regexp: add SubexpNames

Fixes #2440.

R=r, dsymonds
CC=golang-dev
https://golang.org/cl/5559043

src/pkg/regexp/all_test.go

@@ -289,30 +289,45 @@ func TestLiteralPrefix(t *testing.T) {
 	}
 }
 
-type numSubexpCase struct {
-	input    string
-	expected int
+type subexpCase struct {
+	input string
+	num   int
+	names []string
 }
 
-var numSubexpCases = []numSubexpCase{
-	{``, 0},
-	{`.*`, 0},
-	{`abba`, 0},
-	{`ab(b)a`, 1},
-	{`ab(.*)a`, 1},
-	{`(.*)ab(.*)a`, 2},
-	{`(.*)(ab)(.*)a`, 3},
-	{`(.*)((a)b)(.*)a`, 4},
-	{`(.*)(\(ab)(.*)a`, 3},
-	{`(.*)(\(a\)b)(.*)a`, 3},
+var subexpCases = []subexpCase{
+	{``, 0, nil},
+	{`.*`, 0, nil},
+	{`abba`, 0, nil},
+	{`ab(b)a`, 1, []string{"", ""}},
+	{`ab(.*)a`, 1, []string{"", ""}},
+	{`(.*)ab(.*)a`, 2, []string{"", "", ""}},
+	{`(.*)(ab)(.*)a`, 3, []string{"", "", "", ""}},
+	{`(.*)((a)b)(.*)a`, 4, []string{"", "", "", "", ""}},
+	{`(.*)(\(ab)(.*)a`, 3, []string{"", "", "", ""}},
+	{`(.*)(\(a\)b)(.*)a`, 3, []string{"", "", "", ""}},
+	{`(?P<foo>.*)(?P<bar>(a)b)(?P<foo>.*)a`, 4, []string{"", "foo", "bar", "", "foo"}},
 }
 
-func TestNumSubexp(t *testing.T) {
-	for _, c := range numSubexpCases {
+func TestSubexp(t *testing.T) {
+	for _, c := range subexpCases {
 		re := MustCompile(c.input)
 		n := re.NumSubexp()
-		if n != c.expected {
-			t.Errorf("NumSubexp for %q returned %d, expected %d", c.input, n, c.expected)
+		if n != c.num {
+			t.Errorf("%q: NumSubexp = %d, want %d", c.input, n, c.num)
+			continue
+		}
+		names := re.SubexpNames()
+		if len(names) != 1+n {
+			t.Errorf("%q: len(SubexpNames) = %d, want %d", c.input, len(names), n)
+			continue
+		}
+		if c.names != nil {
+			for i := 0; i < 1+n; i++ {
+				if names[i] != c.names[i] {
+					t.Errorf("%q: SubexpNames[%d] = %q, want %q", c.input, i, names[i], c.names[i])
+				}
+			}
 		}
 	}
 }

src/pkg/regexp/regexp.go

@@ -85,6 +85,7 @@ type Regexp struct {
 	prefixRune     rune           // first rune in prefix
 	cond           syntax.EmptyOp // empty-width conditions required at start of match
 	numSubexp      int
+	subexpNames    []string
 	longest        bool
 
 	// cache of machines for running regexp
@@ -140,17 +141,20 @@ func compile(expr string, mode syntax.Flags, longest bool) (*Regexp, error) {
 		return nil, err
 	}
 	maxCap := re.MaxCap()
+	capNames := re.CapNames()
+
 	re = re.Simplify()
 	prog, err := syntax.Compile(re)
 	if err != nil {
 		return nil, err
 	}
 	regexp := &Regexp{
-		expr:      expr,
-		prog:      prog,
-		numSubexp: maxCap,
-		cond:      prog.StartCond(),
-		longest:   longest,
+		expr:        expr,
+		prog:        prog,
+		numSubexp:   maxCap,
+		subexpNames: capNames,
+		cond:        prog.StartCond(),
+		longest:     longest,
 	}
 	regexp.prefix, regexp.prefixComplete = prog.Prefix()
 	if regexp.prefix != "" {
@@ -223,6 +227,15 @@ func (re *Regexp) NumSubexp() int {
 	return re.numSubexp
 }
 
+// SubexpNames returns the names of the parenthesized subexpressions
+// in this Regexp.  The name for the first sub-expression is names[1],
+// so that if m is a match slice, the name for m[i] is SubexpNames()[i].
+// Since the Regexp as a whole cannot be named, names[0] is always
+// the empty string.  The slice should not be modified.
+func (re *Regexp) SubexpNames() []string {
+	return re.subexpNames
+}
+
 const endOfText rune = -1
 
 // input abstracts different representations of the input text. It provides

src/pkg/regexp/syntax/regexp.go

@@ -303,3 +303,19 @@ func (re *Regexp) MaxCap() int {
 	}
 	return m
 }
+
+// CapNames walks the regexp to find the names of capturing groups.
+func (re *Regexp) CapNames() []string {
+	names := make([]string, re.MaxCap()+1)
+	re.capNames(names)
+	return names
+}
+
+func (re *Regexp) capNames(names []string) {
+	if re.Op == OpCapture {
+		names[re.Cap] = re.Name
+	}
+	for _, sub := range re.Sub {
+		sub.capNames(names)
+	}
+}