(flags & GLOB_APPEND)) {
g->gl_offs = offs;
g->gl_pathc = 0;
g->gl_pathv = NULL;
}
if (*pat) {
char *p = strdup(pat);
if (!p) return GLOB_NOSPACE;
buf[0] = 0;
size_t pos = 0;
char *s = p;
if ((flags & (GLOB_TILDE | GLOB_TILDE_CHECK)) && *p == '~')
error = expand_tilde(&s, buf, &pos);
if (!error)
error = do_glob(buf, pos, 0, s, flags, errfunc, &tail);
free(p);
}
if (error == GLOB_NOSPACE) {
freelist(&head);
return error;
}
for (cnt=0, tail=head.next; tail; tail=tail->next, cnt++);
if (!cnt) {
if (flags & GLOB_NOCHECK) {
tail = &head;
if (append(&tail, pat, strlen(pat), 0))
return GLOB_NOSPACE;
cnt++;
} else
return GLOB_NOMATCH;
}
if (flags & GLOB_APPEND) {
char **pathv = realloc(g->gl_pathv, (offs + g->gl_pathc + cnt + 1) * sizeof(char *));
if (!pathv) {
freelist(&head);
return GLOB_NOSPACE;
}
g->gl_pathv = pathv;
offs += g->gl_pathc;
} else {
g->gl_pathv = malloc((offs + cnt + 1) * sizeof(char *));
if (!g->gl_pathv) {
freelist(&head);
return GLOB_NOSPACE;
}
for (i=0; i<offs; i++)
g->gl_pathv[i] = NULL;
}
for (i=0, tail=head.next; i<cnt; tail=tail->next, i++)
g->gl_pathv[offs + i] = tail->name;
g->gl_pathv[offs + i] = NULL;
g->gl_pathc += cnt;
if (!(flags & GLOB_NOSORT))
qsort(g->gl_pathv+offs, cnt, sizeof(char *), sort);
return error;
}
void globfree(glob_t *g)
{
size_t i;
for (i=0; i<g->gl_pathc; i++)
free(g->gl_pathv[g->gl_offs + i] - offsetof(struct match, name));
free(g->gl_pathv);
g->gl_pathc = 0;
g->gl_pathv = NULL;
}
// weak_alias(glob, glob64);
// weak_alias(globfree, globfree64);
File diff suppressed because it is too large Load Diff
@@ -0,0 +1,37 @@
#include <string.h>
#include <regex.h>
#include <stdio.h>
// #include "locale_impl.h"
/* Error message strings for error codes listed in `regex.h'. This list
needs to be in sync with the codes listed there, naturally. */
/* Converted to single string by Rich Felker to remove the need for
* data relocations at runtime, 27 Feb 2006. */
static const char messages[] = {
"No error\0"
"No match\0"
"Invalid regexp\0"
"Unknown collating element\0"
"Unknown character class name\0"
"Trailing backslash\0"
"Invalid back reference\0"
"Missing ']'\0"
"Missing ')'\0"
"Missing '}'\0"
"Invalid contents of {}\0"
"Invalid character range\0"
"Out of memory\0"
"Repetition not preceded by valid expression\0"
"\0Unknown error"
};
size_t regerror(int e, const regex_t *restrict preg, char *restrict buf, size_t size)
{
const char *s;
for (s=messages; e && *s; e--, s+=strlen(s)+1);
if (!*s) s++;
// s = LCTRANS_CUR(s);
return 1+snprintf(buf, size, "%s", s);
}
File diff suppressed because it is too large Load Diff
@@ -0,0 +1,158 @@
/*
tre-mem.c - TRE memory allocator
Copyright (c) 2001-2009 Ville Laurikari <vl@iki.fi>
All rights reserved.
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions
are met:
1. Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
2. Redistributions in binary form must reproduce the above copyright
notice, this list of conditions and the following disclaimer in the
documentation and/or other materials provided with the distribution.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDER AND CONTRIBUTORS
``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/
/*
This memory allocator is for allocating small memory blocks efficiently
in terms of memory overhead and execution speed. The allocated blocks
cannot be freed individually, only all at once. There can be multiple
allocators, though.
*/
#include <stdlib.h>
#include <string.h>
#include "tre.h"
/*
This memory allocator is for allocating small memory blocks efficiently
in terms of memory overhead and execution speed. The allocated blocks
cannot be freed individually, only all at once. There can be multiple
allocators, though.
*/
/* Returns a new memory allocator or NULL if out of memory. */
tre_mem_t
tre_mem_new_impl(int provided, void *provided_block)
{
tre_mem_t mem;
if (provided)
{
mem = provided_block;
memset(mem, 0, sizeof(*mem));
}
else
mem = xcalloc(1, sizeof(*mem));
if (mem == NULL)
return NULL;
return mem;
}
/* Frees the memory allocator and all memory allocated with it. */
void
tre_mem_destroy(tre_mem_t mem)
{
tre_list_t *tmp, *l = mem->blocks;
while (l != NULL)
{
xfree(l->data);
tmp = l->next;
xfree(l);
l = tmp;
}
xfree(mem);
}
/* Allocates a block of `size' bytes from `mem'. Returns a pointer to the
allocated block or NULL if an underlying malloc() failed. */
void *
tre_mem_alloc_impl(tre_mem_t mem, int provided, void *provided_block,
int zero, size_t size)
{
void *ptr;
if (mem->failed)
{
return NULL;
}
if (mem->n < size)
{
/* We need more memory than is available in the current block.
Allocate a new block. */
tre_list_t *l;
if (provided)
{
if (provided_block == NULL)
{
mem->failed = 1;
return NULL;
}
mem->ptr = provided_block;
mem->n = TRE_MEM_BLOCK_SIZE;
}
else
{
int block_size;
if (size * 8 > TRE_MEM_BLOCK_SIZE)
block_size = size * 8;
else
block_size = TRE_MEM_BLOCK_SIZE;
l = xmalloc(sizeof(*l));
if (l == NULL)
{
mem->failed = 1;
return NULL;
}
l->data = xmalloc(block_size);
if (l->data == NULL)
{
xfree(l);
mem->failed = 1;
return NULL;
}
l->next = NULL;
if (mem->current != NULL)
mem->current->next = l;
if (mem->blocks == NULL)
mem->blocks = l;
mem->current = l;
mem->ptr = l->data;
mem->n = block_size;
}
}
/* Make sure the next pointer will be aligned. */
size += ALIGN(mem->ptr + size, long);
/* Allocate from current block. */
ptr = mem->ptr;
mem->ptr += size;
mem->n -= size;
/* Set to zero if needed. */
if (zero)
memset(ptr, 0, size);
return ptr;
}
@@ -0,0 +1,241 @@
/* Taken from musl tre.h */
/*
tre-internal.h - TRE internal definitions
Copyright (c) 2001-2009 Ville Laurikari <vl@iki.fi>
All rights reserved.
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions
are met:
1. Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
2. Redistributions in binary form must reproduce the above copyright
notice, this list of conditions and the following disclaimer in the
documentation and/or other materials provided with the distribution.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDER AND CONTRIBUTORS
``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/
#include <regex.h>
#include <wchar.h>
#include <wctype.h>
#define hidden __attribute__((__visibility__("hidden")))
/* TODO: These should probably go in limits.h */
#define CHARCLASS_NAME_MAX 14
#define RE_DUP_MAX 255
#undef TRE_MBSTATE
#define NDEBUG
#define TRE_REGEX_T_FIELD __opaque
typedef int reg_errcode_t;
typedef wchar_t tre_char_t;
#define DPRINT(msg) do { } while(0)
#define elementsof(x) ( sizeof(x) / sizeof(x[0]) )
#define tre_mbrtowc(pwc, s, n, ps) (mbtowc((pwc), (s), (n)))
/* Wide characters. */
typedef wint_t tre_cint_t;
#define TRE_CHAR_MAX 0x10ffff
#define tre_isalnum iswalnum
#define tre_isalpha iswalpha
#define tre_isblank iswblank
#define tre_iscntrl iswcntrl
#define tre_isdigit iswdigit
#define tre_isgraph iswgraph
#define tre_islower iswlower
#define tre_isprint iswprint
#define tre_ispunct iswpunct
#define tre_isspace iswspace
#define tre_isupper iswupper
#define tre_isxdigit iswxdigit
#define tre_tolower towlower
#define tre_toupper towupper
#define tre_strlen wcslen
/* Use system provided iswctype() and wctype(). */
typedef wctype_t tre_ctype_t;
#define tre_isctype iswctype
#define tre_ctype wctype
/* Returns number of bytes to add to (char *)ptr to make it
properly aligned for the type. */
#define ALIGN(ptr, type) \
((((long)ptr) % sizeof(type)) \
? (sizeof(type) - (((long)ptr) % sizeof(type))) \
: 0)
#undef MAX
#undef MIN
#define MAX(a, b) (((a) >= (b)) ? (a) : (b))
#define MIN(a, b) (((a) <= (b)) ? (a) : (b))
/* TNFA transition type. A TNFA state is an array of transitions,
the terminator is a transition with NULL `state'. */
typedef struct tnfa_transition tre_tnfa_transition_t;
struct tnfa_transition {
/* Range of accepted characters. */
tre_cint_t code_min;
tre_cint_t code_max;
/* Pointer to the destination state. */
tre_tnfa_transition_t *state;
/* ID number of the destination state. */
int state_id;
/* -1 terminated array of tags (or NULL). */
int *tags;
/* Assertion bitmap. */
int assertions;
/* Assertion parameters. */
union {
/* Character class assertion. */
tre_ctype_t class;
/* Back reference assertion. */
int backref;
} u;
/* Negative character class assertions. */
tre_ctype_t *neg_classes;
};
/* Assertions. */
#define ASSERT_AT_BOL 1 /* Beginning of line. */
#define ASSERT_AT_EOL 2 /* End of line. */
#define ASSERT_CHAR_CLASS 4 /* Character class in `class'. */
#define ASSERT_CHAR_CLASS_NEG 8 /* Character classes in `neg_classes'. */
#define ASSERT_AT_BOW 16 /* Beginning of word. */
#define ASSERT_AT_EOW 32 /* End of word. */
#define ASSERT_AT_WB 64 /* Word boundary. */
#define ASSERT_AT_WB_NEG 128 /* Not a word boundary. */
#define ASSERT_BACKREF 256 /* A back reference in `backref'. */
#define ASSERT_LAST 256
/* Tag directions. */
typedef enum {
TRE_TAG_MINIMIZE = 0,
TRE_TAG_MAXIMIZE = 1
} tre_tag_direction_t;
/* Instructions to compute submatch register values from tag values
after a successful match. */
struct tre_submatch_data {
/* Tag that gives the value for rm_so (submatch start offset). */
int so_tag;
/* Tag that gives the value for rm_eo (submatch end offset). */
int eo_tag;
/* List of submatches this submatch is contained in. */
int *parents;
};
typedef struct tre_submatch_data tre_submatch_data_t;
/* TNFA definition. */
typedef struct tnfa tre_tnfa_t;
struct tnfa {
tre_tnfa_transition_t *transitions;
unsigned int num_transitions;
tre_tnfa_transition_t *initial;
tre_tnfa_transition_t *final;
tre_submatch_data_t *submatch_data;
char *firstpos_chars;
int first_char;
unsigned int num_submatches;
tre_tag_direction_t *tag_directions;
int *minimal_tags;
int num_tags;
int num_minimals;
int end_tag;
int num_states;
int cflags;
int have_backrefs;
int have_approx;
};
/* from tre-mem.h: */
#define TRE_MEM_BLOCK_SIZE 1024
typedef struct tre_list {
void *data;
struct tre_list *next;
} tre_list_t;
typedef struct tre_mem_struct {
tre_list_t *blocks;
tre_list_t *current;
char *ptr;
size_t n;
int failed;
void **provided;
} *tre_mem_t;
#ifndef __MLIBC_ABI_ONLY
#define tre_mem_new_impl __tre_mem_new_impl
#define tre_mem_alloc_impl __tre_mem_alloc_impl
#define tre_mem_destroy __tre_mem_destroy
hidden tre_mem_t tre_mem_new_impl(int provided, void *provided_block);
hidden void *tre_mem_alloc_impl(tre_mem_t mem, int provided, void *provided_block,
int zero, size_t size);
/* Returns a new memory allocator or NULL if out of memory. */
#define tre_mem_new() tre_mem_new_impl(0, NULL)
/* Allocates a block of `size' bytes from `mem'. Returns a pointer to the
allocated block or NULL if an underlying malloc() failed. */
#define tre_mem_alloc(mem, size) tre_mem_alloc_impl(mem, 0, NULL, 0, size)
/* Allocates a block of `size' bytes from `mem'. Returns a pointer to the
allocated block or NULL if an underlying malloc() failed. The memory
is set to zero. */
#define tre_mem_calloc(mem, size) tre_mem_alloc_impl(mem, 0, NULL, 1, size)
#ifdef TRE_USE_ALLOCA
/* alloca() versions. Like above, but memory is allocated with alloca()
instead of malloc(). */
#define tre_mem_newa() \
tre_mem_new_impl(1, alloca(sizeof(struct tre_mem_struct)))
#define tre_mem_alloca(mem, size) \
((mem)->n >= (size) \
? tre_mem_alloc_impl((mem), 1, NULL, 0, (size)) \
: tre_mem_alloc_impl((mem), 1, alloca(TRE_MEM_BLOCK_SIZE), 0, (size)))
#endif /* TRE_USE_ALLOCA */
/* Frees the memory allocator and all memory allocated with it. */
hidden void tre_mem_destroy(tre_mem_t mem);
#define xmalloc malloc
#define xcalloc calloc
#define xfree free
#define xrealloc realloc
#endif /* !__MLIBC_ABI_ONLY */