add IsPunctuation2() that also works on unicode punctuation (fixes #309)

inline_test.go

@@ -15,10 +15,7 @@ func TestEmphasis(t *testing.T) {
 	doTestsInlineParam(t, tests, TestParams{})
 }
 
-// TODO: to fix this we would have to update IsPunctuation() to handle
-// `—` which looks like `-` but is a unicode 3-byte thingy and
-// currently IsPunctuation() only handles 1-byte ascii
-func Disabled_TestBug309(t *testing.T) {
+func TestBug309(t *testing.T) {
 	var tests = []string{
 		`*f*—`,
 		"<p><em>f</em>—</p>\n",

parser/inline.go

@@ -1208,7 +1208,8 @@ func helperEmphasis(p *Parser, data []byte, c byte) (int, ast.Node) {
 		if data[i] == c && !IsSpace(data[i-1]) {
 
 			if p.extensions&NoIntraEmphasis != 0 {
-				if !(i+1 == len(data) || IsSpace(data[i+1]) || IsPunctuation(data[i+1])) {
+				rest := data[i+1:]
+				if !(len(rest) == 0 || IsSpace(rest[0]) || IsPunctuation2(rest)) {
 					continue
 				}
 			}

parser/parser.go

@@ -8,6 +8,8 @@ import (
 	"fmt"
 	"strconv"
 	"strings"
+	"unicode"
+	"unicode/utf8"
 
 	"github.com/gomarkdown/markdown/ast"
 )
@@ -727,6 +729,20 @@ func IsPunctuation(c byte) bool {
 	return false
 }
 
+func IsPunctuation2(d []byte) bool {
+	if len(d) == 0 {
+		return false
+	}
+	if IsPunctuation(d[0]) {
+		return true
+	}
+	r, _ := utf8.DecodeRune(d)
+	if r == utf8.RuneError {
+		return false
+	}
+	return unicode.IsPunct(r)
+}
+
 // IsSpace returns true if c is a white-space charactr
 func IsSpace(c byte) bool {
 	return c == ' ' || c == '\t' || c == '\n' || c == '\r' || c == '\f' || c == '\v'