Annotation of src/usr.bin/file/ascmagic.c, Revision 1.3
1.3 ! millert 1: /* $OpenBSD: ascmagic.c,v 1.2 1996/06/26 05:32:54 deraadt Exp $ */
! 2:
1.1 deraadt 3: /*
4: * ASCII magic -- file types that we know based on keywords
5: * that can appear anywhere in the file.
6: *
7: * Copyright (c) Ian F. Darwin, 1987.
8: * Written by Ian F. Darwin.
9: *
10: * This software is not subject to any license of the American Telephone
11: * and Telegraph Company or of the Regents of the University of California.
12: *
13: * Permission is granted to anyone to use this software for any purpose on
14: * any computer system, and to alter it and redistribute it freely, subject
15: * to the following restrictions:
16: *
17: * 1. The author is not responsible for the consequences of use of this
18: * software, no matter how awful, even if they arise from flaws in it.
19: *
20: * 2. The origin of this software must not be misrepresented, either by
21: * explicit claim or by omission. Since few users ever read sources,
22: * credits must appear in the documentation.
23: *
24: * 3. Altered versions must be plainly marked as such, and must not be
25: * misrepresented as being the original software. Since few users
26: * ever read sources, credits must appear in the documentation.
27: *
28: * 4. This notice may not be removed or altered.
29: */
30:
31: #include <stdio.h>
32: #include <string.h>
33: #include <ctype.h>
34: #include <stdlib.h>
35: #include <unistd.h>
36: #include "file.h"
37: #include "names.h"
38:
39: #ifndef lint
1.3 ! millert 40: static char *moduleid = "$OpenBSD: ascmagic.c,v 1.2 1996/06/26 05:32:54 deraadt Exp $";
1.1 deraadt 41: #endif /* lint */
42:
43: /* an optimisation over plain strcmp() */
44: #define STREQ(a, b) (*(a) == *(b) && strcmp((a), (b)) == 0)
45:
46: int
47: ascmagic(buf, nbytes)
48: unsigned char *buf;
49: int nbytes; /* size actually read */
50: {
51: int i, has_escapes = 0;
52: unsigned char *s;
53: char nbuf[HOWMANY+1]; /* one extra for terminating '\0' */
54: char *token;
55: register struct names *p;
56:
57: /*
58: * Do the tar test first, because if the first file in the tar
59: * archive starts with a dot, we can confuse it with an nroff file.
60: */
61: switch (is_tar(buf, nbytes)) {
62: case 1:
63: ckfputs("tar archive", stdout);
64: return 1;
65: case 2:
66: ckfputs("POSIX tar archive", stdout);
67: return 1;
68: }
69:
70: /*
71: * for troff, look for . + letter + letter or .\";
72: * this must be done to disambiguate tar archives' ./file
73: * and other trash from real troff input.
74: */
75: if (*buf == '.') {
76: unsigned char *tp = buf + 1;
77:
78: while (isascii(*tp) && isspace(*tp))
79: ++tp; /* skip leading whitespace */
80: if ((isascii(*tp) && (isalnum(*tp) || *tp=='\\') &&
81: isascii(tp[1]) && (isalnum(tp[1]) || tp[1] == '"'))) {
82: ckfputs("troff or preprocessor input text", stdout);
83: return 1;
84: }
85: }
86: if ((*buf == 'c' || *buf == 'C') &&
87: isascii(buf[1]) && isspace(buf[1])) {
88: ckfputs("fortran program text", stdout);
89: return 1;
90: }
91:
1.3 ! millert 92:
! 93: /* Make sure we are dealing with ascii text before looking for tokens */
! 94: for (i = 0; i < nbytes; i++) {
! 95: if (!isascii(buf[i]))
! 96: return 0; /* not all ASCII */
! 97: }
! 98:
1.1 deraadt 99: /* look for tokens from names.h - this is expensive! */
100: /* make a copy of the buffer here because strtok() will destroy it */
101: s = (unsigned char*) memcpy(nbuf, buf, nbytes);
102: s[nbytes] = '\0';
103: has_escapes = (memchr(s, '\033', nbytes) != NULL);
104: while ((token = strtok((char *) s, " \t\n\r\f")) != NULL) {
105: s = NULL; /* make strtok() keep on tokin' */
106: for (p = names; p < names + NNAMES; p++) {
107: if (STREQ(p->name, token)) {
108: ckfputs(types[p->type], stdout);
109: if (has_escapes)
110: ckfputs(" (with escape sequences)",
111: stdout);
112: return 1;
113: }
114: }
115: }
116:
117: /* all else fails, but it is ASCII... */
118: ckfputs("ASCII text", stdout);
119: if (has_escapes) {
120: ckfputs(" (with escape sequences)", stdout);
121: }
122: return 1;
123: }
124:
125: