rearrange some constants.  unicode package now defines MaxRune and ReplacementChar.
utf8 package imports unicode to get those definitions.
regenerate dependencies.

R=rsc
DELTA=41  (19 added, 3 deleted, 19 changed)
OCL=34123
CL=34129
diff --git a/src/pkg/Make.deps b/src/pkg/Make.deps
index 0ae5ddf..b77974f 100644
--- a/src/pkg/Make.deps
+++ b/src/pkg/Make.deps
@@ -48,14 +48,14 @@
 rpc.install: bufio.install fmt.install gob.install http.install io.install log.install net.install os.install reflect.install sort.install strconv.install strings.install sync.install template.install unicode.install utf8.install
 runtime.install:
 sort.install:
-strconv.install: bytes.install math.install os.install utf8.install
+strconv.install: bytes.install math.install os.install unicode.install utf8.install
 strings.install: utf8.install
 sync.install:
 syscall.install: sync.install
 tabwriter.install: bytes.install container/vector.install io.install os.install utf8.install
 template.install: bytes.install container/vector.install fmt.install io.install os.install reflect.install runtime.install strings.install
 testing.install: flag.install fmt.install os.install runtime.install utf8.install
-testing/iotest.install: io.install log.install os.install
+testing/iotest.install: bytes.install io.install log.install os.install
 time.install: io.install once.install os.install syscall.install
 unicode.install:
-utf8.install:
+utf8.install: unicode.install
diff --git a/src/pkg/strconv/quote.go b/src/pkg/strconv/quote.go
index f970ef5..e343f67 100644
--- a/src/pkg/strconv/quote.go
+++ b/src/pkg/strconv/quote.go
@@ -6,6 +6,7 @@
 
 import (
 	"os";
+	"unicode";
 	"utf8";
 )
 
@@ -175,7 +176,7 @@
 			value = v;
 			break;
 		}
-		if v > utf8.RuneMax {
+		if v > unicode.MaxRune {
 			err = os.EINVAL;
 			return;
 		}
diff --git a/src/pkg/strings/strings_test.go b/src/pkg/strings/strings_test.go
index 714507b..cd9679e 100644
--- a/src/pkg/strings/strings_test.go
+++ b/src/pkg/strings/strings_test.go
@@ -2,9 +2,10 @@
 // Use of this source code is governed by a BSD-style
 // license that can be found in the LICENSE file.
 
-package strings
+package strings_test
 
 import (
+	. "strings";
 	"testing";
 )
 
@@ -92,7 +93,7 @@
 }
 func TestExplode(t *testing.T) {
 	for _, tt := range explodetests {
-		a := explode(tt.s, tt.n);
+		a := Split(tt.s, "", tt.n);
 		if !eq(a, tt.a) {
 			t.Errorf("explode(%q, %d) = %v; want %v", tt.s, tt.n, a, tt.a);
 			continue;
diff --git a/src/pkg/unicode/digit_test.go b/src/pkg/unicode/digit_test.go
index 19a55bb6..a63404e 100644
--- a/src/pkg/unicode/digit_test.go
+++ b/src/pkg/unicode/digit_test.go
@@ -2,9 +2,12 @@
 // Use of this source code is governed by a BSD-style
 // license that can be found in the LICENSE file.
 
-package unicode
+package unicode_test
 
-import "testing"
+import (
+	"testing";
+	. "unicode";
+)
 
 var testDigit = []int {
 	0x0030,
diff --git a/src/pkg/unicode/letter.go b/src/pkg/unicode/letter.go
index c68ec47..f67b7e5 100644
--- a/src/pkg/unicode/letter.go
+++ b/src/pkg/unicode/letter.go
@@ -3,9 +3,14 @@
 // license that can be found in the LICENSE file.
 
 // This package provides data and functions to test some properties of Unicode code points.
-// It is rudimentary but will improve.
 package unicode
 
+const (
+	MaxRune	= 0x10FFFF;	// Maximum valid Unicode code point.
+	ReplacementChar	= 0xFFFD;	// Represents invalid code points.
+)
+
+
 // The representation of a range of Unicode code points.  The range runs from Lo to Hi
 // inclusive and has the specified stride.
 type Range struct {
@@ -42,8 +47,7 @@
 // this CaseRange represents a sequence of the form (say)
 // Upper Lower Upper Lower.
 const (
-	MaxChar		= 0x10FFFF;	// Maximum valid Unicode character value.
-	UpperLower      = MaxChar + 1;	// (Cannot be a valid delta.)
+	UpperLower      = MaxRune + 1;	// (Cannot be a valid delta.)
 )
 
 // Is tests whether rune is in the specified table of ranges.
@@ -113,10 +117,10 @@
 	return Is(Letter, rune);
 }
 
-// To maps the rune to the specified case, UpperCase, LowerCase, or TitleCase
+// To maps the rune to the specified case: UpperCase, LowerCase, or TitleCase
 func To(_case int, rune int) int {
 	if _case < 0 || MaxCase <= _case {
-		return 0xFFFD	// as reasonable an error as any
+		return ReplacementChar	// as reasonable an error as any
 	}
 	// binary search over ranges
 	lo := 0;
@@ -126,7 +130,7 @@
 		r := CaseRanges[m];
 		if r.Lo <= rune && rune <= r.Hi {
 			delta := int(r.Delta[_case]);
-			if delta > MaxChar {
+			if delta > MaxRune {
 				// In an Upper-Lower sequence, which always starts with
 				// an UpperCase letter, the real deltas always look like:
 				//	{0, 1, 0}    UpperCase (Lower is next)
diff --git a/src/pkg/unicode/letter_test.go b/src/pkg/unicode/letter_test.go
index 13daa03..0ccb29f 100644
--- a/src/pkg/unicode/letter_test.go
+++ b/src/pkg/unicode/letter_test.go
@@ -2,9 +2,12 @@
 // Use of this source code is governed by a BSD-style
 // license that can be found in the LICENSE file.
 
-package unicode
+package unicode_test
 
-import "testing"
+import (
+	"testing";
+	. "unicode";
+)
 
 var upperTest = []int{
 	0x41,
diff --git a/src/pkg/unicode/script_test.go b/src/pkg/unicode/script_test.go
index 390e47b..da525c5 100644
--- a/src/pkg/unicode/script_test.go
+++ b/src/pkg/unicode/script_test.go
@@ -2,9 +2,12 @@
 // Use of this source code is governed by a BSD-style
 // license that can be found in the LICENSE file.
 
-package unicode
+package unicode_test
 
-import "testing"
+import (
+	"testing";
+	. "unicode";
+)
 
 type T struct {
 	rune	int;
diff --git a/src/pkg/utf8/utf8.go b/src/pkg/utf8/utf8.go
index 9c2ac79..62adcd9 100644
--- a/src/pkg/utf8/utf8.go
+++ b/src/pkg/utf8/utf8.go
@@ -6,11 +6,12 @@
 // This package calls a Unicode character a rune for brevity.
 package utf8
 
+import "unicode"	// only needed for a couple of constants
+
 // Numbers fundamental to the encoding.
 const (
-	RuneError = 0xFFFD;	// the "error" Rune or "replacement character".
+	RuneError = unicode.ReplacementChar;	// the "error" Rune or "replacement character".
 	RuneSelf = 0x80;	// characters below Runeself are represented as themselves in a single byte.
-	RuneMax = 0x10FFFF;	// maximum Unicode code point.
 	UTFMax = 4;	// maximum number of bytes of a UTF-8 encoded Unicode character.
 )
 
@@ -239,7 +240,7 @@
 		return 2;
 	}
 
-	if rune > RuneMax {
+	if rune > unicode.MaxRune {
 		rune = RuneError
 	}