In pax interchange format, use UTF8 for writing

link names, usernames, or group names that contain
non-ASCII characters.

In particular, this corrects an inconsistency reported
by Ed Maste when archiving symlinks with odd characters:
long symlinks would get preserved, short ones would
be changed.

3 files changed, 70 insertions, 21 deletions

diff --git a/lib/libarchive/archive_entry.c b/lib/libarchive/archive_entry.c
index 1a0febd..82d76b4 100644
--- a/lib/libarchive/archive_entry.c
+++ b/lib/libarchive/archive_entry.c
@@ -203,6 +203,8 @@ aes_copy(struct aes *dest, struct aes *src)
 static const char *
 aes_get_mbs(struct aes *aes)
 {
+	if (aes->aes_mbs == NULL && aes->aes_wcs == NULL)
+		return NULL;
 	if (aes->aes_mbs == NULL && aes->aes_wcs != NULL) {
 		/*
 		 * XXX Need to estimate the number of byte in the
@@ -224,6 +226,8 @@ aes_get_mbs(struct aes *aes)
 static const wchar_t *
 aes_get_wcs(struct aes *aes)
 {
+	if (aes->aes_wcs == NULL && aes->aes_mbs == NULL)
+		return NULL;
 	if (aes->aes_wcs == NULL && aes->aes_mbs != NULL) {
 		/*
 		 * No single byte will be more than one wide character,
@@ -457,12 +461,24 @@ archive_entry_gname(struct archive_entry *entry)
 	return (aes_get_mbs(&entry->ae_gname));
 }
 
+const wchar_t *
+archive_entry_gname_w(struct archive_entry *entry)
+{
+	return (aes_get_wcs(&entry->ae_gname));
+}
+
 const char *
 archive_entry_hardlink(struct archive_entry *entry)
 {
 	return (aes_get_mbs(&entry->ae_hardlink));
 }
 
+const wchar_t *
+archive_entry_hardlink_w(struct archive_entry *entry)
+{
+	return (aes_get_wcs(&entry->ae_hardlink));
+}
+
 ino_t
 archive_entry_ino(struct archive_entry *entry)
 {
@@ -536,6 +552,12 @@ archive_entry_symlink(struct archive_entry *entry)
 	return (aes_get_mbs(&entry->ae_symlink));
 }
 
+const wchar_t *
+archive_entry_symlink_w(struct archive_entry *entry)
+{
+	return (aes_get_wcs(&entry->ae_symlink));
+}
+
 uid_t
 archive_entry_uid(struct archive_entry *entry)
 {
@@ -548,6 +570,12 @@ archive_entry_uname(struct archive_entry *entry)
 	return (aes_get_mbs(&entry->ae_uname));
 }
 
+const wchar_t *
+archive_entry_uname_w(struct archive_entry *entry)
+{
+	return (aes_get_wcs(&entry->ae_uname));
+}
+
 /*
  * Functions to set archive_entry properties.
  */
diff --git a/lib/libarchive/archive_entry.h b/lib/libarchive/archive_entry.h
index 8d9fdef..a35f0df 100644
--- a/lib/libarchive/archive_entry.h
+++ b/lib/libarchive/archive_entry.h
@@ -79,7 +79,9 @@ void			 archive_entry_fflags(struct archive_entry *,
 const char		*archive_entry_fflags_text(struct archive_entry *);
 gid_t			 archive_entry_gid(struct archive_entry *);
 const char		*archive_entry_gname(struct archive_entry *);
+const wchar_t		*archive_entry_gname_w(struct archive_entry *);
 const char		*archive_entry_hardlink(struct archive_entry *);
+const wchar_t		*archive_entry_hardlink_w(struct archive_entry *);
 ino_t			 archive_entry_ino(struct archive_entry *);
 mode_t			 archive_entry_mode(struct archive_entry *);
 time_t			 archive_entry_mtime(struct archive_entry *);
@@ -92,8 +94,10 @@ dev_t			 archive_entry_rdevminor(struct archive_entry *);
 int64_t			 archive_entry_size(struct archive_entry *);
 const struct stat	*archive_entry_stat(struct archive_entry *);
 const char		*archive_entry_symlink(struct archive_entry *);
+const wchar_t		*archive_entry_symlink_w(struct archive_entry *);
 uid_t			 archive_entry_uid(struct archive_entry *);
 const char		*archive_entry_uname(struct archive_entry *);
+const wchar_t		*archive_entry_uname_w(struct archive_entry *);
 
 /*
  * Set fields in an archive_entry.
diff --git a/lib/libarchive/archive_write_set_format_pax.c b/lib/libarchive/archive_write_set_format_pax.c
index 547ada9..75dfe30 100644
--- a/lib/libarchive/archive_write_set_format_pax.c
+++ b/lib/libarchive/archive_write_set_format_pax.c
@@ -62,6 +62,7 @@ static int		 archive_write_pax_header(struct archive *,
 static char		*build_pax_attribute_name(char *dest, const char *src);
 static char		*build_ustar_entry_name(char *dest, const char *src, const char *insert);
 static char		*format_int(char *dest, int64_t);
+static int		 has_non_ASCII(const wchar_t *);
 static int		 write_nulls(struct archive *, size_t);
 
 /*
@@ -315,7 +316,7 @@ archive_write_pax_header(struct archive *a,
 	struct archive_entry *entry_main;
 	const char *linkname, *p;
 	const char *hardlink;
-	const wchar_t *wp, *wp2;
+	const wchar_t *wp;
 	const char *suffix_start;
 	int need_extension, r, ret;
 	struct pax *pax;
@@ -375,36 +376,42 @@ archive_write_pax_header(struct archive *a,
 		/* Find the largest suffix that fits in 'name' field. */
 		suffix_start = strchr(p + strlen(p) - 100 - 1, '/');
 
-	/* Find non-ASCII character, if any. */
-	wp2 = wp;
-	while (*wp2 != L'\0' && *wp2 < 128)
-		wp2++;
-
 	/*
 	 * If name is too long, or has non-ASCII characters, add
 	 * 'path' to pax extended attrs.
 	 */
-	if (suffix_start == NULL || suffix_start - p > 155 || *wp2 != L'\0') {
+	if (suffix_start == NULL || suffix_start - p > 155 || has_non_ASCII(wp)) {
 		add_pax_attr_w(&(pax->pax_header), "path", wp);
 		archive_entry_set_pathname(entry_main,
 		    build_ustar_entry_name(ustar_entry_name, p, NULL));
 		need_extension = 1;
 	}
 
-	/* If link name is too long, add 'linkpath' to pax extended attrs. */
+	/* If link name is too long or has non-ASCII characters, add
+	 * 'linkpath' to pax extended attrs. */
 	linkname = hardlink;
 	if (linkname == NULL)
 		linkname = archive_entry_symlink(entry_main);
 
-	if (linkname != NULL && strlen(linkname) > 100) {
-		add_pax_attr(&(pax->pax_header), "linkpath", linkname);
+	if (linkname != NULL) {
+		/* There is a link name, get the wide version as well. */
 		if (hardlink != NULL)
-			archive_entry_set_hardlink(entry_main,
-			    "././@LongHardLink");
+			wp = archive_entry_hardlink_w(entry_main);
 		else
-			archive_entry_set_symlink(entry_main,
-			    "././@LongSymLink");
-		need_extension = 1;
+			wp = archive_entry_symlink_w(entry_main);
+
+		/* If the link is long or has a non-ASCII character,
+		 * store it as a pax extended attribute. */
+		if (strlen(linkname) > 100 || has_non_ASCII(wp)) {
+			add_pax_attr_w(&(pax->pax_header), "linkpath", wp);
+			if (hardlink != NULL)
+				archive_entry_set_hardlink(entry_main,
+				    "././@LongHardLink");
+			else
+				archive_entry_set_symlink(entry_main,
+				    "././@LongSymLink");
+			need_extension = 1;
+		}
 	}
 
 	/* If file size is too large, add 'size' to pax extended attrs. */
@@ -419,11 +426,12 @@ archive_write_pax_header(struct archive *a,
 		need_extension = 1;
 	}
 
-	/* If group name is too large, add 'gname' to pax extended attrs. */
-	/* TODO: If gname has non-ASCII characters, use pax attribute. */
+	/* If group name is too large or has non-ASCII characters, add
+	 * 'gname' to pax extended attrs. */
 	p = archive_entry_gname(entry_main);
-	if (p != NULL && strlen(p) > 31) {
-		add_pax_attr(&(pax->pax_header), "gname", p);
+	wp = archive_entry_gname_w(entry_main);
+	if (p != NULL && (strlen(p) > 31 || has_non_ASCII(wp))) {
+		add_pax_attr_w(&(pax->pax_header), "gname", wp);
 		archive_entry_set_gname(entry_main, NULL);
 		need_extension = 1;
 	}
@@ -437,8 +445,9 @@ archive_write_pax_header(struct archive *a,
 	/* If user name is too large, add 'uname' to pax extended attrs. */
 	/* TODO: If uname has non-ASCII characters, use pax attribute. */
 	p = archive_entry_uname(entry_main);
-	if (p != NULL && strlen(p) > 31) {
-		add_pax_attr(&(pax->pax_header), "uname", p);
+	wp = archive_entry_uname_w(entry_main);
+	if (p != NULL && (strlen(p) > 31 || has_non_ASCII(wp))) {
+		add_pax_attr_w(&(pax->pax_header), "uname", wp);
 		archive_entry_set_uname(entry_main, NULL);
 		need_extension = 1;
 	}
@@ -1001,3 +1010,11 @@ archive_write_pax_data(struct archive *a, const void *buff, size_t s)
 	pax->entry_bytes_remaining -= s;
 	return (ret);
 }
+
+static int
+has_non_ASCII(const wchar_t *wp)
+{
+	while (*wp != L'\0' && *wp < 128)
+		wp++;
+	return (*wp != L'\0');
+}