From: Rob Pike <r@golang.org>
Date: Tue, 1 Sep 2009 20:46:59 +0000 (-0700)
Subject: casing operations for byte arrays
X-Git-Tag: weekly.2009-11-06~679
X-Git-Url: http://www.git.cypherpunks.su/?a=commitdiff_plain;h=2f5e75859b8bcb1ad3b8a8d3c4db078ecc5a6158;p=gostls13.git

casing operations for byte arrays

R=rsc
DELTA=186  (181 added, 0 deleted, 5 changed)
OCL=34203
CL=34203
---

diff --git a/src/pkg/bytes/bytes.go b/src/pkg/bytes/bytes.go
index e5e8bffd8c..5375fecaa2 100644
--- a/src/pkg/bytes/bytes.go
+++ b/src/pkg/bytes/bytes.go
@@ -6,7 +6,10 @@
 // Analagous to the facilities of the strings package.
 package bytes
 
-import "utf8"
+import (
+	"unicode";
+	"utf8";
+)
 
 // Compare returns an integer comparing the two byte arrays lexicographically.
 // The result will be 0 if a==b, -1 if a < b, and +1 if a > b
@@ -177,3 +180,83 @@ func HasPrefix(s, prefix []byte) bool {
 func HasSuffix(s, suffix []byte) bool {
 	return len(s) >= len(suffix) && Equal(s[len(s)-len(suffix):len(s)], suffix)
 }
+
+// Map returns a copy of the byte array s with all its characters modified
+// according to the mapping function.
+func Map(mapping func(rune int) int, s []byte) []byte {
+	// In the worst case, the array can grow when mapped, making
+	// things unpleasant.  But it's so rare we barge in assuming it's
+	// fine.  It could also shrink but that falls out naturally.
+	maxbytes := len(s);	// length of b
+	nbytes := 0;	// number of bytes encoded in b
+	b := make([]byte, maxbytes);
+	for wid, i := 0, 0; i < len(s); i += wid {
+		wid = 1;
+		rune := int(s[i]);
+		if rune < utf8.RuneSelf {
+			rune = mapping(rune);
+		} else {
+			rune, wid = utf8.DecodeRune(s[i:len(s)]);
+		}
+		rune = mapping(rune);
+		if nbytes + utf8.RuneLen(rune) > maxbytes {
+			// Grow the buffer.
+			maxbytes = maxbytes*2 + utf8.UTFMax;
+			nb := make([]byte, maxbytes);
+			for i, c := range b[0:nbytes] {
+				nb[i] = c
+			}
+			b = nb;
+		}
+		nbytes += utf8.EncodeRune(rune, b[nbytes:maxbytes]);
+	}
+	return b[0:nbytes];
+}
+
+// ToUpper returns a copy of the byte array s with all Unicode letters mapped to their upper case.
+func ToUpper(s []byte) []byte {
+	return Map(unicode.ToUpper, s)
+}
+
+// ToUpper returns a copy of the byte array s with all Unicode letters mapped to their lower case.
+func ToLower(s []byte) []byte {
+	return Map(unicode.ToLower, s)
+}
+
+// ToTitle returns a copy of the byte array s with all Unicode letters mapped to their title case.
+func Title(s []byte) []byte {
+	return Map(unicode.ToTitle, s)
+}
+
+// Trim returns a slice of the string s, with all leading and trailing white space
+// removed, as defined by Unicode.
+func TrimSpace(s []byte) []byte {
+	start, end := 0, len(s);
+	for wid := 0; start < end; start += wid {
+		wid = 1;
+		rune := int(s[start]);
+		if rune >= utf8.RuneSelf {
+			rune, wid = utf8.DecodeRune(s[start:end])
+		}
+		if !unicode.IsSpace(rune) {
+			break;
+		}
+	}
+	for wid := 0; start < end; end -= wid {
+		wid = 1;
+		rune := int(s[end-1]);
+		if rune >= utf8.RuneSelf {
+			// Back up carefully looking for beginning of rune. Mustn't pass start.
+			for wid = 2; start <= end-wid && !utf8.RuneStart(s[end-wid]); wid++ {
+			}
+			if start > end-wid {	// invalid UTF-8 sequence; stop processing
+				return s[start:end]
+			}
+			rune, wid = utf8.DecodeRune(s[end-wid:end]);
+		}
+		if !unicode.IsSpace(rune) {
+			break;
+		}
+	}
+	return s[start:end];
+}
diff --git a/src/pkg/bytes/bytes_test.go b/src/pkg/bytes/bytes_test.go
index e37767d9a2..a7667ec21e 100644
--- a/src/pkg/bytes/bytes_test.go
+++ b/src/pkg/bytes/bytes_test.go
@@ -8,6 +8,7 @@ import (
 	. "bytes";
 	"strings";
 	"testing";
+	"unicode";
 )
 
 func eq(a, b []string) bool {
@@ -163,3 +164,100 @@ func TestCopy(t *testing.T) {
 		}
 	}
 }
+
+// Test case for any function which accepts and returns a byte array.
+// For ease of creation, we write the byte arrays as strings.
+type StringTest struct {
+	in, out string;
+}
+
+var upperTests = []StringTest {
+	StringTest{"", ""},
+	StringTest{"abc", "ABC"},
+	StringTest{"AbC123", "ABC123"},
+	StringTest{"azAZ09_", "AZAZ09_"},
+	StringTest{"\u0250\u0250\u0250\u0250\u0250", "\u2C6F\u2C6F\u2C6F\u2C6F\u2C6F"},	// grows one byte per char
+}
+
+var lowerTests = []StringTest {
+	StringTest{"", ""},
+	StringTest{"abc", "abc"},
+	StringTest{"AbC123", "abc123"},
+	StringTest{"azAZ09_", "azaz09_"},
+	StringTest{"\u2C6D\u2C6D\u2C6D\u2C6D\u2C6D", "\u0251\u0251\u0251\u0251\u0251"},	// shrinks one byte per char
+}
+
+const space = "\t\v\r\f\n\u0085\u00a0\u2000\u3000"
+
+var trimSpaceTests = []StringTest {
+	StringTest{"", ""},
+	StringTest{"abc", "abc"},
+	StringTest{space + "abc" + space, "abc"},
+	StringTest{" ", ""},
+	StringTest{" \t\r\n \t\t\r\r\n\n ", ""},
+	StringTest{" \t\r\n x\t\t\r\r\n\n ", "x"},
+	StringTest{" \u2000\t\r\n x\t\t\r\r\ny\n \u3000", "x\t\t\r\r\ny"},
+	StringTest{"1 \t\r\n2", "1 \t\r\n2"},
+	StringTest{" x\x80", "x\x80"},	// invalid UTF-8 on end
+	StringTest{" x\xc0", "x\xc0"},	// invalid UTF-8 on end
+}
+
+// Bytes returns a new slice containing the bytes in s.
+// Borrowed from strings to avoid dependency.
+func Bytes(s string) []byte {
+	b := make([]byte, len(s));
+	for i := 0; i < len(s); i++ {
+		b[i] = s[i];
+	}
+	return b;
+}
+
+// Execute f on each test case.  funcName should be the name of f; it's used
+// in failure reports.
+func runStringTests(t *testing.T, f func([]byte) []byte, funcName string, testCases []StringTest) {
+	for i, tc := range testCases {
+		actual := string(f(Bytes(tc.in)));
+		if actual != tc.out {
+			t.Errorf("%s(%q) = %q; want %q", funcName, tc.in, actual, tc.out);
+		}
+	}
+}
+
+func tenRunes(rune int) string {
+	r := make([]int, 10);
+	for i := range r {
+		r[i] = rune
+	}
+	return string(r)
+}
+
+func TestMap(t *testing.T) {
+	// Run a couple of awful growth/shrinkage tests
+	a := tenRunes('a');
+	// 1.  Grow.  This triggers two reallocations in Map.
+	maxRune := func(rune int) int { return unicode.MaxRune };
+	m := Map(maxRune, Bytes(a));
+	expect := tenRunes(unicode.MaxRune);
+	if string(m) != expect {
+		t.Errorf("growing: expected %q got %q", expect, m);
+	}
+	// 2. Shrink
+	minRune := func(rune int) int { return 'a' };
+	m = Map(minRune, Bytes(tenRunes(unicode.MaxRune)));
+	expect = a;
+	if string(m) != expect {
+		t.Errorf("shrinking: expected %q got %q", expect, m);
+	}
+}
+
+func TestToUpper(t *testing.T) {
+	runStringTests(t, ToUpper, "ToUpper", upperTests);
+}
+
+func TestToLower(t *testing.T) {
+	runStringTests(t, ToLower, "ToLower", lowerTests);
+}
+
+func TestTrimSpace(t *testing.T) {
+	runStringTests(t, TrimSpace, "TrimSpace", trimSpaceTests);
+}
diff --git a/src/pkg/strings/strings.go b/src/pkg/strings/strings.go
index eaa6a71a1f..f0f0761576 100644
--- a/src/pkg/strings/strings.go
+++ b/src/pkg/strings/strings.go
@@ -149,7 +149,7 @@ func HasSuffix(s, suffix string) bool {
 }
 
 // Map returns a copy of the string s with all its characters modified
-// according to mapping function.
+// according to the mapping function.
 func Map(mapping func(rune int) int, s string) string {
 	// In the worst case, the string can grow when mapped, making
 	// things unpleasant.  But it's so rare we barge in assuming it's
@@ -177,17 +177,17 @@ func Map(mapping func(rune int) int, s string) string {
 	return string(b[0:nbytes]);
 }
 
-// ToUpper returns a copy of the string s with all letters mapped to their upper case.
+// ToUpper returns a copy of the string s with all Unicode letters mapped to their upper case.
 func ToUpper(s string) string {
 	return Map(unicode.ToUpper, s)
 }
 
-// ToUpper returns a copy of the string s with all letters mapped to their lower case.
+// ToUpper returns a copy of the string s with all Unicode letters mapped to their lower case.
 func ToLower(s string) string {
 	return Map(unicode.ToLower, s)
 }
 
-// ToTitle returns a copy of the string s with all letters mapped to their title case.
+// ToTitle returns a copy of the string s with all Unicode letters mapped to their title case.
 func Title(s string) string {
 	return Map(unicode.ToTitle, s)
 }