mirrors/php-src
0
0
Fork 0
mirror of https://github.com/php/php-src.git synced 2024-11-30 21:35:36 +08:00
Code Issues Packages Projects Releases Wiki Activity

fileinfo: Port libmagic 5.40

Browse Source 
Signed-off-by: Anatol Belski <ab@php.net>
This commit is contained in:
Anatol Belski 2021-04-01 20:15:45 +02:00
parent 22019a1edd
commit 3b9173dc8f
21 changed files with 174944 additions and 148884 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

315926
ext/fileinfo/data_file.c
View File

File diff suppressed because it is too large Load Diff

820
ext/fileinfo/libmagic.patch
View File

File diff suppressed because it is too large Load Diff

22
ext/fileinfo/libmagic/apprentice.c
View File

@ -34,11 +34,10 @@
#include "file.h" #include "file.h"
#ifndef lint #ifndef lint
FILE_RCSID("@(#)$File: apprentice.c,v 1.297 2020/05/09 18:57:15 christos Exp $") FILE_RCSID("@(#)$File: apprentice.c,v 1.301 2021/02/23 00:51:11 christos Exp $")
#endif /* lint */ #endif /* lint */
#include "magic.h" #include "magic.h"
#include "patchlevel.h"
#include <stdlib.h> #include <stdlib.h>
#if defined(__hpux) && !defined(HAVE_STRTOULL) #if defined(__hpux) && !defined(HAVE_STRTOULL)
@ -536,6 +535,7 @@ file_ms_alloc(int flags)
ms->elf_notes_max = FILE_ELF_NOTES_MAX; ms->elf_notes_max = FILE_ELF_NOTES_MAX;
ms->regex_max = FILE_REGEX_MAX; ms->regex_max = FILE_REGEX_MAX;
ms->bytes_max = FILE_BYTES_MAX; ms->bytes_max = FILE_BYTES_MAX;
ms->encoding_max = FILE_ENCODING_MAX;
return ms; return ms;
free: free:
efree(ms); efree(ms);
@ -1416,7 +1416,10 @@ apprentice_load(struct magic_set *ms, const char *fn, int action)
*/ */
set_last_default(ms, mset[j].me, mset[j].count); set_last_default(ms, mset[j].me, mset[j].count);
/* coalesce per file arrays into a single one */ /* coalesce per file arrays into a single one, if needed */
if (mset[j].count == 0)
continue;
if (coalesce_entries(ms, mset[j].me, mset[j].count, if (coalesce_entries(ms, mset[j].me, mset[j].count,
&map->magic[j], &map->nmagic[j]) == -1) { &map->magic[j], &map->nmagic[j]) == -1) {
errs++; errs++;
@ -2086,6 +2089,13 @@ parse(struct magic_set *ms, struct magic_entry *me, const char *line,
return -1; return -1;
} }
if (m->type == FILE_NAME && cont_level != 0) {
if (ms->flags & MAGIC_CHECK)
file_magwarn(ms, "`name%s' entries can only be "
"declared at top level", l);
return -1;
}
/* New-style anding: "0 byte&0x80 =0x80 dynamically linked" */ /* New-style anding: "0 byte&0x80 =0x80 dynamically linked" */
/* New and improved: ~ & | ^ + - * / % -- exciting, isn't it? */ /* New and improved: ~ & | ^ + - * / % -- exciting, isn't it? */
@ -2699,7 +2709,7 @@ getvalue(struct magic_set *ms, struct magic *m, const char **p, int action)
ull = CAST(uint64_t, strtoull(*p, &ep, 0)); ull = CAST(uint64_t, strtoull(*p, &ep, 0));
m->value.q = file_signextend(ms, m, ull); m->value.q = file_signextend(ms, m, ull);
if (*p == ep) { if (*p == ep) {
file_magwarn(ms, "Unparseable number `%s'", *p); file_magwarn(ms, "Unparsable number `%s'", *p);
} else { } else {
size_t ts = typesize(m->type); size_t ts = typesize(m->type);
uint64_t x; uint64_t x;
@ -3101,8 +3111,8 @@ internal_loaded:
else else
version = ptr[1]; version = ptr[1];
if (version != VERSIONNO) { if (version != VERSIONNO) {
file_error(ms, 0, "File %d.%d supports only version %d magic " file_error(ms, 0, "File %d supports only version %d magic "
"files. `%s' is version %d", FILE_VERSION_MAJOR, patchlevel, "files. `%s' is version %d", MAGIC_VERSION,
VERSIONNO, dbname, version); VERSIONNO, dbname, version);
goto error; goto error;
} }

76
ext/fileinfo/libmagic/ascmagic.c
View File

@ -35,7 +35,7 @@
#include "file.h" #include "file.h"
#ifndef lint #ifndef lint
FILE_RCSID("@(#)$File: ascmagic.c,v 1.107 2020/06/08 19:58:36 christos Exp $") FILE_RCSID("@(#)$File: ascmagic.c,v 1.109 2021/02/05 23:01:40 christos Exp $")
#endif /* lint */ #endif /* lint */
#include "magic.h" #include "magic.h"
@ -50,7 +50,8 @@ FILE_RCSID("@(#)$File: ascmagic.c,v 1.107 2020/06/08 19:58:36 christos Exp $")
#define ISSPC(x) ((x) == ' ' || (x) == '\t' || (x) == '\r' || (x) == '\n' \ #define ISSPC(x) ((x) == ' ' || (x) == '\t' || (x) == '\r' || (x) == '\n' \
|| (x) == 0x85 || (x) == '\f') || (x) == 0x85 || (x) == '\f')
private unsigned char *encode_utf8(unsigned char *, size_t, unicodechar *, size_t); private unsigned char *encode_utf8(unsigned char *, size_t, file_unichar_t *,
size_t);
private size_t trim_nuls(const unsigned char *, size_t); private size_t trim_nuls(const unsigned char *, size_t);
/* /*
@ -69,7 +70,7 @@ trim_nuls(const unsigned char *buf, size_t nbytes)
protected int protected int
file_ascmagic(struct magic_set *ms, const struct buffer *b, int text) file_ascmagic(struct magic_set *ms, const struct buffer *b, int text)
{ {
unicodechar *ubuf = NULL; file_unichar_t *ubuf = NULL;
size_t ulen = 0; size_t ulen = 0;
int rv = 1; int rv = 1;
struct buffer bb; struct buffer bb;
@ -101,9 +102,9 @@ file_ascmagic(struct magic_set *ms, const struct buffer *b, int text)
} }
protected int protected int
file_ascmagic_with_encoding(struct magic_set *ms, file_ascmagic_with_encoding(struct magic_set *ms, const struct buffer *b,
const struct buffer *b, unicodechar *ubuf, size_t ulen, const char *code, file_unichar_t *ubuf, size_t ulen, const char *code, const char *type,
const char *type, int text) int text)
{ {
struct buffer bb; struct buffer bb;
const unsigned char *buf = CAST(const unsigned char *, b->fbuf); const unsigned char *buf = CAST(const unsigned char *, b->fbuf);
@ -127,7 +128,7 @@ file_ascmagic_with_encoding(struct magic_set *ms,
int executable = 0; int executable = 0;
size_t last_line_end = CAST(size_t, -1); size_t last_line_end = CAST(size_t, -1);
int has_long_lines = 0; size_t has_long_lines = 0;
nbytes = trim_nuls(buf, nbytes); nbytes = trim_nuls(buf, nbytes);
@ -190,8 +191,11 @@ file_ascmagic_with_encoding(struct magic_set *ms,
} }
/* If this line is _longer_ than MAXLINELEN, remember it. */ /* If this line is _longer_ than MAXLINELEN, remember it. */
if (i > last_line_end + MAXLINELEN) if (i > last_line_end + MAXLINELEN) {
has_long_lines = 1; size_t ll = i - last_line_end;
if (ll > has_long_lines)
has_long_lines = ll;
}
if (ubuf[i] == '\033') if (ubuf[i] == '\033')
has_escapes = 1; has_escapes = 1;
@ -269,7 +273,8 @@ file_ascmagic_with_encoding(struct magic_set *ms,
goto done; goto done;
if (has_long_lines) if (has_long_lines)
if (file_printf(ms, ", with very long lines") == -1) if (file_printf(ms, ", with very long lines (%zu)",
has_long_lines) == -1)
goto done; goto done;
/* /*
@ -281,7 +286,8 @@ file_ascmagic_with_encoding(struct magic_set *ms,
if (file_printf(ms, ", with") == -1) if (file_printf(ms, ", with") == -1)
goto done; goto done;
if (n_crlf == 0 && n_cr == 0 && n_nel == 0 && n_lf == 0) { if (n_crlf == 0 && n_cr == 0 &&
n_nel == 0 && n_lf == 0) {
if (file_printf(ms, " no") == -1) if (file_printf(ms, " no") == -1)
goto done; goto done;
} else { } else {
@ -335,7 +341,7 @@ done:
* after end of string, or NULL if an invalid character is found. * after end of string, or NULL if an invalid character is found.
*/ */
private unsigned char * private unsigned char *
encode_utf8(unsigned char *buf, size_t len, unicodechar *ubuf, size_t ulen) encode_utf8(unsigned char *buf, size_t len, file_unichar_t *ubuf, size_t ulen)
{ {
size_t i; size_t i;
unsigned char *end = buf + len; unsigned char *end = buf + len;
@ -345,43 +351,45 @@ encode_utf8(unsigned char *buf, size_t len, unicodechar *ubuf, size_t ulen)
if (end - buf < 1) if (end - buf < 1)
return NULL; return NULL;
*buf++ = CAST(unsigned char, ubuf[i]); *buf++ = CAST(unsigned char, ubuf[i]);
} else if (ubuf[i] <= 0x7ff) { continue;
}
if (ubuf[i] <= 0x7ff) {
if (end - buf < 2) if (end - buf < 2)
return NULL; return NULL;
*buf++ = CAST(unsigned char, (ubuf[i] >> 6) + 0xc0); *buf++ = CAST(unsigned char, (ubuf[i] >> 6) + 0xc0);
*buf++ = CAST(unsigned char, (ubuf[i] & 0x3f) + 0x80); goto out1;
} else if (ubuf[i] <= 0xffff) { }
if (ubuf[i] <= 0xffff) {
if (end - buf < 3) if (end - buf < 3)
return NULL; return NULL;
*buf++ = CAST(unsigned char, (ubuf[i] >> 12) + 0xe0); *buf++ = CAST(unsigned char, (ubuf[i] >> 12) + 0xe0);
*buf++ = CAST(unsigned char, ((ubuf[i] >> 6) & 0x3f) + 0x80); goto out2;
*buf++ = CAST(unsigned char, (ubuf[i] & 0x3f) + 0x80); }
} else if (ubuf[i] <= 0x1fffff) { if (ubuf[i] <= 0x1fffff) {
if (end - buf < 4) if (end - buf < 4)
return NULL; return NULL;
*buf++ = CAST(unsigned char, (ubuf[i] >> 18) + 0xf0); *buf++ = CAST(unsigned char, (ubuf[i] >> 18) + 0xf0);
*buf++ = CAST(unsigned char, ((ubuf[i] >> 12) & 0x3f) + 0x80); goto out3;
*buf++ = CAST(unsigned char, ((ubuf[i] >> 6) & 0x3f) + 0x80); }
*buf++ = CAST(unsigned char, (ubuf[i] & 0x3f) + 0x80); if (ubuf[i] <= 0x3ffffff) {
} else if (ubuf[i] <= 0x3ffffff) {
if (end - buf < 5) if (end - buf < 5)
return NULL; return NULL;
*buf++ = CAST(unsigned char, (ubuf[i] >> 24) + 0xf8); *buf++ = CAST(unsigned char, (ubuf[i] >> 24) + 0xf8);
*buf++ = CAST(unsigned char, ((ubuf[i] >> 18) & 0x3f) + 0x80); goto out4;
*buf++ = CAST(unsigned char, ((ubuf[i] >> 12) & 0x3f) + 0x80); }
*buf++ = CAST(unsigned char, ((ubuf[i] >> 6) & 0x3f) + 0x80); if (ubuf[i] <= 0x7fffffff) {
*buf++ = CAST(unsigned char, (ubuf[i] & 0x3f) + 0x80);
} else if (ubuf[i] <= 0x7fffffff) {
if (end - buf < 6) if (end - buf < 6)
return NULL; return NULL;
*buf++ = CAST(unsigned char, (ubuf[i] >> 30) + 0xfc); *buf++ = CAST(unsigned char, (ubuf[i] >> 30) + 0xfc);
*buf++ = CAST(unsigned char, ((ubuf[i] >> 24) & 0x3f) + 0x80); goto out5;
*buf++ = CAST(unsigned char, ((ubuf[i] >> 18) & 0x3f) + 0x80); }
*buf++ = CAST(unsigned char, ((ubuf[i] >> 12) & 0x3f) + 0x80); /* Invalid character */
*buf++ = CAST(unsigned char, ((ubuf[i] >> 6) & 0x3f) + 0x80); return NULL;
*buf++ = CAST(unsigned char, (ubuf[i] & 0x3f) + 0x80); out5: *buf++ = CAST(unsigned char, ((ubuf[i] >> 24) & 0x3f) + 0x80);
} else /* Invalid character */ out4: *buf++ = CAST(unsigned char, ((ubuf[i] >> 18) & 0x3f) + 0x80);
return NULL; out3: *buf++ = CAST(unsigned char, ((ubuf[i] >> 12) & 0x3f) + 0x80);
out2: *buf++ = CAST(unsigned char, ((ubuf[i] >> 6) & 0x3f) + 0x80);
out1: *buf++ = CAST(unsigned char, ((ubuf[i] >> 0) & 0x3f) + 0x80);
} }
return buf; return buf;

27
ext/fileinfo/libmagic/compress.c
View File

@ -35,7 +35,7 @@
#include "file.h" #include "file.h"
#ifndef lint #ifndef lint
FILE_RCSID("@(#)$File: compress.c,v 1.127 2020/05/31 00:11:06 christos Exp $") FILE_RCSID("@(#)$File: compress.c,v 1.129 2020/12/08 21:26:00 christos Exp $")
#endif #endif
#include "magic.h" #include "magic.h"
@ -72,7 +72,7 @@ typedef void (*sig_t)(int);
#include <bzlib.h> #include <bzlib.h>
#endif #endif
#if defined(HAVE_XZLIB_H) && defined(XZLIBSUPPORT) #if defined(HAVE_LZMA_H) && defined(XZLIBSUPPORT)
#define BUILTIN_XZLIB #define BUILTIN_XZLIB
#include <lzma.h> #include <lzma.h>
#endif #endif
@ -847,8 +847,23 @@ uncompressbuf(int fd, size_t bytes_max, size_t method, const unsigned char *old,
for (i = 0; i < __arraycount(fdp); i++) for (i = 0; i < __arraycount(fdp); i++)
fdp[i][0] = fdp[i][1] = -1; fdp[i][0] = fdp[i][1] = -1;
if ((fd == -1 && pipe(fdp[STDIN_FILENO]) == -1) || /*
pipe(fdp[STDOUT_FILENO]) == -1 || pipe(fdp[STDERR_FILENO]) == -1) { * There are multithreaded users who run magic_file()
* from dozens of threads. If two parallel magic_file() calls
* analyze two large compressed files, both will spawn
* an uncompressing child here, which writes out uncompressed data.
* We read some portion, then close the pipe, then waitpid() the child.
* If uncompressed data is larger, child shound get EPIPE and exit.
* However, with *parallel* calls OTHER child may unintentionally
* inherit pipe fds, thus keeping pipe open and making writes in
* our child block instead of failing with EPIPE!
* (For the bug to occur, two threads must mutually inherit their pipes,
* and both must have large outputs. Thus it happens not that often).
* To avoid this, be sure to create pipes with O_CLOEXEC.
*/
if ((fd == -1 && file_pipe_closexec(fdp[STDIN_FILENO]) == -1) ||
file_pipe_closexec(fdp[STDOUT_FILENO]) == -1 ||
file_pipe_closexec(fdp[STDERR_FILENO]) == -1) {
closep(fdp[STDIN_FILENO]); closep(fdp[STDIN_FILENO]);
closep(fdp[STDOUT_FILENO]); closep(fdp[STDOUT_FILENO]);
return makeerror(newch, n, "Cannot create pipe, %s", return makeerror(newch, n, "Cannot create pipe, %s",
@ -879,16 +894,20 @@ uncompressbuf(int fd, size_t bytes_max, size_t method, const unsigned char *old,
if (fdp[STDIN_FILENO][1] > 2) if (fdp[STDIN_FILENO][1] > 2)
(void) close(fdp[STDIN_FILENO][1]); (void) close(fdp[STDIN_FILENO][1]);
} }
file_clear_closexec(STDIN_FILENO);
///FIXME: if one of the fdp[i][j] is 0 or 1, this can bomb spectacularly ///FIXME: if one of the fdp[i][j] is 0 or 1, this can bomb spectacularly
if (copydesc(STDOUT_FILENO, fdp[STDOUT_FILENO][1])) if (copydesc(STDOUT_FILENO, fdp[STDOUT_FILENO][1]))
(void) close(fdp[STDOUT_FILENO][1]); (void) close(fdp[STDOUT_FILENO][1]);
if (fdp[STDOUT_FILENO][0] > 2) if (fdp[STDOUT_FILENO][0] > 2)
(void) close(fdp[STDOUT_FILENO][0]); (void) close(fdp[STDOUT_FILENO][0]);
file_clear_closexec(STDOUT_FILENO);
if (copydesc(STDERR_FILENO, fdp[STDERR_FILENO][1])) if (copydesc(STDERR_FILENO, fdp[STDERR_FILENO][1]))
(void) close(fdp[STDERR_FILENO][1]); (void) close(fdp[STDERR_FILENO][1]);
if (fdp[STDERR_FILENO][0] > 2) if (fdp[STDERR_FILENO][0] > 2)
(void) close(fdp[STDERR_FILENO][0]); (void) close(fdp[STDERR_FILENO][0]);
file_clear_closexec(STDERR_FILENO);
(void)execvp(compr[method].argv[0], (void)execvp(compr[method].argv[0],
RCAST(char *const *, RCAST(intptr_t, compr[method].argv))); RCAST(char *const *, RCAST(intptr_t, compr[method].argv)));

1
ext/fileinfo/libmagic/config.h
View File

@ -1 +0,0 @@
#include "php.h"

3
ext/fileinfo/libmagic/der.c
View File

@ -35,7 +35,7 @@
#include "file.h" #include "file.h"
#ifndef lint #ifndef lint
FILE_RCSID("@(#)$File: der.c,v 1.20 2020/06/07 19:10:37 christos Exp $") FILE_RCSID("@(#)$File: der.c,v 1.21 2020/06/15 00:58:10 christos Exp $")
#endif #endif
#else #else
#define SIZE_T_FORMAT "z" #define SIZE_T_FORMAT "z"
@ -249,7 +249,6 @@ der_data(char *buf, size_t blen, uint32_t tag, const void *q, uint32_t len)
return snprintf(buf, blen, return snprintf(buf, blen,
"20%c%c-%c%c-%c%c %c%c:%c%c:%c%c GMT", d[0], d[1], d[2], "20%c%c-%c%c-%c%c %c%c:%c%c:%c%c GMT", d[0], d[1], d[2],
d[3], d[4], d[5], d[6], d[7], d[8], d[9], d[10], d[11]); d[3], d[4], d[5], d[6], d[7], d[8], d[9], d[10], d[11]);
break;
default: default:
break; break;
} }

235
ext/fileinfo/libmagic/encoding.c
View File

@ -35,7 +35,7 @@
#include "file.h" #include "file.h"
#ifndef lint #ifndef lint
FILE_RCSID("@(#)$File: encoding.c,v 1.21 2019/06/08 20:49:14 christos Exp $") FILE_RCSID("@(#)$File: encoding.c,v 1.27 2021/02/05 21:33:49 christos Exp $")
#endif /* lint */ #endif /* lint */
#include "magic.h" #include "magic.h"
@ -43,14 +43,20 @@ FILE_RCSID("@(#)$File: encoding.c,v 1.21 2019/06/08 20:49:14 christos Exp $")
#include <stdlib.h> #include <stdlib.h>
private int looks_ascii(const unsigned char *, size_t, unicodechar *, size_t *); private int looks_ascii(const unsigned char *, size_t, file_unichar_t *,
private int looks_utf8_with_BOM(const unsigned char *, size_t, unicodechar *, size_t *);
private int looks_utf8_with_BOM(const unsigned char *, size_t, file_unichar_t *,
size_t *);
private int looks_utf7(const unsigned char *, size_t, file_unichar_t *,
size_t *);
private int looks_ucs16(const unsigned char *, size_t, file_unichar_t *,
size_t *);
private int looks_ucs32(const unsigned char *, size_t, file_unichar_t *,
size_t *);
private int looks_latin1(const unsigned char *, size_t, file_unichar_t *,
size_t *);
private int looks_extended(const unsigned char *, size_t, file_unichar_t *,
size_t *); size_t *);
private int looks_utf7(const unsigned char *, size_t, unicodechar *, size_t *);
private int looks_ucs16(const unsigned char *, size_t, unicodechar *, size_t *);
private int looks_ucs32(const unsigned char *, size_t, unicodechar *, size_t *);
private int looks_latin1(const unsigned char *, size_t, unicodechar *, size_t *);
private int looks_extended(const unsigned char *, size_t, unicodechar *, size_t *);
private void from_ebcdic(const unsigned char *, size_t, unsigned char *); private void from_ebcdic(const unsigned char *, size_t, unsigned char *);
#ifdef DEBUG_ENCODING #ifdef DEBUG_ENCODING
@ -62,19 +68,20 @@ private void from_ebcdic(const unsigned char *, size_t, unsigned char *);
/* /*
* Try to determine whether text is in some character code we can * Try to determine whether text is in some character code we can
* identify. Each of these tests, if it succeeds, will leave * identify. Each of these tests, if it succeeds, will leave
* the text converted into one-unicodechar-per-character Unicode in * the text converted into one-file_unichar_t-per-character Unicode in
* ubuf, and the number of characters converted in ulen. * ubuf, and the number of characters converted in ulen.
*/ */
protected int protected int
file_encoding(struct magic_set *ms, const struct buffer *b, unicodechar **ubuf, file_encoding(struct magic_set *ms, const struct buffer *b,
size_t *ulen, const char **code, const char **code_mime, const char **type) file_unichar_t **ubuf, size_t *ulen, const char **code,
const char **code_mime, const char **type)
{ {
const unsigned char *buf = CAST(const unsigned char *, b->fbuf); const unsigned char *buf = CAST(const unsigned char *, b->fbuf);
size_t nbytes = b->flen; size_t nbytes = b->flen;
size_t mlen; size_t mlen;
int rv = 1, ucs_type; int rv = 1, ucs_type;
unsigned char *nbuf = NULL; unsigned char *nbuf = NULL;
unicodechar *udefbuf; file_unichar_t *udefbuf;
size_t udeflen; size_t udeflen;
if (ubuf == NULL) if (ubuf == NULL)
@ -87,8 +94,12 @@ file_encoding(struct magic_set *ms, const struct buffer *b, unicodechar **ubuf,
*code = "unknown"; *code = "unknown";
*code_mime = "binary"; *code_mime = "binary";
if (nbytes > ms->encoding_max)
nbytes = ms->encoding_max;
mlen = (nbytes + 1) * sizeof((*ubuf)[0]); mlen = (nbytes + 1) * sizeof((*ubuf)[0]);
if ((*ubuf = CAST(unicodechar *, ecalloc(CAST(size_t, 1), mlen))) == NULL) { *ubuf = CAST(file_unichar_t *, ecalloc(CAST(size_t, 1), mlen));
if (*ubuf == NULL) {
file_oomem(ms, mlen); file_oomem(ms, mlen);
goto done; goto done;
} }
@ -102,7 +113,7 @@ file_encoding(struct magic_set *ms, const struct buffer *b, unicodechar **ubuf,
if (looks_ascii(buf, nbytes, *ubuf, ulen)) { if (looks_ascii(buf, nbytes, *ubuf, ulen)) {
if (looks_utf7(buf, nbytes, *ubuf, ulen) > 0) { if (looks_utf7(buf, nbytes, *ubuf, ulen) > 0) {
DPRINTF(("utf-7 %" SIZE_T_FORMAT "u\n", *ulen)); DPRINTF(("utf-7 %" SIZE_T_FORMAT "u\n", *ulen));
*code = "UTF-7 Unicode"; *code = "Unicode text, UTF-7";
*code_mime = "utf-7"; *code_mime = "utf-7";
} else { } else {
DPRINTF(("ascii %" SIZE_T_FORMAT "u\n", *ulen)); DPRINTF(("ascii %" SIZE_T_FORMAT "u\n", *ulen));
@ -111,27 +122,27 @@ file_encoding(struct magic_set *ms, const struct buffer *b, unicodechar **ubuf,
} }
} else if (looks_utf8_with_BOM(buf, nbytes, *ubuf, ulen) > 0) { } else if (looks_utf8_with_BOM(buf, nbytes, *ubuf, ulen) > 0) {
DPRINTF(("utf8/bom %" SIZE_T_FORMAT "u\n", *ulen)); DPRINTF(("utf8/bom %" SIZE_T_FORMAT "u\n", *ulen));
*code = "UTF-8 Unicode (with BOM)"; *code = "Unicode text, UTF-8 (with BOM)";
*code_mime = "utf-8"; *code_mime = "utf-8";
} else if (file_looks_utf8(buf, nbytes, *ubuf, ulen) > 1) { } else if (file_looks_utf8(buf, nbytes, *ubuf, ulen) > 1) {
DPRINTF(("utf8 %" SIZE_T_FORMAT "u\n", *ulen)); DPRINTF(("utf8 %" SIZE_T_FORMAT "u\n", *ulen));
*code = "UTF-8 Unicode"; *code = "Unicode text, UTF-8";
*code_mime = "utf-8"; *code_mime = "utf-8";
} else if ((ucs_type = looks_ucs32(buf, nbytes, *ubuf, ulen)) != 0) { } else if ((ucs_type = looks_ucs32(buf, nbytes, *ubuf, ulen)) != 0) {
if (ucs_type == 1) { if (ucs_type == 1) {
*code = "Little-endian UTF-32 Unicode"; *code = "Unicode text, UTF-32, little-endian";
*code_mime = "utf-32le"; *code_mime = "utf-32le";
} else { } else {
*code = "Big-endian UTF-32 Unicode"; *code = "Unicode text, UTF-32, big-endian";
*code_mime = "utf-32be"; *code_mime = "utf-32be";
} }
DPRINTF(("ucs32 %" SIZE_T_FORMAT "u\n", *ulen)); DPRINTF(("ucs32 %" SIZE_T_FORMAT "u\n", *ulen));
} else if ((ucs_type = looks_ucs16(buf, nbytes, *ubuf, ulen)) != 0) { } else if ((ucs_type = looks_ucs16(buf, nbytes, *ubuf, ulen)) != 0) {
if (ucs_type == 1) { if (ucs_type == 1) {
*code = "Little-endian UTF-16 Unicode"; *code = "Unicode text, UTF-16, little-endian";
*code_mime = "utf-16le"; *code_mime = "utf-16le";
} else { } else {
*code = "Big-endian UTF-16 Unicode"; *code = "Unicode text, UTF-16, big-endian";
*code_mime = "utf-16be"; *code_mime = "utf-16be";
} }
DPRINTF(("ucs16 %" SIZE_T_FORMAT "u\n", *ulen)); DPRINTF(("ucs16 %" SIZE_T_FORMAT "u\n", *ulen));
@ -249,64 +260,40 @@ private char text_chars[256] = {
I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, I /* 0xfX */ I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, I /* 0xfX */
}; };
private int #define LOOKS(NAME, COND) \
looks_ascii(const unsigned char *buf, size_t nbytes, unicodechar *ubuf, private int \
size_t *ulen) looks_ ## NAME(const unsigned char *buf, size_t nbytes, file_unichar_t *ubuf, \
{ size_t *ulen) \
size_t i; { \
size_t i, u; \
*ulen = 0; unsigned char dist[256]; \
memset(dist, 0, sizeof(dist)); \
for (i = 0; i < nbytes; i++) { \
int t = text_chars[buf[i]]; *ulen = 0; \
\
if (t != T) for (i = 0; i < nbytes; i++) { \
return 0; int t = text_chars[buf[i]]; \
\
ubuf[(*ulen)++] = buf[i]; if (COND) \
} return 0; \
\
return 1; ubuf[(*ulen)++] = buf[i]; \
dist[buf[i]]++; \
} \
u = 0; \
for (i = 0; i < __arraycount(dist); i++) { \
if (dist[i]) \
u += dist[i]; \
} \
if (u < 3) \
return 0; \
\
return 1; \
} }
private int LOOKS(ascii, t != T)
looks_latin1(const unsigned char *buf, size_t nbytes, unicodechar *ubuf, size_t *ulen) LOOKS(latin1, t != T && t != I)
{ LOOKS(extended, t != T && t != I && t != X)
size_t i;
*ulen = 0;
for (i = 0; i < nbytes; i++) {
int t = text_chars[buf[i]];
if (t != T && t != I)
return 0;
ubuf[(*ulen)++] = buf[i];
}
return 1;
}
private int
looks_extended(const unsigned char *buf, size_t nbytes, unicodechar *ubuf,
size_t *ulen)
{
size_t i;
*ulen = 0;
for (i = 0; i < nbytes; i++) {
int t = text_chars[buf[i]];
if (t != T && t != I && t != X)
return 0;
ubuf[(*ulen)++] = buf[i];
}
return 1;
}
/* /*
* Decide whether some text looks like UTF-8. Returns: * Decide whether some text looks like UTF-8. Returns:
@ -319,12 +306,65 @@ looks_extended(const unsigned char *buf, size_t nbytes, unicodechar *ubuf,
* If ubuf is non-NULL on entry, text is decoded into ubuf, *ulen; * If ubuf is non-NULL on entry, text is decoded into ubuf, *ulen;
* ubuf must be big enough! * ubuf must be big enough!
*/ */
// from: https://golang.org/src/unicode/utf8/utf8.go
#define XX 0xF1 // invalid: size 1
#define AS 0xF0 // ASCII: size 1
#define S1 0x02 // accept 0, size 2
#define S2 0x13 // accept 1, size 3
#define S3 0x03 // accept 0, size 3
#define S4 0x23 // accept 2, size 3
#define S5 0x34 // accept 3, size 4
#define S6 0x04 // accept 0, size 4
#define S7 0x44 // accept 4, size 4
#define LOCB 0x80
#define HICB 0xBF
// first is information about the first byte in a UTF-8 sequence.
static const uint8_t first[] = {
// 1 2 3 4 5 6 7 8 9 A B C D E F
AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, // 0x00-0x0F
AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, // 0x10-0x1F
AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, // 0x20-0x2F
AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, // 0x30-0x3F
AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, // 0x40-0x4F
AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, // 0x50-0x5F
AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, // 0x60-0x6F
AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, AS, // 0x70-0x7F
// 1 2 3 4 5 6 7 8 9 A B C D E F
XX, XX, XX, XX, XX, XX, XX, XX, XX, XX, XX, XX, XX, XX, XX, XX, // 0x80-0x8F
XX, XX, XX, XX, XX, XX, XX, XX, XX, XX, XX, XX, XX, XX, XX, XX, // 0x90-0x9F
XX, XX, XX, XX, XX, XX, XX, XX, XX, XX, XX, XX, XX, XX, XX, XX, // 0xA0-0xAF
XX, XX, XX, XX, XX, XX, XX, XX, XX, XX, XX, XX, XX, XX, XX, XX, // 0xB0-0xBF
XX, XX, S1, S1, S1, S1, S1, S1, S1, S1, S1, S1, S1, S1, S1, S1, // 0xC0-0xCF
S1, S1, S1, S1, S1, S1, S1, S1, S1, S1, S1, S1, S1, S1, S1, S1, // 0xD0-0xDF
S2, S3, S3, S3, S3, S3, S3, S3, S3, S3, S3, S3, S3, S4, S3, S3, // 0xE0-0xEF
S5, S6, S6, S6, S7, XX, XX, XX, XX, XX, XX, XX, XX, XX, XX, XX, // 0xF0-0xFF
};
// acceptRange gives the range of valid values for the second byte in a UTF-8
// sequence.
struct accept_range {
uint8_t lo; // lowest value for second byte.
uint8_t hi; // highest value for second byte.
} accept_ranges[16] = {
// acceptRanges has size 16 to avoid bounds checks in the code that uses it.
{ LOCB, HICB },
{ 0xA0, HICB },
{ LOCB, 0x9F },
{ 0x90, HICB },
{ LOCB, 0x8F },
};
protected int protected int
file_looks_utf8(const unsigned char *buf, size_t nbytes, unicodechar *ubuf, size_t *ulen) file_looks_utf8(const unsigned char *buf, size_t nbytes, file_unichar_t *ubuf,
size_t *ulen)
{ {
size_t i; size_t i;
int n; int n;
unicodechar c; file_unichar_t c;
int gotone = 0, ctrl = 0; int gotone = 0, ctrl = 0;
if (ubuf) if (ubuf)
@ -346,6 +386,10 @@ file_looks_utf8(const unsigned char *buf, size_t nbytes, unicodechar *ubuf, size
return -1; return -1;
} else { /* 11xxxxxx begins UTF-8 */ } else { /* 11xxxxxx begins UTF-8 */
int following; int following;
uint8_t x = first[buf[i]];
const struct accept_range *ar = &accept_ranges[x >> 4];
if (x == XX)
return -1;
if ((buf[i] & 0x20) == 0) { /* 110xxxxx */ if ((buf[i] & 0x20) == 0) { /* 110xxxxx */
c = buf[i] & 0x1f; c = buf[i] & 0x1f;
@ -370,6 +414,10 @@ file_looks_utf8(const unsigned char *buf, size_t nbytes, unicodechar *ubuf, size
if (i >= nbytes) if (i >= nbytes)
goto done; goto done;
if (n == 0 &&
(buf[i] < ar->lo || buf[i] > ar->hi))
return -1;
if ((buf[i] & 0x80) == 0 || (buf[i] & 0x40)) if ((buf[i] & 0x80) == 0 || (buf[i] & 0x40))
return -1; return -1;
@ -391,8 +439,8 @@ done:
* rest of the text. * rest of the text.
*/ */
private int private int
looks_utf8_with_BOM(const unsigned char *buf, size_t nbytes, unicodechar *ubuf, looks_utf8_with_BOM(const unsigned char *buf, size_t nbytes,
size_t *ulen) file_unichar_t *ubuf, size_t *ulen)
{ {
if (nbytes > 3 && buf[0] == 0xef && buf[1] == 0xbb && buf[2] == 0xbf) if (nbytes > 3 && buf[0] == 0xef && buf[1] == 0xbb && buf[2] == 0xbf)
return file_looks_utf8(buf + 3, nbytes - 3, ubuf, ulen); return file_looks_utf8(buf + 3, nbytes - 3, ubuf, ulen);
@ -401,7 +449,8 @@ looks_utf8_with_BOM(const unsigned char *buf, size_t nbytes, unicodechar *ubuf,
} }
private int private int
looks_utf7(const unsigned char *buf, size_t nbytes, unicodechar *ubuf, size_t *ulen) looks_utf7(const unsigned char *buf, size_t nbytes, file_unichar_t *ubuf,
size_t *ulen)
{ {
if (nbytes > 4 && buf[0] == '+' && buf[1] == '/' && buf[2] == 'v') if (nbytes > 4 && buf[0] == '+' && buf[1] == '/' && buf[2] == 'v')
switch (buf[3]) { switch (buf[3]) {
@ -420,7 +469,7 @@ looks_utf7(const unsigned char *buf, size_t nbytes, unicodechar *ubuf, size_t *u
} }
private int private int
looks_ucs16(const unsigned char *bf, size_t nbytes, unicodechar *ubf, looks_ucs16(const unsigned char *bf, size_t nbytes, file_unichar_t *ubf,
size_t *ulen) size_t *ulen)
{ {
int bigend; int bigend;
@ -443,10 +492,10 @@ looks_ucs16(const unsigned char *bf, size_t nbytes, unicodechar *ubf,
if (bigend) if (bigend)
ubf[(*ulen)++] = bf[i + 1] ubf[(*ulen)++] = bf[i + 1]
| (CAST(unicodechar, bf[i]) << 8); | (CAST(file_unichar_t, bf[i]) << 8);
else else
ubf[(*ulen)++] = bf[i] ubf[(*ulen)++] = bf[i]
| (CAST(unicodechar, bf[i + 1]) << 8); | (CAST(file_unichar_t, bf[i + 1]) << 8);
if (ubf[*ulen - 1] == 0xfffe) if (ubf[*ulen - 1] == 0xfffe)
return 0; return 0;
@ -459,7 +508,7 @@ looks_ucs16(const unsigned char *bf, size_t nbytes, unicodechar *ubf,
} }
private int private int
looks_ucs32(const unsigned char *bf, size_t nbytes, unicodechar *ubf, looks_ucs32(const unsigned char *bf, size_t nbytes, file_unichar_t *ubf,
size_t *ulen) size_t *ulen)
{ {
int bigend; int bigend;
@ -481,15 +530,15 @@ looks_ucs32(const unsigned char *bf, size_t nbytes, unicodechar *ubf,
/* XXX fix to properly handle chars > 65536 */ /* XXX fix to properly handle chars > 65536 */
if (bigend) if (bigend)
ubf[(*ulen)++] = CAST(unicodechar, bf[i + 3]) ubf[(*ulen)++] = CAST(file_unichar_t, bf[i + 3])
| (CAST(unicodechar, bf[i + 2]) << 8) | (CAST(file_unichar_t, bf[i + 2]) << 8)
| (CAST(unicodechar, bf[i + 1]) << 16) | (CAST(file_unichar_t, bf[i + 1]) << 16)
| (CAST(unicodechar, bf[i]) << 24); | (CAST(file_unichar_t, bf[i]) << 24);
else else
ubf[(*ulen)++] = CAST(unicodechar, bf[i + 0]) ubf[(*ulen)++] = CAST(file_unichar_t, bf[i + 0])
| (CAST(unicodechar, bf[i + 1]) << 8) | (CAST(file_unichar_t, bf[i + 1]) << 8)
| (CAST(unicodechar, bf[i + 2]) << 16) | (CAST(file_unichar_t, bf[i + 2]) << 16)
| (CAST(unicodechar, bf[i + 3]) << 24); | (CAST(file_unichar_t, bf[i + 3]) << 24);
if (ubf[*ulen - 1] == 0xfffe) if (ubf[*ulen - 1] == 0xfffe)
return 0; return 0;

39
ext/fileinfo/libmagic/file.h
View File

@ -27,7 +27,7 @@
*/ */
/* /*
* file.h - definitions for file(1) program * file.h - definitions for file(1) program
* @(#)$File: file.h,v 1.220 2020/06/08 17:38:27 christos Exp $ * @(#)$File: file.h,v 1.225 2021/02/05 22:29:07 christos Exp $
*/ */
#ifndef __file_h__ #ifndef __file_h__
@ -35,6 +35,7 @@
#include "config.h" #include "config.h"
#include "php.h"
#include "ext/standard/php_string.h" #include "ext/standard/php_string.h"
#include "ext/pcre/php_pcre.h" #include "ext/pcre/php_pcre.h"
@ -136,6 +137,14 @@
#define MAX(a,b) (((a) > (b)) ? (a) : (b)) #define MAX(a,b) (((a) > (b)) ? (a) : (b))
#endif #endif
#ifndef O_CLOEXEC
# define O_CLOEXEC 0
#endif
#ifndef FD_CLOEXEC
# define FD_CLOEXEC 1
#endif
#define FILE_BADSIZE CAST(size_t, ~0ul) #define FILE_BADSIZE CAST(size_t, ~0ul)
#define MAXDESC 64 /* max len of text description/MIME type */ #define MAXDESC 64 /* max len of text description/MIME type */
#define MAXMIME 80 /* max len of text MIME type */ #define MAXMIME 80 /* max len of text MIME type */
@ -403,14 +412,16 @@ struct level_info {
#endif #endif
}; };
struct cont {
size_t len;
struct level_info *li;
};
#define MAGIC_SETS 2 #define MAGIC_SETS 2
struct magic_set { struct magic_set {
struct mlist *mlist[MAGIC_SETS]; /* list of regular entries */ struct mlist *mlist[MAGIC_SETS]; /* list of regular entries */
struct cont { struct cont c;
size_t len;
struct level_info *li;
} c;
struct out { struct out {
char *buf; /* Accumulation buffer */ char *buf; /* Accumulation buffer */
size_t blen; /* Length of buffer */ size_t blen; /* Length of buffer */
@ -445,6 +456,7 @@ struct magic_set {
uint16_t elf_notes_max; uint16_t elf_notes_max;
uint16_t regex_max; uint16_t regex_max;
size_t bytes_max; /* number of bytes to read from file */ size_t bytes_max; /* number of bytes to read from file */
size_t encoding_max; /* bytes to look for encoding */
#ifndef FILE_BYTES_MAX #ifndef FILE_BYTES_MAX
# define FILE_BYTES_MAX (1024 * 1024) /* how much of the file to look at */ # define FILE_BYTES_MAX (1024 * 1024) /* how much of the file to look at */
#endif #endif
@ -454,11 +466,13 @@ struct magic_set {
#define FILE_INDIR_MAX 50 #define FILE_INDIR_MAX 50
#define FILE_NAME_MAX 50 #define FILE_NAME_MAX 50
#define FILE_REGEX_MAX 8192 #define FILE_REGEX_MAX 8192
#define FILE_ENCODING_MAX (64 * 1024)
}; };
/* Type for Unicode characters */ /* Type for Unicode characters */
typedef unsigned long unicodechar; typedef unsigned long file_unichar_t;
struct stat;
#define FILE_T_LOCAL 1 #define FILE_T_LOCAL 1
#define FILE_T_WINDOWS 2 #define FILE_T_WINDOWS 2
protected const char *file_fmttime(char *, size_t, uint64_t, int); protected const char *file_fmttime(char *, size_t, uint64_t, int);
@ -468,6 +482,8 @@ protected int file_buffer(struct magic_set *, php_stream *, zend_stat_t *, const
size_t); size_t);
protected int file_fsmagic(struct magic_set *, const char *, zend_stat_t *); protected int file_fsmagic(struct magic_set *, const char *, zend_stat_t *);
protected int file_pipe2file(struct magic_set *, int, const void *, size_t); protected int file_pipe2file(struct magic_set *, int, const void *, size_t);
protected int file_vprintf(struct magic_set *, const char *, va_list)
__attribute__((__format__(__printf__, 2, 0)));
protected int file_separator(struct magic_set *); protected int file_separator(struct magic_set *);
protected char *file_copystr(char *, size_t, size_t, const char *); protected char *file_copystr(char *, size_t, size_t, const char *);
protected int file_checkfmt(char *, size_t, const char *); protected int file_checkfmt(char *, size_t, const char *);
@ -486,15 +502,17 @@ protected int file_zmagic(struct magic_set *, const struct buffer *,
protected int file_ascmagic(struct magic_set *, const struct buffer *, protected int file_ascmagic(struct magic_set *, const struct buffer *,
int); int);
protected int file_ascmagic_with_encoding(struct magic_set *, protected int file_ascmagic_with_encoding(struct magic_set *,
const struct buffer *, unicodechar *, size_t, const char *, const char *, int); const struct buffer *, file_unichar_t *, size_t, const char *, const char *, int);
protected int file_encoding(struct magic_set *, const struct buffer *, protected int file_encoding(struct magic_set *, const struct buffer *,
unicodechar **, size_t *, const char **, const char **, const char **); file_unichar_t **, size_t *, const char **, const char **, const char **);
protected int file_is_json(struct magic_set *, const struct buffer *); protected int file_is_json(struct magic_set *, const struct buffer *);
protected int file_is_csv(struct magic_set *, const struct buffer *, int); protected int file_is_csv(struct magic_set *, const struct buffer *, int);
protected int file_is_tar(struct magic_set *, const struct buffer *); protected int file_is_tar(struct magic_set *, const struct buffer *);
protected int file_softmagic(struct magic_set *, const struct buffer *, protected int file_softmagic(struct magic_set *, const struct buffer *,
uint16_t *, uint16_t *, int, int); uint16_t *, uint16_t *, int, int);
protected int file_apprentice(struct magic_set *, const char *, int); protected int file_apprentice(struct magic_set *, const char *, int);
protected int buffer_apprentice(struct magic_set *, struct magic **,
size_t *, size_t);
protected int file_magicfind(struct magic_set *, const char *, struct mlist *); protected int file_magicfind(struct magic_set *, const char *, struct mlist *);
protected uint64_t file_signextend(struct magic_set *, struct magic *, protected uint64_t file_signextend(struct magic_set *, struct magic *,
uint64_t); uint64_t);
@ -510,7 +528,7 @@ protected size_t file_mbswidth(const char *);
protected const char *file_getbuffer(struct magic_set *); protected const char *file_getbuffer(struct magic_set *);
protected ssize_t sread(int, void *, size_t, int); protected ssize_t sread(int, void *, size_t, int);
protected int file_check_mem(struct magic_set *, unsigned int); protected int file_check_mem(struct magic_set *, unsigned int);
protected int file_looks_utf8(const unsigned char *, size_t, unicodechar *, protected int file_looks_utf8(const unsigned char *, size_t, file_unichar_t *,
size_t *); size_t *);
protected size_t file_pstring_length_size(struct magic_set *, protected size_t file_pstring_length_size(struct magic_set *,
const struct magic *); const struct magic *);
@ -521,6 +539,9 @@ protected char * file_printable(char *, size_t, const char *, size_t);
protected int file_os2_apptype(struct magic_set *, const char *, const void *, protected int file_os2_apptype(struct magic_set *, const char *, const void *,
size_t); size_t);
#endif /* __EMX__ */ #endif /* __EMX__ */
protected int file_pipe_closexec(int *);
protected int file_clear_closexec(int);
protected char *file_strtrim(char *);
protected void buffer_init(struct buffer *, int, const zend_stat_t *, protected void buffer_init(struct buffer *, int, const zend_stat_t *,
const void *, size_t); const void *, size_t);

156
ext/fileinfo/libmagic/funcs.c
View File

@ -27,7 +27,7 @@
#include "file.h" #include "file.h"
#ifndef lint #ifndef lint
FILE_RCSID("@(#)$File: funcs.c,v 1.115 2020/02/20 15:50:20 christos Exp $") FILE_RCSID("@(#)$File: funcs.c,v 1.121 2021/02/05 22:29:07 christos Exp $")
#endif /* lint */ #endif /* lint */
#include "magic.h" #include "magic.h"
@ -36,6 +36,9 @@ FILE_RCSID("@(#)$File: funcs.c,v 1.115 2020/02/20 15:50:20 christos Exp $")
#include <stdlib.h> #include <stdlib.h>
#include <string.h> #include <string.h>
#include <ctype.h> #include <ctype.h>
#ifdef HAVE_UNISTD_H
#include <unistd.h> /* for pipe2() */
#endif
#if defined(HAVE_WCHAR_H) #if defined(HAVE_WCHAR_H)
#include <wchar.h> #include <wchar.h>
#endif #endif
@ -100,7 +103,7 @@ file_checkfmt(char *msg, size_t mlen, const char *fmt)
if (*++p == '%') if (*++p == '%')
continue; continue;
// Skip uninteresting. // Skip uninteresting.
while (strchr("0.'+- ", *p) != NULL) while (strchr("#0.'+- ", *p) != NULL)
p++; p++;
if (*p == '*') { if (*p == '*') {
if (msg) if (msg)
@ -126,27 +129,56 @@ file_checkfmt(char *msg, size_t mlen, const char *fmt)
return 0; return 0;
} }
/*
* Like printf, only we append to a buffer.
*/
protected int
file_vprintf(struct magic_set *ms, const char *fmt, va_list ap)
{
size_t len;
char *buf, *newstr;
char tbuf[1024];
if (ms->event_flags & EVENT_HAD_ERR)
return 0;
if (file_checkfmt(tbuf, sizeof(tbuf), fmt)) {
file_clearbuf(ms);
file_error(ms, 0, "Bad magic format `%s' (%s)", fmt, tbuf);
return -1;
}
len = vspprintf(&buf, 0, fmt, ap);
if (len > 1024 || len + ms->o.blen > 1024 * 1024) {
size_t blen = ms->o.blen;
if (buf) efree(buf);
file_clearbuf(ms);
file_error(ms, 0, "Output buffer space exceeded %d+%zu", len,
blen);
return -1;
}
if (ms->o.buf != NULL) {
len = spprintf(&newstr, 0, "%s%s", ms->o.buf, buf);
efree(buf);
efree(ms->o.buf);
buf = newstr;
}
ms->o.buf = buf;
ms->o.blen = len;
return 0;
}
protected int protected int
file_printf(struct magic_set *ms, const char *fmt, ...) file_printf(struct magic_set *ms, const char *fmt, ...)
{ {
int rv;
va_list ap; va_list ap;
char *buf = NULL, *newstr;
va_start(ap, fmt); va_start(ap, fmt);
vspprintf(&buf, 0, fmt, ap); rv = file_vprintf(ms, fmt, ap);
va_end(ap); va_end(ap);
return rv;
if (ms->o.buf != NULL) {
spprintf(&newstr, 0, "%s%s", ms->o.buf, (buf ? buf : ""));
if (buf) {
efree(buf);
}
efree(ms->o.buf);
ms->o.buf = newstr;
} else {
ms->o.buf = buf;
}
return 0;
} }
/* /*
@ -157,30 +189,18 @@ private void
file_error_core(struct magic_set *ms, int error, const char *f, va_list va, file_error_core(struct magic_set *ms, int error, const char *f, va_list va,
size_t lineno) size_t lineno)
{ {
char *buf = NULL;
/* Only the first error is ok */ /* Only the first error is ok */
if (ms->event_flags & EVENT_HAD_ERR) if (ms->event_flags & EVENT_HAD_ERR)
return; return;
if (lineno != 0) { if (lineno != 0) {
efree(ms->o.buf); file_clearbuf(ms);
ms->o.buf = NULL; (void)file_printf(ms, "line %" SIZE_T_FORMAT "u:", lineno);
file_printf(ms, "line %" SIZE_T_FORMAT "u:", lineno);
} }
if (ms->o.buf && *ms->o.buf)
vspprintf(&buf, 0, f, va); (void)file_printf(ms, " ");
va_end(va); (void)file_vprintf(ms, f, va);
if (error > 0)
if (error > 0) { (void)file_printf(ms, " (%s)", strerror(error));
file_printf(ms, "%s (%s)", (*buf ? buf : ""), strerror(error));
} else if (*buf) {
file_printf(ms, "%s", buf);
}
if (buf) {
efree(buf);
}
ms->event_flags |= EVENT_HAD_ERR; ms->event_flags |= EVENT_HAD_ERR;
ms->error = error; ms->error = error;
} }
@ -228,11 +248,31 @@ file_badread(struct magic_set *ms)
} }
#ifndef COMPILE_ONLY #ifndef COMPILE_ONLY
#define FILE_SEPARATOR "\n- "
protected int protected int
file_separator(struct magic_set *ms) file_separator(struct magic_set *ms)
{ {
return file_printf(ms, "\n- "); return file_printf(ms, FILE_SEPARATOR);
}
static void
trim_separator(struct magic_set *ms)
{
size_t l;
if (ms->o.buf == NULL)
return;
l = strlen(ms->o.buf);
if (l < sizeof(FILE_SEPARATOR))
return;
l -= sizeof(FILE_SEPARATOR) - 1;
if (strcmp(ms->o.buf + l, FILE_SEPARATOR) != 0)
return;
ms->o.buf[l] = '\0';
} }
static int static int
@ -450,6 +490,7 @@ simple:
rv = -1; rv = -1;
} }
done: done:
trim_separator(ms);
if ((ms->flags & MAGIC_MIME_ENCODING) != 0) { if ((ms->flags & MAGIC_MIME_ENCODING) != 0) {
if (ms->flags & MAGIC_MIME_TYPE) if (ms->flags & MAGIC_MIME_TYPE)
if (file_printf(ms, "; charset=") == -1) if (file_printf(ms, "; charset=") == -1)
@ -598,7 +639,7 @@ file_check_mem(struct magic_set *ms, unsigned int level)
protected size_t protected size_t
file_printedlen(const struct magic_set *ms) file_printedlen(const struct magic_set *ms)
{ {
return ms->o.buf == NULL ? 0 : strlen(ms->o.buf); return ms->o.blen;
} }
protected int protected int
@ -717,7 +758,7 @@ struct guid {
protected int protected int
file_parse_guid(const char *s, uint64_t *guid) file_parse_guid(const char *s, uint64_t *guid)
{ {
struct guid *g = CAST(struct guid *, guid); struct guid *g = CAST(struct guid *, CAST(void *, guid));
return sscanf(s, return sscanf(s,
"%8x-%4hx-%4hx-%2hhx%2hhx-%2hhx%2hhx%2hhx%2hhx%2hhx%2hhx", "%8x-%4hx-%4hx-%2hhx%2hhx-%2hhx%2hhx%2hhx%2hhx%2hhx%2hhx",
&g->data1, &g->data2, &g->data3, &g->data4[0], &g->data4[1], &g->data1, &g->data2, &g->data3, &g->data4[0], &g->data4[1],
@ -728,7 +769,8 @@ file_parse_guid(const char *s, uint64_t *guid)
protected int protected int
file_print_guid(char *str, size_t len, const uint64_t *guid) file_print_guid(char *str, size_t len, const uint64_t *guid)
{ {
const struct guid *g = CAST(const struct guid *, guid); const struct guid *g = CAST(const struct guid *,
CAST(const void *, guid));
return snprintf(str, len, "%.8X-%.4hX-%.4hX-%.2hhX%.2hhX-" return snprintf(str, len, "%.8X-%.4hX-%.4hX-%.2hhX%.2hhX-"
"%.2hhX%.2hhX%.2hhX%.2hhX%.2hhX%.2hhX", "%.2hhX%.2hhX%.2hhX%.2hhX%.2hhX%.2hhX",
@ -736,3 +778,39 @@ file_print_guid(char *str, size_t len, const uint64_t *guid)
g->data4[2], g->data4[3], g->data4[4], g->data4[5], g->data4[2], g->data4[3], g->data4[4], g->data4[5],
g->data4[6], g->data4[7]); g->data4[6], g->data4[7]);
} }
protected int
file_pipe_closexec(int *fds)
{
#ifdef HAVE_PIPE2
return pipe2(fds, O_CLOEXEC);
#else
if (pipe(fds) == -1)
return -1;
(void)fcntl(fds[0], F_SETFD, FD_CLOEXEC);
(void)fcntl(fds[1], F_SETFD, FD_CLOEXEC);
return 0;
#endif
}
protected int
file_clear_closexec(int fd) {
return fcntl(fd, F_SETFD, 0);
}
protected char *
file_strtrim(char *str)
{
char *last;
while (isspace(CAST(unsigned char, *str)))
str++;
last = str;
while (*last)
last++;
--last;
while (isspace(CAST(unsigned char, *last)))
last--;
*++last = '\0';
return str;
}

7
ext/fileinfo/libmagic/is_csv.c
View File

@ -32,7 +32,7 @@
#include "file.h" #include "file.h"
#ifndef lint #ifndef lint
FILE_RCSID("@(#)$File: is_csv.c,v 1.4 2019/06/26 20:31:31 christos Exp $") FILE_RCSID("@(#)$File: is_csv.c,v 1.6 2020/08/09 16:43:36 christos Exp $")
#endif #endif
#include <string.h> #include <string.h>
@ -94,8 +94,7 @@ csv_parse(const unsigned char *uc, const unsigned char *ue)
size_t nf = 0, tf = 0, nl = 0; size_t nf = 0, tf = 0, nl = 0;
while (uc < ue) { while (uc < ue) {
unsigned char c; switch (*uc++) {
switch (c = *uc++) {
case '"': case '"':
// Eat until the matching quote // Eat until the matching quote
uc = eatquote(uc, ue); uc = eatquote(uc, ue);
@ -150,7 +149,7 @@ file_is_csv(struct magic_set *ms, const struct buffer *b, int looks_text)
return 1; return 1;
if (mime) { if (mime) {
if (file_printf(ms, "application/csv") == -1) if (file_printf(ms, "text/csv") == -1)
return -1; return -1;
return 1; return 1;
} }

8
ext/fileinfo/libmagic/magic.c
View File

@ -28,7 +28,7 @@
#include "file.h" #include "file.h"
#ifndef lint #ifndef lint
FILE_RCSID("@(#)$File: magic.c,v 1.112 2020/06/08 19:44:10 christos Exp $") FILE_RCSID("@(#)$File: magic.c,v 1.114 2021/02/05 21:33:49 christos Exp $")
#endif /* lint */ #endif /* lint */
#include "magic.h" #include "magic.h"
@ -348,6 +348,9 @@ magic_setparam(struct magic_set *ms, int param, const void *val)
case MAGIC_PARAM_BYTES_MAX: case MAGIC_PARAM_BYTES_MAX:
ms->bytes_max = *CAST(const size_t *, val); ms->bytes_max = *CAST(const size_t *, val);
return 0; return 0;
case MAGIC_PARAM_ENCODING_MAX:
ms->encoding_max = *CAST(const size_t *, val);
return 0;
default: default:
errno = EINVAL; errno = EINVAL;
return -1; return -1;
@ -381,6 +384,9 @@ magic_getparam(struct magic_set *ms, int param, void *val)
case MAGIC_PARAM_BYTES_MAX: case MAGIC_PARAM_BYTES_MAX:
*CAST(size_t *, val) = ms->bytes_max; *CAST(size_t *, val) = ms->bytes_max;
return 0; return 0;
case MAGIC_PARAM_ENCODING_MAX:
*CAST(size_t *, val) = ms->encoding_max;
return 0;
default: default:
errno = EINVAL; errno = EINVAL;
return -1; return -1;

3
ext/fileinfo/libmagic/magic.h
View File

@ -113,7 +113,7 @@ b\31transp_compression\0\
#define MAGIC_NO_CHECK_FORTRAN 0x000000 /* Don't check ascii/fortran */ #define MAGIC_NO_CHECK_FORTRAN 0x000000 /* Don't check ascii/fortran */
#define MAGIC_NO_CHECK_TROFF 0x000000 /* Don't check ascii/troff */ #define MAGIC_NO_CHECK_TROFF 0x000000 /* Don't check ascii/troff */
#define MAGIC_VERSION 539 /* This implementation */ #define MAGIC_VERSION 540 /* This implementation */
#ifdef __cplusplus #ifdef __cplusplus
@ -150,6 +150,7 @@ int magic_errno(magic_t);
#define MAGIC_PARAM_ELF_NOTES_MAX 4 #define MAGIC_PARAM_ELF_NOTES_MAX 4
#define MAGIC_PARAM_REGEX_MAX 5 #define MAGIC_PARAM_REGEX_MAX 5
#define MAGIC_PARAM_BYTES_MAX 6 #define MAGIC_PARAM_BYTES_MAX 6
#define MAGIC_PARAM_ENCODING_MAX 7
int magic_setparam(magic_t, int, const void *); int magic_setparam(magic_t, int, const void *);
int magic_getparam(magic_t, int, void *); int magic_getparam(magic_t, int, void *);

390
ext/fileinfo/libmagic/patchlevel.h
View File

@ -1,390 +0,0 @@
#define FILE_VERSION_MAJOR 5
#define patchlevel 37
/*
* Patchlevel file for Ian Darwin's MAGIC command.
* $File: patchlevel.h,v 1.68 2008/03/22 21:39:43 christos Exp $
*
* $Log$
* Revision 2.1 2019/05/30 22:27:12 ab
* Update libmagic to 5.37
*
* $Log$
* Revision 2.1 2018/04/26 22:27:12 ab
* Update libmagic to 5.33
*
* $Log$
* Revision 2.0 2017/10/11 22:27:12 ab
* Update libmagic to 5.31
*
* $Log$
* Revision 1.9 2016/11/24 22:27:12 ab
* Update libmagic to 5.29
*
* $Log$
* Revision 1.9 2016/10/11 22:27:12 ab
* Update libmagic to 5.28
*
* $Log$
* Revision 1.9 2015/03/06 22:27:12 ab
* Update libmagic to 5.2X
*
* $Log$
* Revision 1.8 2014/02/18 22:27:12 ab
* Update libmagic to 5.17
*
* $Log$
* Revision 1.7 2013/03/26 22:27:12 ab
* Update libmagic to 5.14
*
* $Log$
* Revision 1.6 2012/03/26 21:01:37 ab
* Update libmagic to 5.11
*
* Revision 1.5 2012/03/25 13:54:37 ab
* Update libmagic to 5.04
*
* Revision 1.4 2009/05/04 20:52:43 scottmac
* Update libmagic to 5.02
*
* Revision 1.3 2009/03/15 23:02:35 scottmac
* Update fileinfo to libmagic 5.00 and remove dependency on dirent.h on Windows
*
* Revision 1.2 2008/11/02 16:09:27 scottmac
* Update libmagic to 4.26 and add support for v6 of the magic file format.
*
* Revision 1.1 2008/07/11 14:13:50 derick
* - Move lib to libmagic
*
* Revision 1.1 2008/07/11 14:10:50 derick
* - Step one for bundling the libmagic library. Some config.m4 issues left.
*
* Revision 1.69 2008/07/02 15:27:05 christos
* welcome to 4.25
*
* Revision 1.68 2008/03/22 21:39:43 christos
* file 4.24
*
* Revision 1.67 2007/12/28 20:08:40 christos
* welcome to 4.23.
*
* Revision 1.66 2007/12/27 16:38:24 christos
* welcome to 4.22
*
* Revision 1.65 2007/05/24 17:22:27 christos
* Welcome to 4.21
*
* Revision 1.64 2007/03/01 22:14:55 christos
* welcome to 4.20
*
* Revision 1.63 2007/01/12 17:38:28 christos
* Use File id.
*
* Revision 1.62 2006/12/11 21:49:58 christos
* time for 4.19
*
* Revision 1.61 2006/10/31 21:18:09 christos
* bump
*
* Revision 1.60 2006/03/02 22:15:12 christos
* welcome to 4.17
*
* Revision 1.59 2005/10/17 17:15:21 christos
* welcome to 4.16
*
* Revision 1.58 2005/08/18 15:52:56 christos
* welcome to 4.15
*
* Revision 1.57 2005/06/25 15:52:14 christos
* Welcome to 4.14
*
* Revision 1.56 2005/02/09 19:25:13 christos
* Welcome to 4.13
*
* Revision 1.55 2004/11/24 18:57:47 christos
* Re-do the autoconf stuff once more; passes make dist now.
*
* Revision 1.54 2004/11/21 05:52:05 christos
* ready for 4.11
*
* Revision 1.53 2004/07/24 20:40:46 christos
* welcome to 4.10
*
* Revision 1.52 2004/04/07 00:32:25 christos
* welcome to 4.09
*
* Revision 1.51 2004/03/22 21:17:11 christos
* welcome to 4.08.
*
* Revision 1.50 2003/12/23 17:34:04 christos
* 4.07
*
* Revision 1.49 2003/10/15 02:08:27 christos
* welcome to 4.06
*
* Revision 1.48 2003/09/12 19:41:14 christos
* this is 4.04
*
* Revision 1.47 2003/05/23 21:38:21 christos
* welcome to 4.03
*
* Revision 1.46 2003/04/02 18:57:43 christos
* prepare for 4.02
*
* Revision 1.45 2003/03/26 15:37:25 christos
* - Pass lint
* - make NULL in magic_file mean stdin
* - Fix "-" argument to file to pass NULL to magic_file
* - avoid pointer casts by using memcpy
* - rename magic_buf -> magic_buffer
* - keep only the first error
* - manual page: new sentence, new line
* - fix typo in api function (magic_buf -> magic_buffer)
*
* Revision 1.44 2003/03/23 22:23:31 christos
* finish librarification.
*
* Revision 1.43 2003/03/23 21:16:26 christos
* update copyrights.
*
* Revision 1.42 2003/03/23 04:06:05 christos
* Library re-organization
*
* Revision 1.41 2003/02/27 20:53:45 christos
* - fix memory allocation problem (Jeff Johnson)
* - fix stack overflow corruption (David Endler)
* - fixes from NetBSD source (Antti Kantee)
* - magic fixes
*
* Revision 1.40 2003/02/08 18:33:53 christos
* - detect inttypes.h too (Dave Love <d.love@dl.ac.uk>)
* - eliminate unsigned char warnings (Petter Reinholdtsen <pere@hungry.com>)
* - better elf PT_NOTE handling (Nalin Dahyabhai <nalin@redhat.com>)
* - add options to format the output differently
* - much more magic.
*
* Revision 1.39 2002/07/03 18:57:52 christos
* - ansify/c99ize
* - more magic
* - better COMPILE_ONLY support.
* - new magic files.
* - fix solaris compilation problems.
*
* Revision 1.38 2002/05/16 18:45:56 christos
* - pt_note elf additions from NetBSD
* - EMX os specific changes (Alexander Mai)
* - stdint.h detection, acconfig.h fixes (Maciej W. Rozycki, Franz Korntner)
* - regex file additions (Kim Cromie)
* - getopt_long support and misc cleanups (Michael Piefel)
* - many magic fixes and additions
*
* Revision 1.37 2001/09/03 14:44:22 christos
* daylight/tm_isdst detection
* magic fixes
* don't eat the whole file if it has only nulls
*
* Revision 1.36 2001/07/22 21:04:15 christos
* - magic fixes
* - add new operators, pascal strings, UTC date printing, $HOME/.magic
* [from "Tom N Harris" <telliamed@mac.com>]
*
* Revision 1.35 2001/04/24 14:40:25 christos
* - rename magic file sgi to mips and fix it
* - add support for building magic.mgc
* - portability fixes for mmap()
* - try gzip before uncompress, because uncompress sometimes hangs
* - be more conservative about pipe reads and writes
* - many magic fixes
*
* Revision 1.34 2001/03/12 05:05:57 christos
* - new compiled magic format
* - lots of magic additions
*
* Revision 1.33 2000/11/13 00:30:50 christos
* - wordperfect magic fix: freebsd pr 9388
* - more msdos fixes from freebsd pr's 20131 and 20812
* - sas and spss magic [Bruce Foster]
* - mkinstalldirs [John Fremlin]
* - sgi opengl fixes [Michael Pruett]
* - netbsd magic fixes [Ignatios Souvatzis]
* - audio additions [Michael Pruett]
* - fix problem with non ansi RCSID [Andreas Ley]
* - oggs magic [Felix von Leitner]
* - gmon magic [Eugen Dedu]
* - TNEF magic [Joomy]
* - netpbm magic and misc other image stuff [Bryan Henderson]
*
* Revision 1.32 2000/08/05 18:24:18 christos
* Correct indianness detection in elf (Charles Hannum)
* FreeBSD elf core support (Guy Harris)
* Use gzip in systems that don't have uncompress (Anthon van der Neut)
* Internationalization/EBCDIC support (Eric Fisher)
* Many many magic changes
*
* Revision 1.31 2000/05/14 17:58:36 christos
* - new magic for claris files
* - new magic for mathematica and maple files
* - new magic for msvc files
* - new -k flag to keep going matching all possible entries
* - add the word executable on #! magic files, and fix the usage of
* the word script
* - lots of other magic fixes
* - fix typo test -> text
*
* Revision 1.30 2000/04/11 02:41:17 christos
* - add support for mime output (-i)
* - make sure we free memory in case realloc fails
* - magic fixes
*
* Revision 1.29 1999/11/28 20:02:29 christos
* new string/[Bcb] magic from anthon, and adjustments to the magic files to
* use it.
*
* Revision 1.28 1999/10/31 22:11:48 christos
* - add "char" type for compatibility with HP/UX
* - recognize HP/UX syntax &=n etc.
* - include errno.h for CYGWIN
* - conditionalize the S_IS* macros
* - revert the SHT_DYNSYM test that broke the linux stripped binaries test
* - lots of Magdir changes
*
* Revision 1.27 1999/02/14 17:21:41 christos
* Automake support and misc cleanups from Rainer Orth
* Enable reading character and block special files from Dale R. Worley
*
* Revision 1.26 1998/09/12 13:19:39 christos
* - add support for bi-endian indirect offsets (Richard Verhoeven)
* - add recognition for bcpl (Joseph Myers)
* - remove non magic files from Magdir to avoid difficulties building
* on os2 where files are case independent
* - magic fixes.
*
* Revision 1.25 1998/06/27 14:04:04 christos
* OLF patch Guy Harris
* Recognize java/html (debian linux)
* Const poisoning (debian linux)
* More magic!
*
* Revision 1.24 1998/02/15 23:20:38 christos
* Autoconf patch: Felix von Leitner <leitner@math.fu-berlin.de>
* More magic fixes
* Elf64 fixes
*
* Revision 1.23 1997/11/05 16:03:37 christos
* - correct elf prps offset for SunOS-2.5.1 [guy@netapp.com]
* - handle 64 bit time_t's correctly [ewt@redhat.com]
* - new mime style magic [clarosse@netvista.net]
* - new TI calculator magic [rmcguire@freenet.columbus.oh.us]
* - new figlet fonts [obrien@freebsd.org]
* - new cisco magic, and elf fixes [jhawk@bbnplanet.com]
* - -b flag addition, and x86 filesystem magic [vax@linkhead.paranoia.com]
* - s/Mpeg/MPEG, header and elf typo fixes [guy@netapp.com]
* - Windows/NT registry files, audio code [guy@netapp.com]
* - libGrx graphics lib fonts [guy@netapp.com]
* - PNG fixes [guy@netapp.com]
* - more m$ document magic [guy@netapp.com]
* - PPD files [guy@netapp.com]
* - archive magic cleanup [guy@netapp.com]
* - linux kernel magic cleanup [guy@netapp.com]
* - lecter magic [guy@netapp.com]
* - vgetty magic [guy@netapp.com]
* - sniffer additions [guy@netapp.com]
*
* Revision 1.22 1997/01/15 17:23:24 christos
* - add support for elf core files: find the program name under SVR4 [Ken Pizzini]
* - print strings only up to the first carriage return [various]
* - freebsd international ascii support [J Wunsch]
* - magic fixes and additions [Guy Harris]
* - 64 bit fixes [Larry Schwimmer]
* - support for both utime and utimes, but don't restore file access times
* by default [various]
* - \xXX only takes 2 hex digits, not 3.
* - re-implement support for core files [Guy Harris]
*
* Revision 1.21 1996/10/05 18:15:29 christos
* Segregate elf stuff and conditionally enable it with -DBUILTIN_ELF
* More magic fixes
*
* Revision 1.20 1996/06/22 22:15:52 christos
* - support relative offsets of the form >&
* - fix bug with truncating magic strings that contain \n
* - file -f - did not read from stdin as documented
* - support elf file parsing using our own elf support.
* - as always magdir fixes and additions.
*
* Revision 1.19 1995/10/27 23:14:46 christos
* Ability to parse colon separated list of magic files
* New LEGAL.NOTICE
* Various magic file changes
*
* Revision 1.18 1995/05/20 22:09:21 christos
* Passed incorrect argument to eatsize().
* Use %ld and %lx where appropriate.
* Remove unused variables
* ELF support for both big and little endian
* Fixes for small files again.
*
* Revision 1.17 1995/04/28 17:29:13 christos
* - Incorrect nroff detection fix from der Mouse
* - Lost and incorrect magic entries.
* - Added ELF stripped binary detection [in C; ugh]
* - Look for $MAGIC to find the magic file.
* - Eat trailing size specifications from numbers i.e. ignore 10L
* - More fixes for very short files
*
* Revision 1.16 1995/03/25 22:06:45 christos
* - use strtoul() where it exists.
* - fix sign-extend bug
* - try to detect tar archives before nroff files, otherwise
* tar files where the first file starts with a . will not work
*
* Revision 1.15 1995/01/21 21:03:35 christos
* Added CSECTION for the file man page
* Added version flag -v
* Fixed bug with -f input flag (from iorio@violet.berkeley.edu)
* Lots of magic fixes and reorganization...
*
* Revision 1.14 1994/05/03 17:58:23 christos
* changes from mycroft@gnu.ai.mit.edu (Charles Hannum) for unsigned
*
* Revision 1.13 1994/01/21 01:27:01 christos
* Fixed null termination bug from Don Seeley at BSDI in ascmagic.c
*
* Revision 1.12 1993/10/27 20:59:05 christos
* Changed -z flag to understand gzip format too.
* Moved builtin compression detection to a table, and move
* the compress magic entry out of the source.
* Made printing of numbers unsigned, and added the mask to it.
* Changed the buffer size to 8k, because gzip will refuse to
* unzip just a few bytes.
*
* Revision 1.11 1993/09/24 18:49:06 christos
* Fixed small bug in softmagic.c introduced by
* copying the data to be examined out of the input
* buffer. Changed the Makefile to use sed to create
* the correct man pages.
*
* Revision 1.10 1993/09/23 21:56:23 christos
* Passed purify. Fixed indirections. Fixed byte order printing.
* Fixed segmentation faults caused by referencing past the end
* of the magic buffer. Fixed bus errors caused by referencing
* unaligned shorts or longs.
*
* Revision 1.9 1993/03/24 14:23:40 ian
* Batch of minor changes from several contributors.
*
* Revision 1.8 93/02/19 15:01:26 ian
* Numerous changes from Guy Harris too numerous to mention but including
* byte-order independance, fixing "old-style masking", etc. etc. A bugfix
* for broken symlinks from martin@@d255s004.zfe.siemens.de.
*
* Revision 1.7 93/01/05 14:57:27 ian
* Couple of nits picked by Christos (again, thanks).
*
* Revision 1.6 93/01/05 13:51:09 ian
* Lotsa work on the Magic directory.
*
* Revision 1.5 92/09/14 14:54:51 ian
* Fix a tiny null-pointer bug in previous fix for tar archive + uncompress.
*
*/

220
ext/fileinfo/libmagic/softmagic.c
View File

@ -32,7 +32,7 @@
#include "file.h" #include "file.h"
#ifndef lint #ifndef lint
FILE_RCSID("@(#)$File: softmagic.c,v 1.299 2020/06/07 21:58:01 christos Exp $") FILE_RCSID("@(#)$File: softmagic.c,v 1.309 2021/02/05 22:29:07 christos Exp $")
#endif /* lint */ #endif /* lint */
#include "magic.h" #include "magic.h"
@ -169,6 +169,8 @@ file_fmtcheck(struct magic_set *ms, const char *desc, const char *def,
#define F(a, b, c) ((b)) #define F(a, b, c) ((b))
#endif #endif
/* NOTE this function has been kept an the state of 5.39 for BC. Observe
* further as the upgrade to 5.41 or above goes. */
/* /*
* Go through the whole list, stopping if you find a match. Process all * Go through the whole list, stopping if you find a match. Process all
* the continuations of that match before returning. * the continuations of that match before returning.
@ -498,6 +500,28 @@ check_fmt(struct magic_set *ms, const char *fmt)
return rv; return rv;
} }
#if !defined(HAVE_STRNDUP) || defined(__aiws__) || defined(_AIX)
# if defined(__aiws__) || defined(_AIX)
# define strndup aix_strndup /* aix is broken */
# endif
char *strndup(const char *, size_t);
char *
strndup(const char *str, size_t n)
{
size_t len;
char *copy;
for (len = 0; len < n && str[len]; len++)
continue;
if ((copy = malloc(len + 1)) == NULL)
return NULL;
(void)memcpy(copy, str, len);
copy[len] = '\0';
return copy;
}
#endif /* HAVE_STRNDUP */
static int static int
varexpand(struct magic_set *ms, char *buf, size_t len, const char *str) varexpand(struct magic_set *ms, char *buf, size_t len, const char *str)
{ {
@ -569,93 +593,58 @@ mprint(struct magic_set *ms, struct magic *m)
else else
desc = ebuf; desc = ebuf;
#define PRINTER(value, format, stype, utype) \
v = file_signextend(ms, m, CAST(uint64_t, value)); \
switch (check_fmt(ms, desc)) { \
case -1: \
return -1; \
case 1: \
if (m->flag & UNSIGNED) { \
(void)snprintf(buf, sizeof(buf), "%" format "u", \
CAST(utype, v)); \
} else { \
(void)snprintf(buf, sizeof(buf), "%" format "d", \
CAST(stype, v)); \
} \
if (file_printf(ms, F(ms, desc, "%s"), buf) == -1) \
return -1; \
break; \
default: \
if (m->flag & UNSIGNED) { \
if (file_printf(ms, F(ms, desc, "%" format "u"), \
CAST(utype, v)) == -1) \
return -1; \
} else { \
if (file_printf(ms, F(ms, desc, "%" format "d"), \
CAST(stype, v)) == -1) \
return -1; \
} \
break; \
} \
t = ms->offset + sizeof(stype); \
break
switch (m->type) { switch (m->type) {
case FILE_BYTE: case FILE_BYTE:
v = file_signextend(ms, m, CAST(uint64_t, p->b)); PRINTER(p->b, "", int8_t, uint8_t);
switch (check_fmt(ms, desc)) {
case -1:
return -1;
case 1:
(void)snprintf(buf, sizeof(buf), "%d",
CAST(unsigned char, v));
if (file_printf(ms, F(ms, desc, "%s"), buf) == -1)
return -1;
break;
default:
if (file_printf(ms, F(ms, desc, "%d"),
CAST(unsigned char, v)) == -1)
return -1;
break;
}
t = ms->offset + sizeof(char);
break;
case FILE_SHORT: case FILE_SHORT:
case FILE_BESHORT: case FILE_BESHORT:
case FILE_LESHORT: case FILE_LESHORT:
v = file_signextend(ms, m, CAST(uint64_t, p->h)); PRINTER(p->h, "", int16_t, uint16_t);
switch (check_fmt(ms, desc)) {
case -1:
return -1;
case 1:
(void)snprintf(buf, sizeof(buf), "%u",
CAST(unsigned short, v));
if (file_printf(ms, F(ms, desc, "%s"), buf) == -1)
return -1;
break;
default:
if (file_printf(ms, F(ms, desc, "%u"),
CAST(unsigned short, v)) == -1)
return -1;
break;
}
t = ms->offset + sizeof(short);
break;
case FILE_LONG: case FILE_LONG:
case FILE_BELONG: case FILE_BELONG:
case FILE_LELONG: case FILE_LELONG:
case FILE_MELONG: case FILE_MELONG:
v = file_signextend(ms, m, CAST(uint64_t, p->l)); PRINTER(p->l, "", int32_t, uint32_t);
switch (check_fmt(ms, desc)) {
case -1:
return -1;
case 1:
(void)snprintf(buf, sizeof(buf), "%u",
CAST(uint32_t, v));
if (file_printf(ms, F(ms, desc, "%s"), buf) == -1)
return -1;
break;
default:
if (file_printf(ms, F(ms, desc, "%u"),
CAST(uint32_t, v)) == -1)
return -1;
break;
}
t = ms->offset + sizeof(int32_t);
break; break;
case FILE_QUAD: case FILE_QUAD:
case FILE_BEQUAD: case FILE_BEQUAD:
case FILE_LEQUAD: case FILE_LEQUAD:
case FILE_OFFSET: case FILE_OFFSET:
v = file_signextend(ms, m, p->q); PRINTER(p->q, INT64_T_FORMAT, long long, unsigned long long);
switch (check_fmt(ms, desc)) {
case -1:
return -1;
case 1:
(void)snprintf(buf, sizeof(buf), "%" INT64_T_FORMAT "u",
CAST(unsigned long long, v));
if (file_printf(ms, F(ms, desc, "%s"), buf) == -1)
return -1;
break;
default:
if (file_printf(ms, F(ms, desc, "%" INT64_T_FORMAT "u"),
CAST(unsigned long long, v)) == -1)
return -1;
break;
}
t = ms->offset + sizeof(int64_t);
break; break;
case FILE_STRING: case FILE_STRING:
@ -678,19 +667,9 @@ mprint(struct magic_set *ms, struct magic *m)
if (*m->value.s == '\0') if (*m->value.s == '\0')
str[strcspn(str, "\r\n")] = '\0'; str[strcspn(str, "\r\n")] = '\0';
if (m->str_flags & STRING_TRIM) { if (m->str_flags & STRING_TRIM)
char *last; str = file_strtrim(str);
while (isspace(CAST(unsigned char, *str)))
str++;
last = str;
while (*last)
last++;
--last;
while (isspace(CAST(unsigned char, *last)))
last--;
*++last = '\0';
}
if (file_printf(ms, F(ms, desc, "%s"), if (file_printf(ms, F(ms, desc, "%s"),
file_printable(sbuf, sizeof(sbuf), str, file_printable(sbuf, sizeof(sbuf), str,
sizeof(p->s) - (str - p->s))) == -1) sizeof(p->s) - (str - p->s))) == -1)
@ -795,14 +774,20 @@ mprint(struct magic_set *ms, struct magic *m)
case FILE_SEARCH: case FILE_SEARCH:
case FILE_REGEX: { case FILE_REGEX: {
char *cp; char *cp, *scp;
int rval; int rval;
cp = estrndup(RCAST(const char *, ms->search.s), cp = strndup(RCAST(const char *, ms->search.s),
ms->search.rm_len); ms->search.rm_len);
if (cp == NULL) {
file_oomem(ms, ms->search.rm_len);
return -1;
}
scp = (m->str_flags & STRING_TRIM) ? file_strtrim(cp) : cp;
rval = file_printf(ms, F(ms, desc, "%s"), rval = file_printf(ms, F(ms, desc, "%s"),
file_printable(sbuf, sizeof(sbuf), cp, ms->search.rm_len)); file_printable(sbuf, sizeof(sbuf), scp, ms->search.rm_len));
efree(cp); free(cp);
if (rval == -1) if (rval == -1)
return -1; return -1;
@ -955,6 +940,7 @@ moffset(struct magic_set *ms, struct magic *m, const struct buffer *b,
case FILE_DEFAULT: case FILE_DEFAULT:
case FILE_INDIRECT: case FILE_INDIRECT:
case FILE_OFFSET: case FILE_OFFSET:
case FILE_USE:
o = ms->offset; o = ms->offset;
break; break;
@ -1541,6 +1527,28 @@ normal:
return 0; return 0;
} }
private int
save_cont(struct magic_set *ms, struct cont *c)
{
size_t len;
*c = ms->c;
len = c->len * sizeof(*c->li);
ms->c.li = CAST(struct level_info *, malloc(len));
if (ms->c.li == NULL) {
ms->c = *c;
return -1;
}
memcpy(ms->c.li, c->li, len);
return 0;
}
private void
restore_cont(struct magic_set *ms, struct cont *c)
{
free(ms->c.li);
ms->c = *c;
}
private int private int
mget(struct magic_set *ms, struct magic *m, const struct buffer *b, mget(struct magic_set *ms, struct magic *m, const struct buffer *b,
const unsigned char *s, size_t nbytes, size_t o, unsigned int cont_level, const unsigned char *s, size_t nbytes, size_t o, unsigned int cont_level,
@ -1548,14 +1556,15 @@ mget(struct magic_set *ms, struct magic *m, const struct buffer *b,
int *printed_something, int *need_separator, int *returnval, int *printed_something, int *need_separator, int *returnval,
int *found_match) int *found_match)
{ {
uint32_t offset = ms->offset; uint32_t eoffset, offset = ms->offset;
struct buffer bb; struct buffer bb;
intmax_t lhs; intmax_t lhs;
file_pushbuf_t *pb; file_pushbuf_t *pb;
int rv, oneed_separator, in_type; int rv, oneed_separator, in_type, nfound_match;
char *rbuf; char *rbuf;
union VALUETYPE *p = &ms->ms_value; union VALUETYPE *p = &ms->ms_value;
struct mlist ml; struct mlist ml;
struct cont c;
if (*indir_count >= ms->indir_max) { if (*indir_count >= ms->indir_max) {
file_error(ms, 0, "indirect count (%hu) exceeded", file_error(ms, 0, "indirect count (%hu) exceeded",
@ -1836,7 +1845,8 @@ mget(struct magic_set *ms, struct magic *m, const struct buffer *b,
if (rv == 1) { if (rv == 1) {
if ((ms->flags & MAGIC_NODESC) == 0 && if ((ms->flags & MAGIC_NODESC) == 0 &&
file_printf(ms, F(ms, m->desc, "%u"), offset) == -1) { file_printf(ms, F(ms, m->desc, "%u"), offset) == -1)
{
if (rbuf) efree(rbuf); if (rbuf) efree(rbuf);
return -1; return -1;
} }
@ -1860,16 +1870,32 @@ mget(struct magic_set *ms, struct magic *m, const struct buffer *b,
file_error(ms, 0, "cannot find entry `%s'", rbuf); file_error(ms, 0, "cannot find entry `%s'", rbuf);
return -1; return -1;
} }
(*name_count)++; if (save_cont(ms, &c) == -1) {
file_error(ms, errno, "can't allocate continuation");
return -1;
}
oneed_separator = *need_separator; oneed_separator = *need_separator;
if (m->flag & NOSPACE) if (m->flag & NOSPACE)
*need_separator = 0; *need_separator = 0;
nfound_match = 0;
(*name_count)++;
eoffset = ms->eoffset;
rv = match(ms, ml.magic, ml.nmagic, b, offset + o, rv = match(ms, ml.magic, ml.nmagic, b, offset + o,
mode, text, flip, indir_count, name_count, mode, text, flip, indir_count, name_count,
printed_something, need_separator, returnval, found_match); printed_something, need_separator, returnval,
&nfound_match);
ms->ms_value.q = nfound_match;
(*name_count)--; (*name_count)--;
*found_match |= nfound_match;
restore_cont(ms, &c);
if (rv != 1) if (rv != 1)
*need_separator = oneed_separator; *need_separator = oneed_separator;
ms->offset = offset;
ms->eoffset = eoffset;
return rv; return rv;
case FILE_NAME: case FILE_NAME:
@ -1934,13 +1960,10 @@ file_strncmp(const char *s1, const char *s2, size_t len, size_t maxlen,
} }
else if ((flags & STRING_COMPACT_WHITESPACE) && else if ((flags & STRING_COMPACT_WHITESPACE) &&
isspace(*a)) { isspace(*a)) {
/* XXX Dirty. The data and the pattern is what is causing this.
Revert _i for the next port and see if it still matters. */
uint32_t _i = 0;
a++; a++;
if (isspace(*b++)) { if (isspace(*b++)) {
if (!isspace(*a)) if (!isspace(*a))
while (EXPECTED(_i++ < 2048) && b < eb && isspace(*b)) while (b < eb && isspace(*b))
b++; b++;
} }
else { else {
@ -2282,9 +2305,10 @@ error_out:
} }
break; break;
} }
case FILE_INDIRECT:
case FILE_USE: case FILE_USE:
return ms->ms_value.q != 0;
case FILE_NAME: case FILE_NAME:
case FILE_INDIRECT:
return 1; return 1;
case FILE_DER: case FILE_DER:
matched = der_cmp(ms, m); matched = der_cmp(ms, m);

117
ext/fileinfo/magicdata.patch
View File

@ -1,99 +1,22 @@
diff -u magic.orig/Magdir/images magic/Magdir/images diff -ur Magdir.orig/mail.news Magdir/mail.news
--- magic.orig/Magdir/images 2020-05-31 12:34:40.000000000 +0200 --- Magdir.orig/mail.news 2021-03-31 01:47:28.000000000 +0200
+++ magic/Magdir/images 2020-07-05 20:00:41.664783368 +0200 +++ Magdir/mail.news 2021-04-05 19:41:55.168556972 +0200
@@ -1,6 +1,6 @@ @@ -1,5 +1,5 @@
#------------------------------------------------------------------------------ #------------------------------------------------------------------------------
-# $File: images,v 1.181 2020/05/30 23:49:03 christos Exp $ -# $File: mail.news,v 1.26 2021/03/21 14:37:03 christos Exp $
+# $File: images,v 1.183 2020/06/26 17:08:32 christos Exp $ +# $File: mail.news,v 1.27 2021/04/05 16:36:14 christos Exp $
# images: file(1) magic for image formats (see also "iff", and "c-lang" for # mail.news: file(1) magic for mail and news
# XPM bitmaps)
# #
@@ -32,22 +32,22 @@ # Unfortunately, saved netnews also has From line added in some news software.
# Prevent conflicts with CRI ADX. @@ -81,4 +81,4 @@
>(2.S-2) belong !0x28632943 # File format spec: https://wiki.dovecot.org/Design/Dcrypt/#File_format
# skip more garbage like *.iso by looking for positive image type # From: Stephen Gildea
->>2 ubyte >0 0 string CRYPTED\003\007 Dovecot encrypted message
+>2 ubyte >0 ->9 byte xu \b, dcrypt version %d
# skip some compiled terminfo like xterm+tmux by looking for image type less equal 33 +>9 byte x \b, dcrypt version %d
->>>2 ubyte <34 diff -ur Magdir.orig/rpm Magdir/rpm
+>>2 ubyte <34 --- Magdir.orig/rpm 2021-02-23 01:49:24.000000000 +0100
# skip arches.3200 , Finder.Root , Slp.1 by looking for low pixel depth 1 8 15 16 24 32 +++ Magdir/rpm 2021-04-05 19:40:55.080911893 +0200
->>>>16 ubyte 1
->>>>>0 use tga-image
->>>>16 ubyte 8
->>>>>0 use tga-image
->>>>16 ubyte 15
->>>>>0 use tga-image
->>>>16 ubyte 16
->>>>>0 use tga-image
->>>>16 ubyte 24
->>>>>0 use tga-image
->>>>16 ubyte 32
->>>>>0 use tga-image
+>>>16 ubyte 1
+>>>>0 use tga-image
+>>>16 ubyte 8
+>>>>0 use tga-image
+>>>16 ubyte 15
+>>>>0 use tga-image
+>>>16 ubyte 16
+>>>>0 use tga-image
+>>>16 ubyte 24
+>>>>0 use tga-image
+>>>16 ubyte 32
+>>>>0 use tga-image
# display tga bitmap image information
0 name tga-image
>2 ubyte <34 Targa image data
@@ -615,7 +615,7 @@
0 leshort 40
# skip bad samples like GAME by looking for valid number of color planes
>12 uleshort 1 Device independent bitmap graphic
-!:mime image/bmp
+!:mime image/x-ms-bmp
!:apple ????BMPp
!:ext dib
>>4 lelong x \b, %d x
@@ -641,7 +641,7 @@
>>18 leshort x \b, %d x
>>20 leshort x %d
>14 leshort 64 PC bitmap, OS/2 2.x format
-!:mime image/bmp
+!:mime image/x-ms-bmp
!:apple ????BMPp
!:ext bmp
# image width and height fields are unsigned integers for OS/2
@@ -662,7 +662,7 @@
#>>(10.l) ubequad !0 \b, bits 0x%16.16llx
# BITMAPV2INFOHEADER adds RGB bit masks
>14 leshort 52 PC bitmap, Adobe Photoshop
-!:mime image/bmp
+!:mime image/x-ms-bmp
!:apple ????BMPp
!:ext bmp
>>18 lelong x \b, %d x
@@ -670,7 +670,7 @@
>>28 leshort x %d
# BITMAPV3INFOHEADER adds alpha channel bit mask
>14 leshort 56 PC bitmap, Adobe Photoshop with alpha channel mask
-!:mime image/bmp
+!:mime image/x-ms-bmp
!:apple ????BMPp
!:ext bmp
>>18 lelong x \b, %d x
@@ -679,7 +679,7 @@
>14 leshort 40
# jump 4 bytes before end of file/header to skip fmt-116-signature-id-118.dib
>>(2.l-4) ulong x PC bitmap, Windows 3.x format
-!:mime image/bmp
+!:mime image/x-ms-bmp
!:apple ????BMPp
>>>18 lelong x \b, %d x
>>>22 lelong x %d
diff -u magic.orig/Magdir/rpm magic/Magdir/rpm
--- magic.orig/Magdir/rpm 2019-02-22 14:06:34.000000000 +0100
+++ magic/Magdir/rpm 2020-07-05 19:38:02.720419674 +0200
@@ -29,6 +29,7 @@ @@ -29,6 +29,7 @@
>>8 beshort 17 SuperH >>8 beshort 17 SuperH
>>8 beshort 18 Xtensa >>8 beshort 18 Xtensa
@ -102,9 +25,9 @@ diff -u magic.orig/Magdir/rpm magic/Magdir/rpm
#delta RPM Daniel Novotny (dnovotny@redhat.com) #delta RPM Daniel Novotny (dnovotny@redhat.com)
0 string drpm Delta RPM 0 string drpm Delta RPM
diff -u magic.orig/Magdir/securitycerts magic/Magdir/securitycerts diff -ur Magdir.orig/securitycerts Magdir/securitycerts
--- magic.orig/Magdir/securitycerts 2019-02-22 14:06:34.000000000 +0100 --- Magdir.orig/securitycerts 2021-02-23 01:49:24.000000000 +0100
+++ magic/Magdir/securitycerts 2020-07-05 19:38:02.720419674 +0200 +++ Magdir/securitycerts 2021-04-05 19:40:55.080911893 +0200
@@ -4,3 +4,5 @@ @@ -4,3 +4,5 @@
0 search/1 -----BEGIN\ CERTIFICATE------ RFC1421 Security Certificate text 0 search/1 -----BEGIN\ CERTIFICATE------ RFC1421 Security Certificate text
0 search/1 -----BEGIN\ NEW\ CERTIFICATE RFC1421 Security Certificate Signing Request text 0 search/1 -----BEGIN\ NEW\ CERTIFICATE RFC1421 Security Certificate Signing Request text

2
ext/fileinfo/tests/bug68819_001.phpt
View File

@ -15,4 +15,4 @@ $type = $finfo->buffer($string);
var_dump($type); var_dump($type);
?> ?>
--EXPECT-- --EXPECT--
string(60) "ASCII text, with very long lines, with CRLF line terminators" string(66) "ASCII text, with very long lines (617), with CRLF line terminators"

2
ext/fileinfo/tests/bug68819_002.phpt
View File

@ -23,4 +23,4 @@ var_dump($type);
?> ?>
--EXPECT-- --EXPECT--
string(60) "ASCII text, with very long lines, with CRLF line terminators" string(67) "ASCII text, with very long lines (8191), with CRLF line terminators"

2
ext/fileinfo/tests/finfo_file_basic.phpt
View File

@ -24,5 +24,5 @@ try {
*** Testing finfo_file() : basic functionality *** *** Testing finfo_file() : basic functionality ***
string(28) "text/x-php; charset=us-ascii" string(28) "text/x-php; charset=us-ascii"
string(22) "PHP script, ASCII text" string(22) "PHP script, ASCII text"
string(25) "text/plain; charset=utf-8" string(28) "text/plain; charset=us-ascii"
finfo_file(): Argument #1 ($finfo) must not contain any null bytes finfo_file(): Argument #1 ($finfo) must not contain any null bytes

2886
ext/fileinfo/tests/magic
View File

File diff suppressed because it is too large Load Diff

2886
ext/fileinfo/tests/magic私はガラスを食べられます
View File

File diff suppressed because it is too large Load Diff