Add support for hackish ucs16 strings.

diff --git a/src/apprentice.c b/src/apprentice.c
index 38f9d51edbd8c825e5bc7d73436db2ad460235bc..08b686361deae2131e5b3ef0a503628db92414c7 100644 (file)
--- a/src/apprentice.c
+++ b/src/apprentice.c
@@ -45,7 +45,7 @@
 #endif
 
 #ifndef        lint
-FILE_RCSID("@(#)$Id: apprentice.c,v 1.80 2004/11/13 08:11:39 christos Exp $")
+FILE_RCSID("@(#)$Id: apprentice.c,v 1.81 2004/11/20 23:50:12 christos Exp $")
 #endif /* lint */
 
 #define        EATAB {while (isascii((unsigned char) *l) && \
@@ -74,6 +74,9 @@ FILE_RCSID("@(#)$Id: apprentice.c,v 1.80 2004/11/13 08:11:39 christos Exp $")
 #define MAXPATHLEN     1024
 #endif
 
+#define IS_STRING(t) ((t) == FILE_STRING || (t) == FILE_PSTRING || \
+    (t) == FILE_BESTRING16 || (t) == FILE_LESTRING16)
+
 private int getvalue(struct magic_set *ms, struct magic *, char **);
 private int hextoint(int);
 private char *getstr(struct magic_set *, char *, char *, int, int *);
@@ -370,6 +373,8 @@ file_signextend(struct magic_set *ms, struct magic *m, uint32_t v)
                        break;
                case FILE_STRING:
                case FILE_PSTRING:
+               case FILE_BESTRING16:
+               case FILE_LESTRING16:
                        break;
                case FILE_REGEX:
                        break;
@@ -542,6 +547,8 @@ parse(struct magic_set *ms, struct magic **magicp, uint32_t *nmagicp, char *l,
 #define NBELDATE       7
 #define NLELDATE       7
 #define NREGEX         5
+#define NBESTRING16    10
+#define NLESTRING16    10
 
        if (*l == 'u') {
                ++l;
@@ -599,7 +606,13 @@ parse(struct magic_set *ms, struct magic **magicp, uint32_t *nmagicp, char *l,
                l += NLELDATE;
        } else if (strncmp(l, "regex", NREGEX)==0) {
                m->type = FILE_REGEX;
-               l += sizeof("regex");
+               l += NREGEX;
+       } else if (strncmp(l, "bestring16", NBESTRING16)==0) {
+               m->type = FILE_BESTRING16;
+               l += NBESTRING16;
+       } else if (strncmp(l, "lestring16", NLESTRING16)==0) {
+               m->type = FILE_LESTRING16;
+               l += NLESTRING16;
        } else {
                if (ms->flags & MAGIC_CHECK)
                        file_magwarn(ms, "type `%s' invalid", l);
@@ -608,14 +621,13 @@ parse(struct magic_set *ms, struct magic **magicp, uint32_t *nmagicp, char *l,
        /* New-style anding: "0 byte&0x80 =0x80 dynamically linked" */
        /* New and improved: ~ & | ^ + - * / % -- exciting, isn't it? */
        if (*l == '~') {
-               if (FILE_STRING != m->type && FILE_PSTRING != m->type)
+               if (!IS_STRING(m->type))
                        m->mask_op = FILE_OPINVERSE;
                ++l;
        }
        if ((t = strchr(fops,  *l)) != NULL) {
                uint32_t op = (uint32_t)(t - fops);
-               if (op != FILE_OPDIVIDE ||
-                   (FILE_STRING != m->type && FILE_PSTRING != m->type)) {
+               if (op != FILE_OPDIVIDE || !IS_STRING(m->type)) {
                        ++l;
                        m->mask_op |= op;
                        val = (uint32_t)strtoul(l, &l, 0);
@@ -666,7 +678,7 @@ parse(struct magic_set *ms, struct magic **magicp, uint32_t *nmagicp, char *l,
                }
                break;
        case '!':
-               if (m->type != FILE_STRING && m->type != FILE_PSTRING) {
+               if (!IS_STRING(m->type)) {
                        m->reln = *l;
                        ++l;
                        break;
@@ -784,6 +796,8 @@ getvalue(struct magic_set *ms, struct magic *m, char **p)
        int slen;
 
        switch (m->type) {
+       case FILE_BESTRING16:
+       case FILE_LESTRING16:
        case FILE_STRING:
        case FILE_PSTRING:
        case FILE_REGEX:
@@ -1210,7 +1224,7 @@ bs1(struct magic *m)
        m->cont_level = swap2(m->cont_level);
        m->offset = swap4((uint32_t)m->offset);
        m->in_offset = swap4((uint32_t)m->in_offset);
-       if (m->type != FILE_STRING)
+       if (IS_STRING(m->type))
                m->value.l = swap4(m->value.l);
        m->mask = swap4(m->mask);
 }

diff --git a/src/file.h b/src/file.h
index 2f10939944e453c148bee3ec6b42a4ad712b599c..f29bba01afc42e54da3e9b11a8011f4bb8e05b6c 100644 (file)
--- a/src/file.h
+++ b/src/file.h
@@ -27,7 +27,7 @@
  */
 /*
  * file.h - definitions for file(1) program
- * @(#)$Id: file.h,v 1.63 2004/11/13 08:11:39 christos Exp $
+ * @(#)$Id: file.h,v 1.64 2004/11/20 23:50:12 christos Exp $
  */
 
 #ifndef __file_h__
@@ -108,6 +108,8 @@ struct magic {
 #define                                FILE_BELDATE    15
 #define                                FILE_LELDATE    16
 #define                                FILE_REGEX      17
+#define                                FILE_BESTRING16 18
+#define                                FILE_LESTRING16 19
 
 #define                                FILE_FORMAT_NAME        \
 /* 0 */                        "invalid 0",            \
@@ -127,7 +129,9 @@ struct magic {
 /* 14 */                       "ldate",                \
 /* 15 */                       "beldate",              \
 /* 16 */                       "leldate",              \
-/* 17 */                       "regex",
+/* 17 */                       "regex",                \
+/* 18 */                       "bestring16",           \
+/* 19 */                       "lestring16",
 
 #define        FILE_FMT_NUM    "cduxXi"
 #define FILE_FMT_STR   "s"     
@@ -150,7 +154,9 @@ struct magic {
 /* 14 */                       FILE_FMT_STR,           \
 /* 15 */                       FILE_FMT_STR,           \
 /* 16 */                       FILE_FMT_STR,           \
-/* 17 */                       FILE_FMT_STR,
+/* 17 */                       FILE_FMT_STR,           \
+/* 18 */                       FILE_FMT_STR,           \
+/* 19 */                       FILE_FMT_STR,
 
        /* Word 3 */
        uint8_t in_op;          /* operator for indirection */

diff --git a/src/softmagic.c b/src/softmagic.c
index 4fc1af834fa281f05d652ff954e0092aa6ebddd1..3cc0c2cfb6d45b2abee6539ec522f593e5deab23 100644 (file)
--- a/src/softmagic.c
+++ b/src/softmagic.c
@@ -39,7 +39,7 @@
 
 
 #ifndef        lint
-FILE_RCSID("@(#)$Id: softmagic.c,v 1.69 2004/11/13 08:11:12 christos Exp $")
+FILE_RCSID("@(#)$Id: softmagic.c,v 1.70 2004/11/20 23:50:13 christos Exp $")
 #endif /* lint */
 
 private int match(struct magic_set *, struct magic *, uint32_t,
@@ -49,6 +49,8 @@ private int mget(struct magic_set *, union VALUETYPE *, const unsigned char *,
 private int mcheck(struct magic_set *, union VALUETYPE *, struct magic *);
 private int32_t mprint(struct magic_set *, union VALUETYPE *, struct magic *);
 private void mdebug(uint32_t, const char *, size_t);
+private int mcopy(struct magic_set *, union VALUETYPE *, int, int,
+    const unsigned char *, size_t, size_t);
 private int mconvert(struct magic_set *, union VALUETYPE *, struct magic *);
 private int check_mem(struct magic_set *, unsigned int);
 
@@ -271,6 +273,8 @@ mprint(struct magic_set *ms, union VALUETYPE *p, struct magic *m)
 
        case FILE_STRING:
        case FILE_PSTRING:
+       case FILE_BESTRING16:
+       case FILE_LESTRING16:
                if (m->reln == '=') {
                        if (file_printf(ms, m->desc, m->value.s) == -1)
                                return -1;
@@ -421,6 +425,8 @@ mconvert(struct magic_set *ms, union VALUETYPE *p, struct magic *m)
                        p->l = ~p->l;
                return 1;
        case FILE_STRING:
+       case FILE_BESTRING16:
+       case FILE_LESTRING16:
                {
                        size_t len;
 
@@ -598,12 +604,10 @@ mdebug(uint32_t offset, const char *str, size_t len)
 }
 
 private int
-mget(struct magic_set *ms, union VALUETYPE *p, const unsigned char *s,
-    struct magic *m, size_t nbytes)
+mcopy(struct magic_set *ms, union VALUETYPE *p, int type, int indir,
+    const unsigned char *s, size_t offset, size_t nbytes)
 {
-       uint32_t offset = m->offset;
-
-       if (m->type == FILE_REGEX) {
+       if (type == FILE_REGEX && indir == 0) {
                /*
                 * offset is interpreted as last line to search,
                 * (starting at 1), not as bytes-from start-of-file
@@ -619,18 +623,53 @@ mget(struct magic_set *ms, union VALUETYPE *p, const unsigned char *s,
                        last = b;
                if (last != NULL)
                        *last = '\0';
-       } else if (offset + sizeof(union VALUETYPE) <= nbytes)
-               memcpy(p, s + offset, sizeof(union VALUETYPE));
-       else {
-               /*
-                * the usefulness of padding with zeroes eludes me, it
-                * might even cause problems
-                */
-               memset(p, 0, sizeof(union VALUETYPE));
-               if (offset < nbytes)
-                       memcpy(p, s + offset, nbytes - offset);
+               return 0;
+       }
+
+       if (indir == 0 && (type == FILE_BESTRING16 || type == FILE_LESTRING16))
+       {
+               const char *src = s + offset;
+               const char *esrc = s + nbytes;
+               char *dst = p->s, *edst = &p->s[sizeof(p->s) - 1];
+
+               if (type == FILE_BESTRING16)
+                       src++;
+
+               for (;src < esrc; src++, dst++) {
+                       if (dst < edst)
+                               *dst = *src++;
+                       else
+                               break;
+                       if (*dst == '\0')
+                               *dst = ' ';
+               }
+               *edst = '\0';
+               return 0;
        }
 
+       if (offset + sizeof(*p) <= nbytes)
+               nbytes = sizeof(*p);
+
+       (void)memcpy(p, s + offset, nbytes);
+
+       /*
+        * the usefulness of padding with zeroes eludes me, it
+        * might even cause problems
+        */
+       if (nbytes < sizeof(*p))
+               (void)memset(p + nbytes, 0, sizeof(*p) - nbytes);
+       return 0;
+}
+
+private int
+mget(struct magic_set *ms, union VALUETYPE *p, const unsigned char *s,
+    struct magic *m, size_t nbytes)
+{
+       uint32_t offset = m->offset;
+
+       if (mcopy(ms, p, m->type, m->flag & INDIR, s, offset, nbytes) == -1)
+               return -1;
+
        /* Verify we have enough data to match magic type */
        switch (m->type) {
                case FILE_BYTE:
@@ -1017,11 +1056,8 @@ mget(struct magic_set *ms, union VALUETYPE *p, const unsigned char *s,
                        break;
                }
 
-               if (nbytes < sizeof(union VALUETYPE) ||
-                   nbytes - sizeof(union VALUETYPE) < offset)
-                       return 0;
-
-               memcpy(p, s + offset, sizeof(union VALUETYPE));
+               if (mcopy(ms, p, m->type, 0, s, offset, nbytes) == -1)
+                       return -1;
 
                if ((ms->flags & MAGIC_DEBUG) != 0) {
                        mdebug(offset, (char *)(void *)p,
@@ -1070,6 +1106,8 @@ mcheck(struct magic_set *ms, union VALUETYPE *p, struct magic *m)
                break;
 
        case FILE_STRING:
+       case FILE_BESTRING16:
+       case FILE_LESTRING16:
        case FILE_PSTRING:
        {
                /*