better detetection/handling of binary files.

make -a do the right thing, doc and implement -U, -I.
add --help and --mmap for compatibility.
some other minor fixes.

some from NetBSD.  ok deraadt@

5 files changed, 75 insertions, 29 deletions

diff --git a/usr.bin/grep/binary.c b/usr.bin/grep/binary.c
index 360d142f4bd..e8ff9188fd8 100644
--- a/usr.bin/grep/binary.c
+++ b/usr.bin/grep/binary.c
@@ -1,4 +1,4 @@
-/*	$OpenBSD: binary.c,v 1.4 2003/06/22 22:38:50 deraadt Exp $	*/
+/*	$OpenBSD: binary.c,v 1.5 2003/06/23 00:55:09 tedu Exp $	*/
 
 /*-
  * Copyright (c) 1999 James Howard and Dag-Erling Coïdan Smørgrav
@@ -83,7 +83,7 @@ mmbin_file(mmf_t *f)
 
 	/* XXX knows too much about mmf internals */
 	for (i = 0; i < BUFFER_SIZE && i < f->len; i++)
-		if (!isprint(f->base[i]))
+		if (!isprint(f->base[i]) && !isspace(f->base[i]))
 			return 1;
 	mmrewind(f);
 	return 0;
diff --git a/usr.bin/grep/grep.1 b/usr.bin/grep/grep.1
index bd619f834c7..1e7108942cb 100644
--- a/usr.bin/grep/grep.1
+++ b/usr.bin/grep/grep.1
@@ -1,4 +1,4 @@
-.\"	$OpenBSD: grep.1,v 1.9 2003/06/22 22:42:00 millert Exp $
+.\"	$OpenBSD: grep.1,v 1.10 2003/06/23 00:55:09 tedu Exp $
 .\" Copyright (c) 1980, 1990, 1993
 .\"	The Regents of the University of California.  All rights reserved.
 .\"
@@ -37,7 +37,7 @@
 .Sh SYNOPSIS
 .Nm grep
 .Op Fl AB Ar num
-.Op Fl CEFGHLPRSVZabchilnoqsvwx
+.Op Fl CEFGHILPRSUVZabchilnoqsvwx
 .Op Fl e Ar pattern
 .Op Fl f Ar file
 .Op Ar
@@ -122,6 +122,8 @@ If
 .Fl R
 is specified, follow symbolic links only if they were explictly listed
 on the command line.
+.It Fl I
+Ignore binary files.
 .It Fl L
 Only the names of files not containing selected lines are written to
 standard output.
@@ -139,6 +141,8 @@ Recursively search subdirectories listed.
 If
 .Fl R
 is specified, all symbolic links are followed.
+.It Fl U
+Search binary files, but do not attempt to print them.
 .It Fl V
 Display version information.
 .It Fl Z
@@ -147,7 +151,7 @@ Force
 to behave as
 .Nm zgrep .
 .It Fl a
-Do not search in binary files.
+Treat all files as text.
 .It Fl b
 The block number on the disk in which a matched pattern is located is
 displayed in front of the respective matched line.
diff --git a/usr.bin/grep/grep.c b/usr.bin/grep/grep.c
index a955223851c..f644f97dd3c 100644
--- a/usr.bin/grep/grep.c
+++ b/usr.bin/grep/grep.c
@@ -1,4 +1,4 @@
-/*	$OpenBSD: grep.c,v 1.8 2003/06/22 23:51:22 tedu Exp $	*/
+/*	$OpenBSD: grep.c,v 1.9 2003/06/23 00:55:09 tedu Exp $	*/
 
 /*-
  * Copyright (c) 1999 James Howard and Dag-Erling Coïdan Smørgrav
@@ -27,6 +27,7 @@
  */
 
 #include <sys/types.h>
+#include <sys/limits.h>
 #include <sys/stat.h>
 
 #include <err.h>
@@ -67,7 +68,6 @@ int	 Vflag;		/* -V: display version information */
 #ifndef NOZ
 int	 Zflag;		/* -Z: decompress input before processing */
 #endif
-int	 aflag;		/* -a: only search ascii files */
 int	 bflag;		/* -b: show block numbers for each match */
 int	 cflag;		/* -c: only show a count of matching lines */
 int	 hflag;		/* -h: don't print filename headers */
@@ -81,6 +81,14 @@ int	 vflag;		/* -v: only show non-matching lines */
 int	 wflag;		/* -w: pattern must start and end on word boundaries */
 int	 xflag;		/* -x: pattern must match entire line */
 
+int binbehave = BIN_FILE_BIN;
+
+enum {
+	BIN_OPT = CHAR_MAX + 1,
+	HELP_OPT,
+	MMAP_OPT
+};
+
 /* Housekeeping */
 int	 first;		/* flag whether or not this is our fist match */
 int	 tail;		/* lines left to print */
@@ -93,29 +101,35 @@ usage(void)
 {
 	fprintf(stderr,
 #ifdef NOZ
-	    "usage: %s [-[AB] num] [-CEFGHLPRSVabchilnoqsvwx]"
+	    "usage: %s [-[AB] num] [-CEFGHILPRSUVabchilnoqsvwx]"
 #else
-	    "usage: %s [-[AB] num] [-CEFGHLPRSVZabchilnoqsvwx]"
+	    "usage: %s [-[AB] num] [-CEFGHILPRSUVZabchilnoqsvwx]"
 #endif
 	    " [-e pattern] [-f file] [file ...]\n", __progname);
 	exit(2);
 }
 
 #ifdef NOZ
-static char *optstr = "0123456789A:B:CEFGHLPSRUVabce:f:hilnoqrsuvwxy";
+static char *optstr = "0123456789A:B:CEFGHILPSRUVabce:f:hilnoqrsuvwxy";
 #else
-static char *optstr = "0123456789A:B:CEFGHLPSRUVZabce:f:hilnoqrsuvwxy";
+static char *optstr = "0123456789A:B:CEFGHILPSRUVZabce:f:hilnoqrsuvwxy";
 #endif
 
 struct option long_options[] =
 {
-	{"basic-regexp",        no_argument,       NULL, 'G'},
-	{"extended-regexp",     no_argument,       NULL, 'E'},
-	{"fixed-strings",       no_argument,       NULL, 'F'},
+	{"binary-files",        required_argument, NULL, BIN_OPT},
+	{"help",                no_argument,       NULL, HELP_OPT},
+	{"mmap",                no_argument,       NULL, MMAP_OPT},
 	{"after-context",       required_argument, NULL, 'A'},
 	{"before-context",      required_argument, NULL, 'B'},
 	{"context",             optional_argument, NULL, 'C'},
+	{"devices",             required_argument, NULL, 'D'},
+	{"extended-regexp",     no_argument,       NULL, 'E'},
+	{"fixed-strings",       no_argument,       NULL, 'F'},
+	{"basic-regexp",        no_argument,       NULL, 'G'},
+	{"binary",              no_argument,       NULL, 'U'},
 	{"version",             no_argument,       NULL, 'V'},
+	{"text",                no_argument,       NULL, 'a'},
 	{"byte-offset",         no_argument,       NULL, 'b'},
 	{"count",               no_argument,       NULL, 'c'},
 	{"regexp",              required_argument, NULL, 'e'},
@@ -129,11 +143,9 @@ struct option long_options[] =
 	{"silent",              no_argument,       NULL, 'q'},
 	{"recursive",           no_argument,       NULL, 'r'},
 	{"no-messages",         no_argument,       NULL, 's'},
-	{"text",                no_argument,       NULL, 'a'},
 	{"revert-match",        no_argument,       NULL, 'v'},
 	{"word-regexp",         no_argument,       NULL, 'w'},
 	{"line-regexp",         no_argument,       NULL, 'x'},
-	{"binary",              no_argument,       NULL, 'U'},
 	{"unix-byte-offsets",   no_argument,       NULL, 'u'},
 #ifndef NOZ
 	{"decompress",          no_argument,       NULL, 'Z'},
@@ -230,6 +242,9 @@ main(int argc, char *argv[])
 		case 'H':
 			Hflag++;
 			break;
+		case 'I':
+			binbehave = BIN_FILE_SKIP;
+			break;
 		case 'L':
 			lflag = 0;
 			Lflag = qflag = 1;
@@ -246,13 +261,11 @@ main(int argc, char *argv[])
 			oflag++;
 			break;
 		case 'U':
-		case 'u':
-			/* these are here for compatability */
+			binbehave = BIN_FILE_BIN;
 			break;
 		case 'V':
 			fprintf(stderr, "grep version %u.%u\n", VER_MAJ, VER_MIN);
-			fprintf(stderr, argv[0]);
-			usage();
+			exit(0);
 			break;
 #ifndef NOZ
 		case 'Z':
@@ -260,7 +273,7 @@ main(int argc, char *argv[])
 			break;
 #endif
 		case 'a':
-			aflag = 1;
+			binbehave = BIN_FILE_TEXT;
 			break;
 		case 'b':
 			bflag = 1;
@@ -308,6 +321,21 @@ main(int argc, char *argv[])
 		case 'x':
 			xflag = 1;
 			break;
+		case BIN_OPT:
+			if (strcmp("binary", optarg) == 0)
+				binbehave = BIN_FILE_BIN;
+			else if (strcmp("without-match", optarg) == 0)
+				binbehave = BIN_FILE_SKIP;
+			else if (strcmp("text", optarg) == 0)
+				binbehave = BIN_FILE_TEXT;
+			else
+				errx(2, "Unknown binary-files option");
+			break;
+		case 'u':
+		case MMAP_OPT:
+			/* default, compatibility */
+			break;
+		case HELP_OPT:
 		default:
 			usage();
 		}
diff --git a/usr.bin/grep/grep.h b/usr.bin/grep/grep.h
index 907f6c042d9..d9981e3808b 100644
--- a/usr.bin/grep/grep.h
+++ b/usr.bin/grep/grep.h
@@ -1,4 +1,4 @@
-/*	$OpenBSD: grep.h,v 1.2 2003/06/22 22:38:50 deraadt Exp $	*/
+/*	$OpenBSD: grep.h,v 1.3 2003/06/23 00:55:09 tedu Exp $	*/
 
 /*-
  * Copyright (c) 1999 James Howard and Dag-Erling Coïdan Smørgrav
@@ -35,6 +35,10 @@
 #define VER_MAJ 0
 #define VER_MIN 9
 
+#define BIN_FILE_BIN	0
+#define BIN_FILE_SKIP	1
+#define BIN_FILE_TEXT	2
+
 typedef struct {
 	size_t		 len;
 	int		 line_no;
@@ -48,8 +52,9 @@ extern int	 cflags, eflags;
 
 /* Command line flags */
 extern int	 Aflag, Bflag, Hflag, Lflag, Pflag, Sflag, Rflag, Zflag,
-		 aflag, bflag, cflag, hflag, lflag, nflag, qflag, sflag,
+		 bflag, cflag, hflag, lflag, nflag, qflag, sflag,
 		 vflag, wflag, xflag;
+extern int	 binbehave;
 
 extern int	 first, lead, matchall, patterns, tail;
 extern char    **pattern;
diff --git a/usr.bin/grep/util.c b/usr.bin/grep/util.c
index 2d51301ebb4..730d39fff6d 100644
--- a/usr.bin/grep/util.c
+++ b/usr.bin/grep/util.c
@@ -1,4 +1,4 @@
-/*	$OpenBSD: util.c,v 1.3 2003/06/22 22:38:50 deraadt Exp $	*/
+/*	$OpenBSD: util.c,v 1.4 2003/06/23 00:55:09 tedu Exp $	*/
 
 /*-
  * Copyright (c) 1999 James Howard and Dag-Erling Coïdan Smørgrav
@@ -47,7 +47,7 @@
  */
 
 static int	linesqueued;
-static int	procline(str_t *l);
+static int	procline(str_t *l, int);
 
 int
 grep_tree(char **argv)
@@ -92,7 +92,7 @@ procfile(char *fn)
 {
 	str_t ln;
 	file_t *f;
-	int c, t, z;
+	int c, t, z, nottext;
 
 	if (fn == NULL) {
 		fn = "(standard input)";
@@ -105,7 +105,9 @@ procfile(char *fn)
 			warn("%s", fn);
 		return 0;
 	}
-	if (aflag && grep_bin_file(f)) {
+
+	nottext = grep_bin_file(f);
+	if (nottext && binbehave == BIN_FILE_SKIP) {
 		grep_close(f);
 		return 0;
 	}
@@ -127,7 +129,7 @@ procfile(char *fn)
 
 		z = tail;
 
-		if ((t = procline(&ln)) == 0 && Bflag > 0 && z == 0) {
+		if ((t = procline(&ln, nottext)) == 0 && Bflag > 0 && z == 0) {
 			enqueue(&ln);
 			linesqueued++;
 		}
@@ -146,6 +148,10 @@ procfile(char *fn)
 		printf("%s\n", fn);
 	if (Lflag && c == 0)
 		printf("%s\n", fn);
+	if (c && !cflag && !lflag && !Lflag &&
+	    binbehave == BIN_FILE_BIN && nottext)
+		printf("Binary file %s matches\n", fn);
+
 	return c;
 }
 
@@ -157,7 +163,7 @@ procfile(char *fn)
 #define isword(x) (isalnum(x) || (x) == '_')
 
 static int
-procline(str_t *l)
+procline(str_t *l, int nottext)
 {
 	regmatch_t	pmatch;
 	int		c, i, r, t;
@@ -192,6 +198,9 @@ procline(str_t *l)
 	}
 
 print:
+	if (c && binbehave == BIN_FILE_BIN && nottext)
+		return c; /* Binary file */
+
 	if ((tail > 0 || c) && !cflag && !qflag) {
 		if (c) {
 			if (first > 0 && tail == 0 && (Bflag < linesqueued) && (Aflag || Bflag))