rudimentary string utilities.

R=r
DELTA=314  (306 added, 8 deleted, 0 changed)
OCL=15074
CL=15263

src/lib/make.bash

@@ -6,14 +6,6 @@
 
 echo; echo; echo %%%% making lib %%%%; echo
 
-rm -f *.6
-for i in fmt.go flag.go container/vector.go sort.go
-do
-	base=$(basename $i .go)
-	echo 6g -o $GOROOT/pkg/$base.6 $i
-	6g -o $GOROOT/pkg/$base.6 $i
-done
-
 for i in os math
 do
 	echo; echo; echo %%%% making lib/$i %%%%; echo
@@ -22,3 +14,11 @@ do
 	cd ..
 done
 
+rm -f *.6
+for i in fmt.go flag.go container/vector.go rand.go sort.go strings.go
+do
+	base=$(basename $i .go)
+	echo 6g -o $GOROOT/pkg/$base.6 $i
+	6g -o $GOROOT/pkg/$base.6 $i
+done
+

src/lib/strings.go

+// Copyright 2009 The Go Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package strings
+
+// Count UTF-8 sequences in s.
+// Assumes s is well-formed.
+export func utflen(s string) int {
+	n := 0;
+	for i := 0; i < len(s); i++ {
+		if s[i]&0xC0 != 0x80 {
+			n++
+		}
+	}
+	return n
+}
+
+// Split string into array of UTF-8 sequences (still strings)
+export func explode(s string) *[]string {
+	a := new([]string, utflen(s));
+	j := 0;
+	for i := 0; i < len(a); i++ {
+		ej := j;
+		ej++;
+		for ej < len(s) && (s[ej]&0xC0) == 0x80 {
+			ej++
+		}
+		a[i] = s[j:ej];
+		j = ej
+	}
+	return a
+}
+
+// Count non-overlapping instances of sep in s.
+export func count(s, sep string) int {
+	if sep == "" {
+		return utflen(s)+1
+	}
+	c := sep[0];
+	n := 0;
+	for i := 0; i+len(sep) <= len(s); i++ {
+		if s[i] == c && (len(sep) == 1 || s[i:i+len(sep)] == sep) {
+			n++;
+			i += len(sep)-1
+		}
+	}
+	return n
+}
+
+// Return index of first instance of sep in s.
+export func index(s, sep string) int {
+	if sep == "" {
+		return 0
+	}
+	c := sep[0];
+	for i := 0; i+len(sep) <= len(s); i++ {
+		if s[i] == c && (len(sep) == 1 || s[i:i+len(sep)] == sep) {
+			return i
+		}
+	}
+	return -1
+}
+
+// Split string into list of strings at separators
+export func split(s, sep string) *[]string {
+	if sep == "" {
+		return explode(s)
+	}
+	c := sep[0];
+	start := 0;
+	n := count(s, sep)+1;
+	a := new([]string, n);
+	na := 0;
+	for i := 0; i+len(sep) <= len(s); i++ {
+		if s[i] == c && (len(sep) == 1 || s[i:i+len(sep)] == sep) {
+			a[na] = s[start:i];
+			na++;
+			start = i+len(sep);
+			i += len(sep)-1
+		}
+	}
+	a[na] = s[start:len(s)];
+	return a
+}
+	
+// Join list of strings with separators between them.
+export func join(a *[]string, sep string) string {
+	if len(a) == 0 {
+		return 0
+	}
+	if len(a) == 1 {
+		return a[0]
+	}
+	n := len(sep) * (len(a)-1);
+	for i := 0; i < len(a); i++ {
+		n += len(a[i])
+	}
+
+	b := new([]byte, n);
+	bp := 0;
+	for i := 0; i < len(a); i++ {
+		s := a[i];
+		for j := 0; j < len(s); j++ {
+			b[bp] = s[j];
+			bp++
+		}
+		if i + 1 < len(a) {
+			s = sep;
+			for j := 0; j < len(s); j++ {
+				b[bp] = s[j];
+				bp++
+			}
+		}
+	}
+	return string(b)
+}
+
+// Convert decimal string to integer.
+// TODO: Doesn't check for overflow.
+export func atoi(s string) (i int, ok bool) {
+	// empty string bad
+	if len(s) == 0 { 
+		return 0, false
+	}
+	
+	// pick off leading sign
+	neg := false;
+	if s[0] == '+' {
+		s = s[1:len(s)]
+	} else if s[0] == '-' {
+		neg = true;
+		s = s[1:len(s)]
+	}
+	
+	// empty string bad
+	if len(s) == 0 { 
+		return 0, false
+	}
+
+	// pick off zero
+	if s == "0" {
+		return 0, true
+	}
+	
+	// otherwise, leading zero bad
+	if s[0] == '0' {
+		return 0, false
+	}
+
+	// parse number
+	n := 0;
+	for i := 0; i < len(s); i++ {
+		if s[i] < '0' || s[i] > '9' {
+			return 0, false
+		}
+		n = n*10 + int(s[i] - '0')
+	}
+	if neg {
+		n = -n
+	}
+	return n, true
+}
+
+export func itoa(i int) string {
+	if i == 0 {
+		return "0"
+	}
+	
+	neg := false;	// negative
+	bigneg := false;	// largest negative number
+	if i < 0 {
+		neg = true;
+		i = -i;
+		if i < 0 {
+			bigneg = true;	// is largest negative int
+			i--	// now positive
+		}
+	}
+
+	// Assemble decimal in reverse order.
+	var b [32]byte;
+	bp := len(b);
+	for ; i > 0; i /= 10 {
+		bp--;
+		b[bp] = byte(i%10) + '0'
+	}
+	if neg {	// add sign
+		bp--;
+		b[bp] = '-'
+	}
+	if bigneg {	// account for i-- above
+		b[len(b)-1]++
+	}
+	
+	// BUG return string(b[bp:len(b)])
+	return string((&b)[bp:len(b)])
+}

test/stringslib.go

+// Copyright 2009 The Go Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+// $G $F.go && $L $F.$A && ./$A.out
+
+package main
+
+import strings "strings"
+
+func split(s, sep string) *[]string {
+	a := strings.split(s, sep);
+	b := strings.join(a, sep);
+	if b != s {
+		print("Split: ", s, " ", sep, " got ", len(a), "\n");
+		for i := 0; i < len(a); i++ {
+			print(" a[", i, "] = ", a[i], "\n")
+		}
+		panic("split / join "+s+" "+sep)
+	}
+	return a
+}
+
+func explode(s string) *[]string {
+	a := strings.explode(s);
+	b := strings.join(a, "");
+	if b != s {
+		panic("explode / join "+s)
+	}
+	return a
+}
+
+func itoa(i int) string {
+	s := strings.itoa(i);
+	n, ok := strings.atoi(s);
+	if n != i {
+		print("itoa: ", i, " ", s, "\n");
+		panic("itoa")
+	}
+	return s
+}
+
+func main() {
+	abcd := "abcd";
+	faces := "☺☻☹";
+	commas := "1,2,3,4";
+	dots := "1....2....3....4";
+	if strings.utflen(abcd) != 4 { panic("utflen abcd") }
+	if strings.utflen(faces) != 3 { panic("utflen faces") }
+	if strings.utflen(commas) != 7 { panic("utflen commas") }
+	{
+		a := split(abcd, "a");
+		if len(a) != 2 || a[0] != "" || a[1] != "bcd" { panic("split abcd a") }
+	}
+	{
+		a := split(abcd, "z");
+		if len(a) != 1 || a[0] != "abcd" { panic("split abcd z") }
+	}
+	{
+		a := split(abcd, "");
+		if len(a) != 4 || a[0] != "a" || a[1] != "b" || a[2] != "c" || a[3] != "d" { panic("split abcd empty") }
+	}
+	{
+		a := explode(abcd);
+		if len(a) != 4 || a[0] != "a" || a[1] != "b" || a[2] != "c" || a[3] != "d" { panic("explode abcd") }
+	}
+	{
+		a := split(commas, ",");
+		if len(a) != 4 || a[0] != "1" || a[1] != "2" || a[2] != "3" || a[3] != "4" { panic("split commas") }
+	}
+	{
+		a := split(dots, "...");
+		if len(a) != 4 || a[0] != "1" || a[1] != ".2" || a[2] != ".3" || a[3] != ".4" { panic("split dots") }
+	}
+
+	{
+		a := split(faces, "☹");
+		if len(a) != 2 || a[0] != "☺☻" || a[1] != "" { panic("split faces 1") }
+	}
+	{
+		a := split(faces, "~");
+		if len(a) != 1 || a[0] != faces { panic("split faces ~") }
+	}
+	{
+		a := explode(faces);
+		if len(a) != 3 || a[0] != "☺" || a[1] != "☻" || a[2] != "☹" { panic("explode faces") }
+	}
+	{
+		a := split(faces, "");
+		if len(a) != 3 || a[0] != "☺" || a[1] != "☻" || a[2] != "☹" { panic("split faces empty") }
+	}
+	
+	{
+		n, ok := strings.atoi("0"); if n != 0 || !ok { panic("atoi 0") }
+		n, ok = strings.atoi("-1"); if n != -1 || !ok { panic("atoi -1") }
+		n, ok = strings.atoi("+345"); if n != 345 || !ok { panic("atoi +345") }
+		n, ok = strings.atoi("9999"); if n != 9999 || !ok { panic("atoi 9999") }
+		n, ok = strings.atoi("20ba"); if n != 0 || ok { panic("atoi 20ba") }
+		n, ok = strings.atoi("hello"); if n != 0 || ok { panic("hello") }
+	}
+	
+	if itoa(0) != "0" { panic("itoa 0") }
+	if itoa(12345) != "12345" { panic("itoa 12345") }
+	if itoa(-1<<31) != "-2147483648" { panic("itoa 1<<31") }
+	
+	// should work if int == int64: is there some way to know?
+	// if itoa(-1<<63) != "-9223372036854775808" { panic("itoa 1<<63") }
+}