archive/tar: convert Reader.Next to be loop based

Motivation for change:
* Recursive logic is hard to follow, since it tends to apply
things in reverse. On the other hand, the tar formats tend to
describe meta headers as affecting the next entry.
* Recursion also applies changes in the wrong order. Two test
files are attached that use multiple headers. The previous Go
behavior differs from what GNU and BSD tar do.

Change-Id: Ic1557256fc1363c5cb26570e5d0b9f65a9e57341
Reviewed-on: https://go-review.googlesource.com/14624
Run-TryBot: Joe Tsai <joetsai@digital-static.net>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>

archive/tar/reader_test.go

@@ -288,6 +288,30 @@ var untarTests = []*untarTest{
 			},
 		},
 	},
+	{
+		// Matches the behavior of GNU, BSD, and STAR tar utilities.
+		file: "testdata/gnu-multi-hdrs.tar",
+		headers: []*Header{
+			{
+				Name:     "GNU2/GNU2/long-path-name",
+				Linkname: "GNU4/GNU4/long-linkpath-name",
+				ModTime:  time.Unix(0, 0),
+				Typeflag: '2',
+			},
+		},
+	},
+	{
+		// Matches the behavior of GNU and BSD tar utilities.
+		file: "testdata/pax-multi-hdrs.tar",
+		headers: []*Header{
+			{
+				Name:     "bar",
+				Linkname: "PAX4/PAX4/long-linkpath-name",
+				ModTime:  time.Unix(0, 0),
+				Typeflag: '2',
+			},
+		},
+	},
 	{
 		file: "testdata/neg-size.tar",
 		err:  ErrHeader,