cmd/internal/str: add utilities for quoting and splitting args

JoinAndQuoteFields does the inverse of SplitQuotedFields: it joins a
list of arguments with spaces into one string, quoting arguments that
contain spaces or quotes.

QuotedStringListFlag uses SplitQuotedFields and JoinAndQuoteFields
together to define new flags that accept lists of arguments.

For #41400

Change-Id: I4986b753cb5e6fabb5b489bf26aedab889f853f5
Reviewed-on: https://go-review.googlesource.com/c/go/+/334731
Trust: Jay Conrod <[email protected]>
Trust: Michael Matloob <[email protected]>
Run-TryBot: Jay Conrod <[email protected]>
TryBot-Result: Go Bot <[email protected]>
Reviewed-by: Bryan C. Mills <[email protected]>
Reviewed-by: Michael Matloob <[email protected]>
Reviewed-on: https://go-review.googlesource.com/c/go/+/341935

Author: Jay Conrod

src/cmd/internal/str/str.go

@@ -7,7 +7,9 @@ package str
 
 import (
 	"bytes"
+	"flag"
 	"fmt"
+	"strings"
 	"unicode"
 	"unicode/utf8"
 )
@@ -153,3 +155,73 @@ func SplitQuotedFields(s string) ([]string, error) {
 	}
 	return f, nil
 }
+
+// JoinAndQuoteFields joins a list of arguments into a string that can be parsed
+// with SplitQuotedFields. Arguments are quoted only if necessary; arguments
+// without spaces or quotes are kept as-is. No argument may contain both
+// single and double quotes.
+func JoinAndQuoteFields(args []string) (string, error) {
+	var buf []byte
+	for i, arg := range args {
+		if i > 0 {
+			buf = append(buf, ' ')
+		}
+		var sawSpace, sawSingleQuote, sawDoubleQuote bool
+		for _, c := range arg {
+			switch {
+			case c > unicode.MaxASCII:
+				continue
+			case isSpaceByte(byte(c)):
+				sawSpace = true
+			case c == '\'':
+				sawSingleQuote = true
+			case c == '"':
+				sawDoubleQuote = true
+			}
+		}
+		switch {
+		case !sawSpace && !sawSingleQuote && !sawDoubleQuote:
+			buf = append(buf, []byte(arg)...)
+
+		case !sawSingleQuote:
+			buf = append(buf, '\'')
+			buf = append(buf, []byte(arg)...)
+			buf = append(buf, '\'')
+
+		case !sawDoubleQuote:
+			buf = append(buf, '"')
+			buf = append(buf, []byte(arg)...)
+			buf = append(buf, '"')
+
+		default:
+			return "", fmt.Errorf("argument %q contains both single and double quotes and cannot be quoted", arg)
+		}
+	}
+	return string(buf), nil
+}
+
+// A QuotedStringListFlag parses a list of string arguments encoded with
+// JoinAndQuoteFields. It is useful for flags like cmd/link's -extldflags.
+type QuotedStringListFlag []string
+
+var _ flag.Value = (*QuotedStringListFlag)(nil)
+
+func (f *QuotedStringListFlag) Set(v string) error {
+	fs, err := SplitQuotedFields(v)
+	if err != nil {
+		return err
+	}
+	*f = fs[:len(fs):len(fs)]
+	return nil
+}
+
+func (f *QuotedStringListFlag) String() string {
+	if f == nil {
+		return ""
+	}
+	s, err := JoinAndQuoteFields(*f)
+	if err != nil {
+		return strings.Join(*f, " ")
+	}
+	return s
+}

src/cmd/internal/str/str_test.go

@@ -4,7 +4,11 @@
 
 package str
 
-import "testing"
+import (
+	"reflect"
+	"strings"
+	"testing"
+)
 
 var foldDupTests = []struct {
 	list   []string
@@ -25,3 +29,80 @@ func TestFoldDup(t *testing.T) {
 		}
 	}
 }
+
+func TestSplitQuotedFields(t *testing.T) {
+	for _, test := range []struct {
+		name    string
+		value   string
+		want    []string
+		wantErr string
+	}{
+		{name: "empty", value: "", want: nil},
+		{name: "space", value: " ", want: nil},
+		{name: "one", value: "a", want: []string{"a"}},
+		{name: "leading_space", value: " a", want: []string{"a"}},
+		{name: "trailing_space", value: "a ", want: []string{"a"}},
+		{name: "two", value: "a b", want: []string{"a", "b"}},
+		{name: "two_multi_space", value: "a  b", want: []string{"a", "b"}},
+		{name: "two_tab", value: "a\tb", want: []string{"a", "b"}},
+		{name: "two_newline", value: "a\nb", want: []string{"a", "b"}},
+		{name: "quote_single", value: `'a b'`, want: []string{"a b"}},
+		{name: "quote_double", value: `"a b"`, want: []string{"a b"}},
+		{name: "quote_both", value: `'a '"b "`, want: []string{"a ", "b "}},
+		{name: "quote_contains", value: `'a "'"'b"`, want: []string{`a "`, `'b`}},
+		{name: "escape", value: `\'`, want: []string{`\'`}},
+		{name: "quote_unclosed", value: `'a`, wantErr: "unterminated ' string"},
+	} {
+		t.Run(test.name, func(t *testing.T) {
+			got, err := SplitQuotedFields(test.value)
+			if err != nil {
+				if test.wantErr == "" {
+					t.Fatalf("unexpected error: %v", err)
+				} else if errMsg := err.Error(); !strings.Contains(errMsg, test.wantErr) {
+					t.Fatalf("error %q does not contain %q", errMsg, test.wantErr)
+				}
+				return
+			}
+			if test.wantErr != "" {
+				t.Fatalf("unexpected success; wanted error containing %q", test.wantErr)
+			}
+			if !reflect.DeepEqual(got, test.want) {
+				t.Errorf("got %q; want %q", got, test.want)
+			}
+		})
+	}
+}
+
+func TestJoinAndQuoteFields(t *testing.T) {
+	for _, test := range []struct {
+		name          string
+		args          []string
+		want, wantErr string
+	}{
+		{name: "empty", args: nil, want: ""},
+		{name: "one", args: []string{"a"}, want: "a"},
+		{name: "two", args: []string{"a", "b"}, want: "a b"},
+		{name: "space", args: []string{"a ", "b"}, want: "'a ' b"},
+		{name: "newline", args: []string{"a\n", "b"}, want: "'a\n' b"},
+		{name: "quote", args: []string{`'a `, "b"}, want: `"'a " b`},
+		{name: "unquoteable", args: []string{`'"`}, wantErr: "contains both single and double quotes and cannot be quoted"},
+	} {
+		t.Run(test.name, func(t *testing.T) {
+			got, err := JoinAndQuoteFields(test.args)
+			if err != nil {
+				if test.wantErr == "" {
+					t.Fatalf("unexpected error: %v", err)
+				} else if errMsg := err.Error(); !strings.Contains(errMsg, test.wantErr) {
+					t.Fatalf("error %q does not contain %q", errMsg, test.wantErr)
+				}
+				return
+			}
+			if test.wantErr != "" {
+				t.Fatalf("unexpected success; wanted error containing %q", test.wantErr)
+			}
+			if got != test.want {
+				t.Errorf("got %s; want %s", got, test.want)
+			}
+		})
+	}
+}