compiler: implement range over a string

compiler.go

@@ -2241,25 +2241,26 @@ func (c *Compiler) parseExpr(frame *Frame, expr ssa.Value) (llvm.Value, error) {
 		slice = c.builder.CreateInsertValue(slice, sliceCap, 2, "")
 		return slice, nil
 	case *ssa.Next:
-		if expr.IsString {
-			return llvm.Value{}, errors.New("todo: next: string")
-		} else { // map
-			fn := c.mod.NamedFunction("runtime.hashmapNext")
+		rangeVal := expr.Iter.(*ssa.Range).X
+		llvmRangeVal, err := c.parseExpr(frame, rangeVal)
+		if err != nil {
+			return llvm.Value{}, err
+		}
 		it, err := c.parseExpr(frame, expr.Iter)
 		if err != nil {
 			return llvm.Value{}, err
 		}
-			rangeMap := expr.Iter.(*ssa.Range).X
-			m, err := c.parseExpr(frame, rangeMap)
-			if err != nil {
-				return llvm.Value{}, err
-			}
+		if expr.IsString {
+			fn := c.mod.NamedFunction("runtime.stringNext")
+			return c.builder.CreateCall(fn, []llvm.Value{llvmRangeVal, it}, "range.next"), nil
+		} else { // map
+			fn := c.mod.NamedFunction("runtime.hashmapNext")
 
-			llvmKeyType, err := c.getLLVMType(rangeMap.Type().(*types.Map).Key())
+			llvmKeyType, err := c.getLLVMType(rangeVal.Type().(*types.Map).Key())
 			if err != nil {
 				return llvm.Value{}, err
 			}
-			llvmValueType, err := c.getLLVMType(rangeMap.Type().(*types.Map).Elem())
+			llvmValueType, err := c.getLLVMType(rangeVal.Type().(*types.Map).Elem())
 			if err != nil {
 				return llvm.Value{}, err
 			}
@@ -2268,7 +2269,7 @@ func (c *Compiler) parseExpr(frame *Frame, expr ssa.Value) (llvm.Value, error) {
 			mapKeyPtr := c.builder.CreateBitCast(mapKeyAlloca, c.i8ptrType, "range.keyptr")
 			mapValueAlloca := c.builder.CreateAlloca(llvmValueType, "range.value")
 			mapValuePtr := c.builder.CreateBitCast(mapValueAlloca, c.i8ptrType, "range.valueptr")
-			ok := c.builder.CreateCall(fn, []llvm.Value{m, it, mapKeyPtr, mapValuePtr}, "range.next")
+			ok := c.builder.CreateCall(fn, []llvm.Value{llvmRangeVal, it, mapKeyPtr, mapValuePtr}, "range.next")
 
 			tuple := llvm.Undef(llvm.StructType([]llvm.Type{llvm.Int1Type(), llvmKeyType, llvmValueType}, false))
 			tuple = c.builder.CreateInsertValue(tuple, ok, 0, "")
@@ -2285,12 +2286,15 @@ func (c *Compiler) parseExpr(frame *Frame, expr ssa.Value) (llvm.Value, error) {
 		frame.phis = append(frame.phis, Phi{expr, phi})
 		return phi, nil
 	case *ssa.Range:
+		var iteratorType llvm.Type
 		switch typ := expr.X.Type().Underlying().(type) {
-		case *types.Basic:
-			// string
-			return llvm.Value{}, errors.New("todo: range: string")
+		case *types.Basic: // string
+			iteratorType = c.mod.GetTypeByName("runtime.stringIterator")
 		case *types.Map:
-			iteratorType := c.mod.GetTypeByName("runtime.hashmapIterator")
+			iteratorType = c.mod.GetTypeByName("runtime.hashmapIterator")
+		default:
+			panic("unknown type in range: " + typ.String())
+		}
 		it := c.builder.CreateAlloca(iteratorType, "range.it")
 		zero, err := getZeroValue(iteratorType)
 		if err != nil {
@@ -2298,9 +2302,6 @@ func (c *Compiler) parseExpr(frame *Frame, expr ssa.Value) (llvm.Value, error) {
 		}
 		c.builder.CreateStore(zero, it)
 		return it, nil
-		default:
-			panic("unknown type in range: " + typ.String())
-		}
 	case *ssa.Slice:
 		if expr.Max != nil {
 			return llvm.Value{}, errors.New("todo: full slice expressions (with max): " + expr.Type().String())

src/runtime/string.go

@@ -12,6 +12,12 @@ type _string struct {
 	length lenType
 }
 
+// The iterator state for a range over a string.
+type stringIterator struct {
+	byteindex  lenType
+	rangeindex lenType
+}
+
 // Return true iff the strings match.
 //go:nobounds
 func stringEqual(x, y string) bool {
@@ -75,6 +81,18 @@ func stringFromUnicode(x rune) _string {
 	return _string{ptr: (*byte)(unsafe.Pointer(&array)), length: length}
 }
 
+// Iterate over a string.
+// Returns (ok, key, value).
+func stringNext(s string, it *stringIterator) (bool, int, rune) {
+	if len(s) <= int(it.byteindex) {
+		return false, 0, 0
+	}
+	r, length := decodeUTF8(s, it.byteindex)
+	it.byteindex += length
+	it.rangeindex += 1
+	return true, int(it.rangeindex), r
+}
+
 // Convert a Unicode code point into an array of bytes and its length.
 func encodeUTF8(x rune) ([4]byte, lenType) {
 	// https://stackoverflow.com/questions/6240055/manually-converting-unicode-codepoints-into-utf-8-and-utf-16
@@ -102,3 +120,31 @@ func encodeUTF8(x rune) ([4]byte, lenType) {
 		return [4]byte{0xef, 0xbf, 0xbd, 0}, 3
 	}
 }
+
+// Decode a single UTF-8 character from a string.
+//go:nobounds
+func decodeUTF8(s string, index lenType) (rune, lenType) {
+	remaining := lenType(len(s)) - index // must be >= 1 before calling this function
+	x := s[index]
+	switch {
+	case x&0x80 == 0x00: // 0xxxxxxx
+		return rune(x), 1
+	case x&0xe0 == 0xc0: // 110xxxxx
+		if remaining < 2 {
+			return 0xfffd, 1
+		}
+		return (rune(x&0x1f) << 6) | (rune(s[index+1]) & 0x3f), 2
+	case x&0xf0 == 0xe0: // 1110xxxx
+		if remaining < 3 {
+			return 0xfffd, 1
+		}
+		return (rune(x&0x0f) << 12) | ((rune(s[index+1]) & 0x3f) << 6) | (rune(s[index+2]) & 0x3f), 3
+	case x&0xf8 == 0xf0: // 11110xxx
+		if remaining < 4 {
+			return 0xfffd, 1
+		}
+		return (rune(x&0x07) << 18) | ((rune(s[index+1]) & 0x3f) << 12) | ((rune(s[index+2]) & 0x3f) << 6) | (rune(s[index+3]) & 0x3f), 4
+	default:
+		return 0xfffd, 1
+	}
+}