MFp4: libarchive 2.5.4b. (Still 'b' until I get a bit more

feedback, but the 2.5 branch is shaping up nicely.)

In addition to many small bug fixes and code improvements:
 * Another iteration of versioning; I think I've got it right now.
 * Portability:  A lot of progress on Windows support (though I'm
   not committing all of the Windows support files to FreeBSD CVS)
 * Explicit tracking of MBS, WCS, and UTF-8 versions of strings
   in archive_entry; the archive_entry routines now correctly return
   NULL only when something is unset, setting NULL properly clears
   string values.  Most charset conversions have been pushed down to
   archive_string.
 * Better handling of charset conversion failure when writing or
   reading UTF-8 headers in pax archives
 * archive_entry_linkify() provides multiple strategies for
   hardlink matching to suit different format expectations
 * More accurate bzip2 format detection
 * Joerg Sonnenberger's extensive improvements to mtree support
 * Rough support for self-extracting ZIP archives.  Not an ideal
   approach, but it works for the archives I've tried.
 * New "sparsify" option in archive_write_disk converts blocks of nulls
   into seeks.
 * Better default behavior for the test harness; it now reports
   all failures by default instead of coredumping at the first one.

1 files changed, 223 insertions, 106 deletions

diff --git a/lib/libarchive/archive_entry.h b/lib/libarchive/archive_entry.h
index 8c61b9553b63..c453483a65ac 100644
--- a/lib/libarchive/archive_entry.h
+++ b/lib/libarchive/archive_entry.h
@@ -28,20 +28,74 @@
 #ifndef ARCHIVE_ENTRY_H_INCLUDED
 #define	ARCHIVE_ENTRY_H_INCLUDED
 
+/*
+ * Note: archive_entry.h is for use outside of libarchive; the
+ * configuration headers (config.h, archive_platform.h, etc.) are
+ * purely internal.  Do NOT use HAVE_XXX configuration macros to
+ * control the behavior of this header!  If you must conditionalize,
+ * use predefined compiler and/or platform macros.
+ */
+
 #include <sys/types.h>
 #include <stddef.h>  /* for wchar_t */
 #include <time.h>
+
+/* Get appropriate definitions of standard POSIX-style types. */
+/* These should match the types used in 'struct stat' */
+#ifdef _WIN32
+#define	__LA_UID_T	unsigned int
+#define	__LA_GID_T	unsigned int
+#define	__LA_DEV_T	unsigned int
+#define	__LA_MODE_T	unsigned short
+#else
 #include <unistd.h>
+#define	__LA_UID_T	uid_t
+#define	__LA_GID_T	gid_t
+#define	__LA_DEV_T	dev_t
+#define	__LA_MODE_T	mode_t
+#endif
+
+/*
+ * XXX Is this defined for all Windows compilers?  If so, in what
+ * header?  It would be nice to remove the __LA_INO_T indirection and
+ * just use plain ino_t everywhere.  Likewise for the other types just
+ * above.
+ */
+#define	__LA_INO_T	ino_t
+
+
+/*
+ * On Windows, define LIBARCHIVE_STATIC if you're building or using a
+ * .lib.  The default here assumes you're building a DLL.  Only
+ * libarchive source should ever define __LIBARCHIVE_BUILD.
+ */
+#if ((defined __WIN32__) || (defined _WIN32)) && (!defined LIBARCHIVE_STATIC)
+# ifdef __LIBARCHIVE_BUILD
+#  ifdef __GNUC__
+#   define __LA_DECL	__attribute__((dllexport)) extern
+#  else
+#   define __LA_DECL	__declspec(dllexport)
+#  endif
+# else
+#  ifdef __GNUC__
+#   define __LA_DECL	__attribute__((dllimport)) extern
+#  else
+#   define __LA_DECL	__declspec(dllimport)
+#  endif
+# endif
+#else
+/* Static libraries on all platforms and shared libraries on non-Windows. */
+# define __LA_DECL
+#endif
 
 #ifdef __cplusplus
 extern "C" {
 #endif
 
-
 /*
  * Description of an archive entry.
  *
- * Basically, a "struct stat" with a few text fields added in.
+ * You can think of this as "struct stat" with some text fields added in.
  *
  * TODO: Add "comment", "charset", and possibly other entries that are
  * supported by "pax interchange" format.  However, GNU, ustar, cpio,
@@ -90,50 +144,51 @@ struct archive_entry;
  * Basic object manipulation
  */
 
-struct archive_entry	*archive_entry_clear(struct archive_entry *);
+__LA_DECL struct archive_entry	*archive_entry_clear(struct archive_entry *);
 /* The 'clone' function does a deep copy; all of the strings are copied too. */
-struct archive_entry	*archive_entry_clone(struct archive_entry *);
-void			 archive_entry_free(struct archive_entry *);
-struct archive_entry	*archive_entry_new(void);
+__LA_DECL struct archive_entry	*archive_entry_clone(struct archive_entry *);
+__LA_DECL void			 archive_entry_free(struct archive_entry *);
+__LA_DECL struct archive_entry	*archive_entry_new(void);
 
 /*
  * Retrieve fields from an archive_entry.
  */
 
-time_t			 archive_entry_atime(struct archive_entry *);
-long			 archive_entry_atime_nsec(struct archive_entry *);
-time_t			 archive_entry_ctime(struct archive_entry *);
-long			 archive_entry_ctime_nsec(struct archive_entry *);
-dev_t			 archive_entry_dev(struct archive_entry *);
-dev_t			 archive_entry_devmajor(struct archive_entry *);
-dev_t			 archive_entry_devminor(struct archive_entry *);
-mode_t			 archive_entry_filetype(struct archive_entry *);
-void			 archive_entry_fflags(struct archive_entry *,
+__LA_DECL time_t	 archive_entry_atime(struct archive_entry *);
+__LA_DECL long		 archive_entry_atime_nsec(struct archive_entry *);
+__LA_DECL time_t	 archive_entry_ctime(struct archive_entry *);
+__LA_DECL long		 archive_entry_ctime_nsec(struct archive_entry *);
+__LA_DECL dev_t		 archive_entry_dev(struct archive_entry *);
+__LA_DECL dev_t		 archive_entry_devmajor(struct archive_entry *);
+__LA_DECL dev_t		 archive_entry_devminor(struct archive_entry *);
+__LA_DECL __LA_MODE_T	 archive_entry_filetype(struct archive_entry *);
+__LA_DECL void		 archive_entry_fflags(struct archive_entry *,
 			    unsigned long * /* set */,
 			    unsigned long * /* clear */);
-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 *);
-long			 archive_entry_mtime_nsec(struct archive_entry *);
-unsigned int		 archive_entry_nlink(struct archive_entry *);
-const char		*archive_entry_pathname(struct archive_entry *);
-const wchar_t		*archive_entry_pathname_w(struct archive_entry *);
-dev_t			 archive_entry_rdev(struct archive_entry *);
-dev_t			 archive_entry_rdevmajor(struct archive_entry *);
-dev_t			 archive_entry_rdevminor(struct archive_entry *);
-int64_t			 archive_entry_size(struct archive_entry *);
-const char		*archive_entry_strmode(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 *);
+__LA_DECL const char	*archive_entry_fflags_text(struct archive_entry *);
+__LA_DECL __LA_GID_T	 archive_entry_gid(struct archive_entry *);
+__LA_DECL const char	*archive_entry_gname(struct archive_entry *);
+__LA_DECL const wchar_t	*archive_entry_gname_w(struct archive_entry *);
+__LA_DECL const char	*archive_entry_hardlink(struct archive_entry *);
+__LA_DECL const wchar_t	*archive_entry_hardlink_w(struct archive_entry *);
+__LA_DECL __LA_INO_T	 archive_entry_ino(struct archive_entry *);
+__LA_DECL __LA_MODE_T	 archive_entry_mode(struct archive_entry *);
+__LA_DECL time_t	 archive_entry_mtime(struct archive_entry *);
+__LA_DECL long		 archive_entry_mtime_nsec(struct archive_entry *);
+__LA_DECL unsigned int	 archive_entry_nlink(struct archive_entry *);
+__LA_DECL const char	*archive_entry_pathname(struct archive_entry *);
+__LA_DECL const wchar_t	*archive_entry_pathname_w(struct archive_entry *);
+__LA_DECL dev_t		 archive_entry_rdev(struct archive_entry *);
+__LA_DECL dev_t		 archive_entry_rdevmajor(struct archive_entry *);
+__LA_DECL dev_t		 archive_entry_rdevminor(struct archive_entry *);
+__LA_DECL const char	*archive_entry_sourcepath(struct archive_entry *);
+__LA_DECL int64_t	 archive_entry_size(struct archive_entry *);
+__LA_DECL const char	*archive_entry_strmode(struct archive_entry *);
+__LA_DECL const char	*archive_entry_symlink(struct archive_entry *);
+__LA_DECL const wchar_t	*archive_entry_symlink_w(struct archive_entry *);
+__LA_DECL __LA_UID_T	 archive_entry_uid(struct archive_entry *);
+__LA_DECL const char	*archive_entry_uname(struct archive_entry *);
+__LA_DECL const wchar_t	*archive_entry_uname_w(struct archive_entry *);
 
 /*
  * Set fields in an archive_entry.
@@ -142,48 +197,54 @@ const wchar_t		*archive_entry_uname_w(struct archive_entry *);
  * In contrast, 'copy' functions do copy the object pointed to.
  */
 
-void	archive_entry_set_atime(struct archive_entry *, time_t, long);
-void	archive_entry_set_ctime(struct archive_entry *, time_t, long);
-void	archive_entry_set_dev(struct archive_entry *, dev_t);
-void	archive_entry_set_devmajor(struct archive_entry *, dev_t);
-void	archive_entry_set_devminor(struct archive_entry *, dev_t);
-void	archive_entry_set_filetype(struct archive_entry *, unsigned int);
-void	archive_entry_set_fflags(struct archive_entry *,
+__LA_DECL void	archive_entry_set_atime(struct archive_entry *, time_t, long);
+__LA_DECL void	archive_entry_set_ctime(struct archive_entry *, time_t, long);
+__LA_DECL void	archive_entry_set_dev(struct archive_entry *, dev_t);
+__LA_DECL void	archive_entry_set_devmajor(struct archive_entry *, dev_t);
+__LA_DECL void	archive_entry_set_devminor(struct archive_entry *, dev_t);
+__LA_DECL void	archive_entry_set_filetype(struct archive_entry *, unsigned int);
+__LA_DECL void	archive_entry_set_fflags(struct archive_entry *,
 	    unsigned long /* set */, unsigned long /* clear */);
 /* Returns pointer to start of first invalid token, or NULL if none. */
 /* Note that all recognized tokens are processed, regardless. */
-const wchar_t *archive_entry_copy_fflags_text_w(struct archive_entry *,
+__LA_DECL const char *archive_entry_copy_fflags_text(struct archive_entry *,
+	    const char *);
+__LA_DECL const wchar_t *archive_entry_copy_fflags_text_w(struct archive_entry *,
 	    const wchar_t *);
-void	archive_entry_set_gid(struct archive_entry *, gid_t);
-void	archive_entry_set_gname(struct archive_entry *, const char *);
-void	archive_entry_copy_gname(struct archive_entry *, const char *);
-void	archive_entry_copy_gname_w(struct archive_entry *, const wchar_t *);
-void	archive_entry_set_hardlink(struct archive_entry *, const char *);
-void	archive_entry_copy_hardlink(struct archive_entry *, const char *);
-void	archive_entry_copy_hardlink_w(struct archive_entry *, const wchar_t *);
-void	archive_entry_set_ino(struct archive_entry *, unsigned long);
-void	archive_entry_set_link(struct archive_entry *, const char *);
-void	archive_entry_copy_link(struct archive_entry *, const char *);
-void	archive_entry_copy_link_w(struct archive_entry *, const wchar_t *);
-void	archive_entry_set_mode(struct archive_entry *, mode_t);
-void	archive_entry_set_mtime(struct archive_entry *, time_t, long);
-void	archive_entry_set_nlink(struct archive_entry *, unsigned int);
-void	archive_entry_set_pathname(struct archive_entry *, const char *);
-void	archive_entry_copy_pathname(struct archive_entry *, const char *);
-void	archive_entry_copy_pathname_w(struct archive_entry *, const wchar_t *);
-void	archive_entry_set_perm(struct archive_entry *, mode_t);
-void	archive_entry_set_rdev(struct archive_entry *, dev_t);
-void	archive_entry_set_rdevmajor(struct archive_entry *, dev_t);
-void	archive_entry_set_rdevminor(struct archive_entry *, dev_t);
-void	archive_entry_set_size(struct archive_entry *, int64_t);
-void	archive_entry_set_symlink(struct archive_entry *, const char *);
-void	archive_entry_copy_symlink(struct archive_entry *, const char *);
-void	archive_entry_copy_symlink_w(struct archive_entry *, const wchar_t *);
-void	archive_entry_set_uid(struct archive_entry *, uid_t);
-void	archive_entry_set_uname(struct archive_entry *, const char *);
-void	archive_entry_copy_uname(struct archive_entry *, const char *);
-void	archive_entry_copy_uname_w(struct archive_entry *, const wchar_t *);
-
+__LA_DECL void	archive_entry_set_gid(struct archive_entry *, __LA_GID_T);
+__LA_DECL void	archive_entry_set_gname(struct archive_entry *, const char *);
+__LA_DECL void	archive_entry_copy_gname(struct archive_entry *, const char *);
+__LA_DECL void	archive_entry_copy_gname_w(struct archive_entry *, const wchar_t *);
+__LA_DECL int	archive_entry_update_gname_utf8(struct archive_entry *, const char *);
+__LA_DECL void	archive_entry_set_hardlink(struct archive_entry *, const char *);
+__LA_DECL void	archive_entry_copy_hardlink(struct archive_entry *, const char *);
+__LA_DECL void	archive_entry_copy_hardlink_w(struct archive_entry *, const wchar_t *);
+__LA_DECL void	archive_entry_set_ino(struct archive_entry *, unsigned long);
+__LA_DECL void	archive_entry_set_link(struct archive_entry *, const char *);
+__LA_DECL void	archive_entry_copy_link(struct archive_entry *, const char *);
+__LA_DECL void	archive_entry_copy_link_w(struct archive_entry *, const wchar_t *);
+__LA_DECL int	archive_entry_update_link_utf8(struct archive_entry *, const char *);
+__LA_DECL void	archive_entry_set_mode(struct archive_entry *, __LA_MODE_T);
+__LA_DECL void	archive_entry_set_mtime(struct archive_entry *, time_t, long);
+__LA_DECL void	archive_entry_set_nlink(struct archive_entry *, unsigned int);
+__LA_DECL void	archive_entry_set_pathname(struct archive_entry *, const char *);
+__LA_DECL void	archive_entry_copy_pathname(struct archive_entry *, const char *);
+__LA_DECL void	archive_entry_copy_pathname_w(struct archive_entry *, const wchar_t *);
+__LA_DECL int	archive_entry_update_pathname_utf8(struct archive_entry *, const char *);
+__LA_DECL void	archive_entry_set_perm(struct archive_entry *, __LA_MODE_T);
+__LA_DECL void	archive_entry_set_rdev(struct archive_entry *, dev_t);
+__LA_DECL void	archive_entry_set_rdevmajor(struct archive_entry *, dev_t);
+__LA_DECL void	archive_entry_set_rdevminor(struct archive_entry *, dev_t);
+__LA_DECL void	archive_entry_set_size(struct archive_entry *, int64_t);
+__LA_DECL void	archive_entry_copy_sourcepath(struct archive_entry *, const char *);
+__LA_DECL void	archive_entry_set_symlink(struct archive_entry *, const char *);
+__LA_DECL void	archive_entry_copy_symlink(struct archive_entry *, const char *);
+__LA_DECL void	archive_entry_copy_symlink_w(struct archive_entry *, const wchar_t *);
+__LA_DECL void	archive_entry_set_uid(struct archive_entry *, __LA_UID_T);
+__LA_DECL void	archive_entry_set_uname(struct archive_entry *, const char *);
+__LA_DECL void	archive_entry_copy_uname(struct archive_entry *, const char *);
+__LA_DECL void	archive_entry_copy_uname_w(struct archive_entry *, const wchar_t *);
+__LA_DECL int	archive_entry_update_uname_utf8(struct archive_entry *, const char *);
 /*
  * Routines to bulk copy fields to/from a platform-native "struct
  * stat."  Libarchive used to just store a struct stat inside of each
@@ -193,8 +254,8 @@ void	archive_entry_copy_uname_w(struct archive_entry *, const wchar_t *);
  *
  * TODO: On Linux, provide both stat32 and stat64 versions of these functions.
  */
-const struct stat	*archive_entry_stat(struct archive_entry *);
-void	archive_entry_copy_stat(struct archive_entry *, const struct stat *);
+__LA_DECL const struct stat	*archive_entry_stat(struct archive_entry *);
+__LA_DECL void	archive_entry_copy_stat(struct archive_entry *, const struct stat *);
 
 /*
  * ACL routines.  This used to simply store and return text-format ACL
@@ -242,11 +303,11 @@ void	archive_entry_copy_stat(struct archive_entry *, const struct stat *);
  * POSIX.1e) is useful for handling archive formats that combine
  * default and access information in a single ACL list.
  */
-void	 archive_entry_acl_clear(struct archive_entry *);
-void	 archive_entry_acl_add_entry(struct archive_entry *,
+__LA_DECL void	 archive_entry_acl_clear(struct archive_entry *);
+__LA_DECL void	 archive_entry_acl_add_entry(struct archive_entry *,
 	    int /* type */, int /* permset */, int /* tag */,
 	    int /* qual */, const char * /* name */);
-void	 archive_entry_acl_add_entry_w(struct archive_entry *,
+__LA_DECL void	 archive_entry_acl_add_entry_w(struct archive_entry *,
 	    int /* type */, int /* permset */, int /* tag */,
 	    int /* qual */, const wchar_t * /* name */);
 
@@ -255,11 +316,11 @@ void	 archive_entry_acl_add_entry_w(struct archive_entry *,
  * "next" entry.  The want_type parameter allows you to request only
  * access entries or only default entries.
  */
-int	 archive_entry_acl_reset(struct archive_entry *, int /* want_type */);
-int	 archive_entry_acl_next(struct archive_entry *, int /* want_type */,
+__LA_DECL int	 archive_entry_acl_reset(struct archive_entry *, int /* want_type */);
+__LA_DECL int	 archive_entry_acl_next(struct archive_entry *, int /* want_type */,
 	    int * /* type */, int * /* permset */, int * /* tag */,
 	    int * /* qual */, const char ** /* name */);
-int	 archive_entry_acl_next_w(struct archive_entry *, int /* want_type */,
+__LA_DECL int	 archive_entry_acl_next_w(struct archive_entry *, int /* want_type */,
 	    int * /* type */, int * /* permset */, int * /* tag */,
 	    int * /* qual */, const wchar_t ** /* name */);
 
@@ -276,11 +337,11 @@ int	 archive_entry_acl_next_w(struct archive_entry *, int /* want_type */,
  */
 #define	ARCHIVE_ENTRY_ACL_STYLE_EXTRA_ID	1024
 #define	ARCHIVE_ENTRY_ACL_STYLE_MARK_DEFAULT	2048
-const wchar_t	*archive_entry_acl_text_w(struct archive_entry *,
+__LA_DECL const wchar_t	*archive_entry_acl_text_w(struct archive_entry *,
 		    int /* flags */);
 
 /* Return a count of entries matching 'want_type' */
-int	 archive_entry_acl_count(struct archive_entry *, int /* want_type */);
+__LA_DECL int	 archive_entry_acl_count(struct archive_entry *, int /* want_type */);
 
 /*
  * Private ACL parser.  This is private because it handles some
@@ -295,15 +356,15 @@ int	 archive_entry_acl_count(struct archive_entry *, int /* want_type */);
  * TODO: Move this declaration out of the public header and into
  * a private header.  Warnings above are silly.
  */
-int		 __archive_entry_acl_parse_w(struct archive_entry *,
+__LA_DECL int		 __archive_entry_acl_parse_w(struct archive_entry *,
 		    const wchar_t *, int /* type */);
 
 /*
  * extended attributes
  */
 
-void	 archive_entry_xattr_clear(struct archive_entry *);
-void	 archive_entry_xattr_add_entry(struct archive_entry *,
+__LA_DECL void	 archive_entry_xattr_clear(struct archive_entry *);
+__LA_DECL void	 archive_entry_xattr_add_entry(struct archive_entry *,
 	    const char * /* name */, const void * /* value */,
 	    size_t /* size */);
 
@@ -312,37 +373,93 @@ void	 archive_entry_xattr_add_entry(struct archive_entry *,
  * "next" entry.
  */
 
-int	archive_entry_xattr_count(struct archive_entry *);
-int	archive_entry_xattr_reset(struct archive_entry *);
-int	archive_entry_xattr_next(struct archive_entry *,
+__LA_DECL int	archive_entry_xattr_count(struct archive_entry *);
+__LA_DECL int	archive_entry_xattr_reset(struct archive_entry *);
+__LA_DECL int	archive_entry_xattr_next(struct archive_entry *,
 	    const char ** /* name */, const void ** /* value */, size_t *);
 
 /*
- * Utility to detect hardlinks.
+ * Utility to match up hardlinks.
+ *
+ * The 'struct archive_entry_linkresolver' is a cache of archive entries
+ * for files with multiple links.  Here's how to use it:
+ *   1. Create a lookup object with archive_entry_linkresolver_new()
+ *   2. Tell it the archive format you're using.
+ *   3. Hand each archive_entry to archive_entry_linkify().
+ *      That function will return 0, 1, or 2 entries that should
+ *      be written.
+ *   4. Call archive_entry_linkify(resolver, NULL) until
+ *      no more entries are returned.
+ *   5. Call archive_entry_link_resolver_free(resolver) to free resources.
  *
- * The 'struct archive_hardlink_lookup' is a cache of entry
- * names and dev/ino numbers.  Here's how to use it:
- *   1. Create a lookup object with archive_hardlink_lookup_new()
- *   2. Hand each archive_entry to archive_hardlink_lookup().
- *      That function will return NULL (this is not a hardlink to
- *      a previous entry) or the pathname of the first entry
- *      that matched this.
- *   3. Use archive_hardlink_lookup_free() to release the cache.
+ * The entries returned have their hardlink and size fields updated
+ * appropriately.  If an entry is passed in that does not refer to
+ * a file with multiple links, it is returned unchanged.  The intention
+ * is that you should be able to simply filter all entries through
+ * this machine.
  *
  * To make things more efficient, be sure that each entry has a valid
  * nlinks value.  The hardlink cache uses this to track when all links
  * have been found.  If the nlinks value is zero, it will keep every
  * name in the cache indefinitely, which can use a lot of memory.
+ *
+ * Note that archive_entry_size() is reset to zero if the file
+ * body should not be written to the archive.  Pay attention!
  */
-struct archive_entry_linkresolver;
+__LA_DECL struct archive_entry_linkresolver;
 
-struct archive_entry_linkresolver *archive_entry_linkresolver_new(void);
-void archive_entry_linkresolver_free(struct archive_entry_linkresolver *);
-const char *archive_entry_linkresolve(struct archive_entry_linkresolver *,
-    struct archive_entry *);
+/*
+ * There are three different strategies for marking hardlinks.
+ * The descriptions below name them after the best-known
+ * formats that rely on each strategy:
+ *
+ * "Old cpio" is the simplest, it always returns any entry unmodified.
+ *    As far as I know, only cpio formats use this.  Old cpio archives
+ *    store every link with the full body; the onus is on the dearchiver
+ *    to detect and properly link the files as they are restored.
+ * "tar" is also pretty simple; it caches a copy the first time it sees
+ *    any link.  Subsequent appearances are modified to be hardlink
+ *    references to the first one without any body.  Used by all tar
+ *    formats, although the newest tar formats permit the "old cpio" strategy
+ *    as well.  This strategy is very simple for the dearchiver,
+ *    and reasonably straightforward for the archiver.
+ * "new cpio" is trickier.  It stores the body only with the last
+ *    occurrence.  The complication is that we might not
+ *    see every link to a particular file in a single session, so
+ *    there's no easy way to know when we've seen the last occurrence.
+ *    The solution here is to queue one link until we see the next.
+ *    At the end of the session, you can enumerate any remaining
+ *    entries by calling archive_entry_linkify(NULL) and store those
+ *    bodies.  If you have a file with three links l1, l2, and l3,
+ *    you'll get the following behavior if you see all three links:
+ *           linkify(l1) => NULL   (the resolver stores l1 internally)
+ *           linkify(l2) => l1     (resolver stores l2, you write l1)
+ *           linkify(l3) => l2, l3 (all links seen, you can write both).
+ *    If you only see l1 and l2, you'll get this behavior:
+ *           linkify(l1) => NULL
+ *           linkify(l2) => l1
+ *           linkify(NULL) => l2   (at end, you retrieve remaining links)
+ *    As the name suggests, this strategy is used by newer cpio variants.
+ *    It's noticably more complex for the archiver, slightly more complex
+ *    for the dearchiver than the tar strategy, but makes it straightforward
+ *    to restore a file using any link by simply continuing to scan until
+ *    you see a link that is stored with a body.  In contrast, the tar
+ *    strategy requires you to rescan the archive from the beginning to
+ *    correctly extract an arbitrary link.
+ */
+
+__LA_DECL struct archive_entry_linkresolver *archive_entry_linkresolver_new(void);
+__LA_DECL void archive_entry_linkresolver_set_strategy(
+	struct archive_entry_linkresolver *, int /* format_code */);
+__LA_DECL void archive_entry_linkresolver_free(struct archive_entry_linkresolver *);
+__LA_DECL void archive_entry_linkify(struct archive_entry_linkresolver *,
+    struct archive_entry **, struct archive_entry **);
 
 #ifdef __cplusplus
 }
 #endif
 
+/* This is meaningless outside of this header. */
+#undef __LA_DECL
+
 #endif /* !ARCHIVE_ENTRY_H_INCLUDED */