Add -z flag to pax to allow gzipping of archive output. Add -z and -Z (gzip

and compress) to pax when used in tar mode (invoked as 'tar') for
compatibility with GNU tar.

bzip2 functionality for further GNU tar compatibility will be added at a
later date.

Note in the manpage that -z is non-standard.

Obtained from:	OpenBSD
Reviewed by:	-hackers
MFC after:	2 weeks

5 files changed, 107 insertions, 13 deletions

diff --git a/bin/pax/ar_io.c b/bin/pax/ar_io.c
index 0d02f20..ee4681b 100644
--- a/bin/pax/ar_io.c
+++ b/bin/pax/ar_io.c
@@ -44,16 +44,18 @@ static const char rcsid[] =
 #endif /* not lint */
 
 #include <sys/types.h>
-#include <sys/stat.h>
 #include <sys/ioctl.h>
 #include <sys/mtio.h>
-#include <signal.h>
-#include <string.h>
+#include <sys/stat.h>
+#include <sys/wait.h>
+#include <err.h>
+#include <errno.h>
 #include <fcntl.h>
-#include <unistd.h>
+#include <signal.h>
 #include <stdio.h>
-#include <errno.h>
+#include <string.h>
 #include <stdlib.h>
+#include <unistd.h>
 #include "pax.h"
 #include "extern.h"
 
@@ -79,9 +81,12 @@ static int invld_rec;			/* tape has out of spec record size */
 static int wr_trail = 1;		/* trailer was rewritten in append */
 static int can_unlnk = 0;		/* do we unlink null archives?  */
 char *arcname;		  	/* printable name of archive */
+const char *gzip_program;		/* name of gzip program */
+static pid_t zpid = -1;			/* pid of child process */
 
 static int get_phys __P((void));
 extern sigset_t s_mask;
+static void ar_start_gzip __P((int, const char *, int));
 
 /*
  * ar_open()
@@ -121,6 +126,8 @@ ar_open(name)
 			arcname = STDN;
 		} else if ((arfd = open(name, EXT_MODE, DMOD)) < 0)
 			syswarn(0, errno, "Failed open to read on %s", name);
+		if (arfd != -1 && gzip_program != NULL)
+			ar_start_gzip(arfd, gzip_program, 0);
 		break;
 	case ARCHIVE:
 		if (name == NULL) {
@@ -130,6 +137,8 @@ ar_open(name)
 			syswarn(0, errno, "Failed open to write on %s", name);
 		else
 			can_unlnk = 1;
+		if (arfd != -1 && gzip_program != NULL)
+			ar_start_gzip(arfd, gzip_program, 1);
 		break;
 	case APPND:
 		if (name == NULL) {
@@ -335,6 +344,16 @@ ar_close()
 		can_unlnk = 0;
 	}
 
+	/*
+	 * for a quick extract/list, pax frequently exits before the child
+	 * process is done
+	 */
+	if ((act == LIST || act == EXTRACT) && nflag && zpid > 0) {
+		int status;
+		kill(zpid, SIGINT);
+		waitpid(zpid, &status, 0);
+	}
+
 	(void)close(arfd);
 
 	if (vflag && (artyp == ISTAPE)) {
@@ -1287,3 +1306,46 @@ ar_next()
 	}
 	return(0);
 }
+
+/*
+ * ar_start_gzip()
+ * starts the gzip compression/decompression process as a child, using magic
+ * to keep the fd the same in the calling function (parent).
+ */
+void
+ar_start_gzip(int fd, const char *gzip_program, int wr)
+{
+	int fds[2];
+	char *gzip_flags;
+
+	if (pipe(fds) < 0)
+		err(1, "could not pipe");
+	zpid = fork();
+	if (zpid < 0)
+		err(1, "could not fork");
+
+	/* parent */
+	if (zpid) {
+		if (wr)
+			dup2(fds[1], fd);
+		else
+			dup2(fds[0], fd);
+		close(fds[0]);
+		close(fds[1]);
+	} else {
+		if (wr) {
+			dup2(fds[0], STDIN_FILENO);
+			dup2(fd, STDOUT_FILENO);
+			gzip_flags = "-c";
+		} else {
+			dup2(fds[1], STDOUT_FILENO);
+			dup2(fd, STDIN_FILENO);
+			gzip_flags = "-dc";
+		}
+		close(fds[0]);
+		close(fds[1]);
+		if (execlp(gzip_program, gzip_program, gzip_flags, NULL) < 0)
+			err(1, "could not exec");
+		/* NOTREACHED */
+	}
+}
diff --git a/bin/pax/extern.h b/bin/pax/extern.h
index 528ce6c..9f4097c 100644
--- a/bin/pax/extern.h
+++ b/bin/pax/extern.h
@@ -48,6 +48,7 @@
  * ar_io.c
  */
 extern char *arcname;
+extern const char *gzip_program;
 int ar_open __P((char *));
 void ar_close __P((void));
 void ar_drain __P((void));
diff --git a/bin/pax/options.c b/bin/pax/options.c
index 5b253f9..cbdcb11 100644
--- a/bin/pax/options.c
+++ b/bin/pax/options.c
@@ -78,6 +78,9 @@ static void cpio_options __P((register int, register char **));
 static void cpio_usage __P((void));
 #endif
 
+#define GZIP_CMD	"gzip"		/* command to run as gzip */
+#define COMPRESS_CMD	"compress"	/* command to run as compress */
+
 /*
  *	Format specific routine table - MUST BE IN SORTED ORDER BY NAME
  *	(see pax.h for description of each function)
@@ -192,7 +195,7 @@ pax_options(argc, argv)
 	/*
 	 * process option flags
 	 */
-	while ((c=getopt(argc,argv,"ab:cdf:iklno:p:rs:tuvwx:B:DE:G:HLPT:U:XYZ"))
+	while ((c=getopt(argc,argv,"ab:cdf:iklno:p:rs:tuvwx:zB:DE:G:HLPT:U:XYZ"))
 	    != -1) {
 		switch (c) {
 		case 'a':
@@ -376,6 +379,12 @@ pax_options(argc, argv)
 			(void)fputs("\n\n", stderr);
 			pax_usage();
 			break;
+		case 'z':
+			/*
+			 * use gzip.  Non standard option.
+			 */
+			gzip_program = GZIP_CMD;
+			break;
 		case 'B':
 			/*
 			 * non-standard option on number of bytes written on a
@@ -694,6 +703,12 @@ tar_options(argc, argv)
 			 */
 			act = EXTRACT;
 			break;
+		case 'z':
+			/*
+			 * use gzip.  Non standard option.
+			 */
+			gzip_program = GZIP_CMD;
+			break;
 		case 'B':
 			/*
 			 * Nothing to do here, this is pax default
@@ -723,6 +738,12 @@ tar_options(argc, argv)
 			 */
 			Xflag = 1;
 			break;
+		case 'Z':
+			/*
+			 * use compress.
+			 */
+			gzip_program = COMPRESS_CMD;
+			break;
 		case '0':
 			arcname = DEV_0;
 			break;
@@ -1075,18 +1096,18 @@ void
 pax_usage()
 #endif
 {
-	(void)fputs("usage: pax [-cdnv] [-E limit] [-f archive] ", stderr);
+	(void)fputs("usage: pax [-cdnvz] [-E limit] [-f archive] ", stderr);
 	(void)fputs("[-s replstr] ... [-U user] ...", stderr);
 	(void)fputs("\n	   [-G group] ... ", stderr);
 	(void)fputs("[-T [from_date][,to_date]] ... ", stderr);
 	(void)fputs("[pattern ...]\n", stderr);
-	(void)fputs("       pax -r [-cdiknuvDYZ] [-E limit] ", stderr);
+	(void)fputs("       pax -r [-cdiknuvzDYZ] [-E limit] ", stderr);
 	(void)fputs("[-f archive] [-o options] ... \n", stderr);
 	(void)fputs("	   [-p string] ... [-s replstr] ... ", stderr);
 	(void)fputs("[-U user] ... [-G group] ...\n	   ", stderr);
 	(void)fputs("[-T [from_date][,to_date]] ... ", stderr);
 	(void)fputs(" [pattern ...]\n", stderr);
-	(void)fputs("       pax -w [-dituvHLPX] [-b blocksize] ", stderr);
+	(void)fputs("       pax -w [-dituvzHLPX] [-b blocksize] ", stderr);
 	(void)fputs("[ [-a] [-f archive] ] [-x format] \n", stderr);
 	(void)fputs("	   [-B bytes] [-s replstr] ... ", stderr);
 	(void)fputs("[-o options] ... [-U user] ...", stderr);
@@ -1114,7 +1135,7 @@ void
 tar_usage()
 #endif
 {
-	(void)fputs("usage: tar -{txru}[cevfbmopwBHLPX014578] [tapefile] ",
+	(void)fputs("usage: tar -{txru}[cevfbmopwzBHLPXZ014578] [tapefile] ",
 		 stderr);
 	(void)fputs("[blocksize] file1 file2...\n", stderr);
 	exit(1);
diff --git a/bin/pax/pax.1 b/bin/pax/pax.1
index 052d57b..88a11eb 100644
--- a/bin/pax/pax.1
+++ b/bin/pax/pax.1
@@ -44,7 +44,7 @@
 .Nd read and write file archives and copy directory hierarchies
 .Sh SYNOPSIS
 .Nm
-.Op Fl cdnv
+.Op Fl cdnvz
 .Bk -words
 .Op Fl f Ar archive
 .Ek
@@ -71,7 +71,7 @@
 .Op Ar pattern ...\&
 .Nm
 .Fl r
-.Op Fl cdiknuvDYZ
+.Op Fl cdiknuvzDYZ
 .Bk -words
 .Op Fl f Ar archive
 .Ek
@@ -107,7 +107,7 @@
 .Op Ar pattern ...\&
 .Nm
 .Fl w
-.Op Fl dituvHLPX
+.Op Fl dituvzHLPX
 .Bk -words
 .Op Fl b Ar blocksize
 .Ek
@@ -756,6 +756,12 @@ as the result of any specific archive format restrictions.
 The individual archive formats may impose additional restrictions on use.
 Typical archive format restrictions include (but are not limited to):
 file pathname length, file size, link pathname length and the type of the file.
+.It Fl z
+Use
+.Xr gzip 1
+to compress (decompress) the archive while writing (reading).
+Incompatible with
+.Fl a .
 .It Fl B Ar bytes
 Limit the number of bytes written to a single archive volume to
 .Ar bytes .
@@ -1100,6 +1106,7 @@ utility is a superset of the
 .St -p1003.2
 standard.
 The options
+.Fl z ,
 .Fl B ,
 .Fl D ,
 .Fl E ,
diff --git a/bin/pax/pax.c b/bin/pax/pax.c
index 91719ae..a53ba80 100644
--- a/bin/pax/pax.c
+++ b/bin/pax/pax.c
@@ -53,6 +53,7 @@ static const char rcsid[] =
 #include <sys/stat.h>
 #include <sys/time.h>
 #include <sys/resource.h>
+#include <err.h>
 #include <errno.h>
 #include <locale.h>
 #include <paths.h>
@@ -272,6 +273,8 @@ main(argc, argv)
 		archive();
 		break;
 	case APPND:
+		if (gzip_program != NULL)
+			err(1, "can not gzip while appending");
 		append();
 		break;
 	case COPY: