Add support for RFC 2389 (FEAT) and RFC 2640 (UTF8) to ftpd(8).

The support for RFC 2640 (UTF8) is optional and rudimentary.
The server just advertises its capability to handle UTF-8 file
names and relies on its own 8-bit cleanness, as well as on
the backward compatibility of UTF-8 with ASCII.  So uploaded
files will have UTF-8 names, but the initial server contents
should be prepared in UTF-8 by hand, no on-the-fly conversion
of file names will be done.

PR:		bin/111714
Submitted by:	Zhang Weiwu <see email in the PR>
MFC after:	1 week

3 files changed, 45 insertions, 3 deletions

diff --git a/libexec/ftpd/ftpcmd.y b/libexec/ftpd/ftpcmd.y
index 5235a40..3c6ecb7 100644
--- a/libexec/ftpd/ftpcmd.y
+++ b/libexec/ftpd/ftpcmd.y
@@ -95,6 +95,7 @@ extern	char proctitle[];
 extern	int usedefault;
 extern  char tmpline[];
 extern	int readonly;
+extern	int assumeutf8;
 extern	int noepsv;
 extern	int noretr;
 extern	int noguestretr;
@@ -135,7 +136,7 @@ extern int epsvall;
 	ABOR	DELE	CWD	LIST	NLST	SITE
 	STAT	HELP	NOOP	MKD	RMD	PWD
 	CDUP	STOU	SMNT	SYST	SIZE	MDTM
-	LPRT	LPSV	EPRT	EPSV
+	LPRT	LPSV	EPRT	EPSV	FEAT
 
 	UMASK	IDLE	CHMOD	MDFIVE
 
@@ -687,6 +688,25 @@ cmd
 			if ($4 != NULL)
 				free($4);
 		}
+	| FEAT CRLF
+		{
+			lreply(211, "Extensions supported:");
+#if 0
+			/* XXX these two keywords are non-standard */
+			printf(" EPRT\r\n");
+			if (!noepsv)
+				printf(" EPSV\r\n");
+#endif
+			printf(" MDTM\r\n");
+			printf(" REST STREAM\r\n");
+			printf(" SIZE\r\n");
+			if (assumeutf8) {
+				/* TVFS requires UTF8, see RFC 3659 */
+				printf(" TVFS\r\n");
+				printf(" UTF8\r\n");
+			}
+			reply(211, "End.");
+		}
 	| SYST check_login CRLF
 		{
 			if ($2) {
@@ -1112,6 +1132,7 @@ struct tab cmdtab[] = {		/* In order defined in RFC 765 */
 	{ "NLST", NLST, OSTR, 1,	"[ <sp> path-name ]" },
 	{ "SITE", SITE, SITECMD, 1,	"site-cmd [ <sp> arguments ]" },
 	{ "SYST", SYST, ARGS, 1,	"(get type of operating system)" },
+	{ "FEAT", FEAT, ARGS, 1,	"(get extended features)" },
 	{ "STAT", STAT, OSTR, 1,	"[ <sp> path-name ]" },
 	{ "HELP", HELP, OSTR, 1,	"[ <sp> <string> ]" },
 	{ "NOOP", NOOP, ARGS, 1,	"" },
diff --git a/libexec/ftpd/ftpd.8 b/libexec/ftpd/ftpd.8
index 7f8a3ca..c17be80 100644
--- a/libexec/ftpd/ftpd.8
+++ b/libexec/ftpd/ftpd.8
@@ -40,7 +40,7 @@
 .Nd Internet File Transfer Protocol server
 .Sh SYNOPSIS
 .Nm
-.Op Fl 46ADdEhMmOoRrSUvW
+.Op Fl 468ADdEhMmOoRrSUvW
 .Op Fl l Op Fl l
 .Op Fl a Ar address
 .Op Fl P Ar port
@@ -78,6 +78,21 @@ When
 is specified, accept connections via
 .Dv AF_INET6
 socket.
+.It Fl 8
+The server is running in transparent UTF-8 mode.
+That is, it just encourages RFC\ 2640 compliant clients to send
+paths encoded in UTF-8 in their commands.
+In particular, the names of files uploaded by such clients will
+have UTF-8 encoding.
+The clients will also assume that server replies and existing file
+names have UTF-8 encoding, too.
+The names of files preloaded to the server should meet this expectation
+for the clients to present the correct file names to their users.
+Own messages of
+.Nm
+are always encoded in ASCII, which in fact is a subset of UTF-8.
+.Pp
+Note that this option does not make the server do any encoding conversion.
 .It Fl A
 Allow only anonymous ftp access.
 .It Fl a
@@ -250,6 +265,7 @@ is specified.
 .It DELE Ta "delete a file [RW]"
 .It EPRT Ta "specify data connection port, multiprotocol"
 .It EPSV Ta "prepare for server-to-server transfer, multiprotocol"
+.It FEAT Ta "give information on extended features of server"
 .It HELP Ta "give help information"
 .It LIST Ta "give list files in a directory" Pq Dq Li "ls -lgA"
 .It LPRT Ta "specify data connection port, multiprotocol"
diff --git a/libexec/ftpd/ftpd.c b/libexec/ftpd/ftpd.c
index 1a6a373..5095f20 100644
--- a/libexec/ftpd/ftpd.c
+++ b/libexec/ftpd/ftpd.c
@@ -128,6 +128,7 @@ int	logging;
 int	restricted_data_ports = 1;
 int	paranoid = 1;	  /* be extra careful about security */
 int	anon_only = 0;    /* Only anonymous ftp allowed */
+int	assumeutf8 = 0;   /* Assume that server file names are in UTF-8 */
 int	guest;
 int	dochroot;
 char	*chrootdir;
@@ -308,7 +309,7 @@ main(int argc, char *argv[], char **envp)
 	openlog("ftpd", LOG_PID | LOG_NDELAY, LOG_FTP);
 
 	while ((ch = getopt(argc, argv,
-	                    "46a:AdDEhlmMoOp:P:rRSt:T:u:UvW")) != -1) {
+	                    "468a:AdDEhlmMoOp:P:rRSt:T:u:UvW")) != -1) {
 		switch (ch) {
 		case '4':
 			family = (family == AF_INET6) ? AF_UNSPEC : AF_INET;
@@ -318,6 +319,10 @@ main(int argc, char *argv[], char **envp)
 			family = (family == AF_INET) ? AF_UNSPEC : AF_INET6;
 			break;
 
+		case '8':
+			assumeutf8 = 1;
+			break;
+
 		case 'a':
 			bindname = optarg;
 			break;