Initial commitHEADmain

5 files changed, 284 insertions, 0 deletions

diff --git a/go.mod b/go.mod
new file mode 100644
index 0000000..1c5d030
--- /dev/null
+++ b/go.mod
@@ -0,0 +1,5 @@
+module roseh.moe/cmd/rsh
+
+go 1.26.5
+
+require github.com/google/go-cmp v0.7.0
diff --git a/go.sum b/go.sum
new file mode 100644
index 0000000..40e761a
--- /dev/null
+++ b/go.sum
@@ -0,0 +1,2 @@
+github.com/google/go-cmp v0.7.0 h1:wk8382ETsv4JYUZwIsn6YpYiWiBsYLSJiTsyBybVuN8=
+github.com/google/go-cmp v0.7.0/go.mod h1:pXiqmnSA92OHEEa9HXL2W4E7lf9JzCmGVUdgjX3N/iU=
diff --git a/rsh.go b/rsh.go
new file mode 100644
index 0000000..80c52e0
--- /dev/null
+++ b/rsh.go
@@ -0,0 +1,182 @@
+package main
+
+import (
+	"bufio"
+	"fmt"
+	"os"
+)
+
+type lexer struct {
+	insertConcat bool
+}
+
+func (l *lexer) splitTokens(data []byte, atEOF bool) (advance int, token []byte, err error) {
+	n := 0
+Whitespace:
+	for ; n < len(data); n++ {
+		switch data[n] {
+		case ' ', '\t':
+			l.insertConcat = false
+		case '#':
+			i := n + 1
+			for ; i < len(data) && data[i] != '\n'; i++ {
+			}
+			if i < len(data) {
+				return i + 1, []byte(";"), nil
+			}
+			if atEOF {
+				return i, nil, nil
+			}
+			return n, nil, nil
+		case '\\':
+			if n+1 < len(data) && data[n+1] == '\n' {
+				n++
+				l.insertConcat = false
+				continue
+			}
+			fallthrough
+		default:
+			break Whitespace
+		}
+	}
+	data = data[n:]
+	if len(data) == 0 {
+		return n, nil, nil
+	}
+	switch data[0] {
+	case '$':
+		if l.insertConcat {
+			l.insertConcat = false
+			return n, []byte{}, nil
+		}
+		i := 1
+		for ; i < len(data); i++ {
+			if b := data[i]; !(b == '_' || '0' <= b && b <= '9' || 'a' <= b && b <= 'z' || 'A' <= b && b <= 'Z') {
+				break
+			}
+		}
+		if i == len(data) && !atEOF {
+			return n, nil, nil
+		}
+		if i < len(data) && data[i] == '(' {
+			i++
+		} else {
+			l.insertConcat = true
+		}
+		if i == 1 {
+			return 0, nil, fmt.Errorf(`invalid token "$"`)
+		}
+		return n + i, data[:i], nil
+	case '<', '>':
+		if len(data) == 1 && !atEOF {
+			return n, nil, nil
+		}
+		l.insertConcat = false
+		if len(data) > 1 && data[1] == '(' {
+			return n + 2, data[:2], nil
+		}
+		return n + 1, data[:1], nil
+	case '\'':
+		if l.insertConcat {
+			l.insertConcat = false
+			return n, []byte{}, nil
+		}
+		i := 1
+		for ; i < len(data); i++ {
+			if data[i] == '\'' {
+				if i+1 == len(data) && !atEOF {
+					return n, nil, nil
+				}
+				if i+1 == len(data) || data[i+1] != '\'' {
+					break
+				}
+			}
+		}
+		if i == len(data) {
+			if atEOF {
+				return 0, nil, fmt.Errorf("unterminated string")
+			}
+			return n, nil, nil
+		}
+		i++
+		l.insertConcat = true
+		return n + i, data[:i], nil
+	case '\n':
+		l.insertConcat = false
+		return n + 1, []byte(";"), nil
+	case '&':
+		if len(data) == 1 && !atEOF {
+			return n, nil, nil
+		}
+		l.insertConcat = false
+		if len(data) > 1 && data[1] == '&' {
+			return n + 2, data[:2], nil
+		}
+		return n + 1, data[:1], nil
+	case '|':
+		if len(data) == 1 && !atEOF {
+			return n, nil, nil
+		}
+		l.insertConcat = false
+		if len(data) > 1 && data[1] == '|' {
+			return n + 2, data[:2], nil
+		}
+		return n + 1, data[:1], nil
+	case ';':
+		l.insertConcat = false
+		return n + 1, data[:1], nil
+	case ')':
+		l.insertConcat = true
+		return n + 1, data[:1], nil
+	case '*':
+		if l.insertConcat {
+			l.insertConcat = false
+			return n, []byte{}, nil
+		}
+		l.insertConcat = true
+		return n + 1, data[:1], nil
+	default:
+		if l.insertConcat {
+			l.insertConcat = false
+			return n, []byte{}, nil
+		}
+		i := 1
+	Word:
+		for ; i < len(data); i++ {
+			switch data[i] {
+			case '\\':
+				if i+1 == len(data) && !atEOF {
+					return n, nil, nil
+				}
+				if i+1 < len(data) && data[i+1] == '\n' {
+					break Word
+				}
+			case '&', '|', '*', ')', ';', '$', '<', '>', '\'', '#', ' ', '\t', '\n':
+				break Word
+			}
+		}
+		if i == len(data) && !atEOF {
+			return n, nil, nil
+		}
+		l.insertConcat = true
+		return n + i, data[:i], nil
+	}
+}
+
+func splitTokens() bufio.SplitFunc {
+	return new(lexer).splitTokens
+}
+
+func main() {
+	scanner := bufio.NewScanner(os.Stdin)
+	scanner.Split(splitTokens())
+	var tokens []string
+	for scanner.Scan() {
+		tokens = append(tokens, scanner.Text())
+	}
+	if err := scanner.Err(); err != nil {
+		fmt.Fprintln(os.Stderr, err)
+		os.Exit(1)
+	}
+	fmt.Printf("%q\n", tokens)
+}
diff --git a/rsh_test.go b/rsh_test.go
new file mode 100644
index 0000000..988a2d8
--- /dev/null
+++ b/rsh_test.go
@@ -0,0 +1,93 @@
+package main
+
+import (
+	"bufio"
+	"strings"
+	"testing"
+
+	"github.com/google/go-cmp/cmp"
+)
+
+func TestSplitTokens(t *testing.T) {
+	t.Parallel()
+
+	for _, tc := range []struct {
+		desc  string
+		input string
+		want  []string
+	}{{
+		desc:  "SingleWord",
+		input: "a",
+		want:  []string{"a"},
+	}, {
+		desc:  "SplitWords",
+		input: "a b c",
+		want:  []string{"a", "b", "c"},
+	}, {
+		desc:  "Star",
+		input: "a*b",
+		want:  []string{"a", "", "*", "", "b"},
+	}, {
+		desc:  "Background",
+		input: "a&b",
+		want:  []string{"a", "&", "b"},
+	}, {
+		desc:  "Pipe",
+		input: "a|b",
+		want:  []string{"a", "|", "b"},
+	}, {
+		desc:  "And",
+		input: "a&&b",
+		want:  []string{"a", "&&", "b"},
+	}, {
+		desc:  "Newline",
+		input: "a\n",
+		want:  []string{"a", ";"},
+	}, {
+		desc:  "CaptureOutputAsFile",
+		input: "a<(b)",
+		want:  []string{"a", "<(", "b", ")"},
+	}, {
+		desc:  "Substitution",
+		input: "a$(b)c",
+		want:  []string{"a", "", "$(", "b", ")", "", "c"},
+	}, {
+		desc:  "Comment",
+		input: "abc#de;fg",
+		want:  []string{"abc"},
+	}, {
+		desc:  "CommentNewline",
+		input: "a#b\n",
+		want:  []string{"a", ";"},
+	}, {
+		desc:  "BackslashEOF",
+		input: `a\`,
+		want:  []string{`a\`},
+	}} {
+		t.Run(tc.desc, func(t *testing.T) {
+			scanner := bufio.NewScanner(strings.NewReader(tc.input))
+			scanner.Split(splitTokens())
+			var got []string
+			for scanner.Scan() {
+				got = append(got, scanner.Text())
+			}
+			if err := scanner.Err(); err != nil {
+				t.Fatalf("splitTokens failed: %s", err)
+			}
+			if diff := cmp.Diff(tc.want, got); diff != "" {
+				t.Errorf("splitTokens produced unexpected diff (-want +got):\n%s", diff)
+			}
+		})
+	}
+}
+
+func FuzzSplitTokens(f *testing.F) {
+	testcases := []string{"a", "a b c", "a*b", "a&b", "a|b", "a&&b", "a\n", "a<(b)", "a$(b)c", "abc#de;fg", "a#b\n", `a\`}
+	for _, tc := range testcases {
+		f.Add([]byte(tc))
+	}
+	f.Fuzz(func(t *testing.T, input []byte) {
+		splitTokens()(input, true)
+		splitTokens()(input, false)
+	})
+}
diff --git a/testdata/fuzz/FuzzSplitTokens/5e6d51ab3f025967 b/testdata/fuzz/FuzzSplitTokens/5e6d51ab3f025967
new file mode 100644
index 0000000..0f66b13
--- /dev/null
+++ b/testdata/fuzz/FuzzSplitTokens/5e6d51ab3f025967
@@ -0,0 +1,2 @@
+go test fuzz v1
+[]byte("'0000000")