Merge branch 'main' into use-host-header-option

Author: wxiaoguang

modules/markup/html.go

@@ -71,7 +71,8 @@ var globalVars = sync.OnceValue(func() *globalVarsType {
 	// it is still accepted by the CommonMark specification, as well as the HTML5 spec:
 	//   http://spec.commonmark.org/0.28/#email-address
 	//   https://html.spec.whatwg.org/multipage/input.html#e-mail-state-(type%3Demail)
-	v.emailRegex = regexp.MustCompile("(?:\\s|^|\\(|\\[)([a-zA-Z0-9.!#$%&'*+\\/=?^_`{|}~-]+@[a-zA-Z0-9](?:[a-zA-Z0-9-]{0,61}[a-zA-Z0-9])?(?:\\.[a-zA-Z0-9]{2,}(?:[a-zA-Z0-9-]{0,61}[a-zA-Z0-9])?)+)(?:\\s|$|\\)|\\]|;|,|\\?|!|\\.(\\s|$))")
+	// At the moment, we use stricter rule for rendering purpose: only allow the "name" part starting after the word boundary
+	v.emailRegex = regexp.MustCompile(`\b([-\w.!#$%&'*+/=?^{|}~]*@[a-zA-Z0-9](?:[a-zA-Z0-9-]{0,61}[a-zA-Z0-9])?(?:\.[a-zA-Z0-9]{2,}(?:[a-zA-Z0-9-]{0,61}[a-zA-Z0-9])?)+)\b`)
 
 	// emojiShortCodeRegex find emoji by alias like :smile:
 	v.emojiShortCodeRegex = regexp.MustCompile(`:[-+\w]+:`)

modules/markup/html_email.go

@@ -3,7 +3,11 @@
 
 package markup
 
-import "golang.org/x/net/html"
+import (
+	"strings"
+
+	"golang.org/x/net/html"
+)
 
 // emailAddressProcessor replaces raw email addresses with a mailto: link.
 func emailAddressProcessor(ctx *RenderContext, node *html.Node) {
@@ -14,6 +18,14 @@ func emailAddressProcessor(ctx *RenderContext, node *html.Node) {
 			return
 		}
 
+		var nextByte byte
+		if len(node.Data) > m[3] {
+			nextByte = node.Data[m[3]]
+		}
+		if strings.IndexByte(":/", nextByte) != -1 {
+			// for cases: "[email protected]:owner/repo.git", "https://[email protected]/owner/repo.git"
+			return
+		}
 		mail := node.Data[m[2]:m[3]]
 		replaceContent(node, m[2], m[3], createLink(ctx, "mailto:"+mail, mail, "" /*mailto*/))
 		node = node.NextSibling.NextSibling

modules/markup/html_test.go

@@ -225,10 +225,10 @@ func TestRender_email(t *testing.T) {
 	test := func(input, expected string) {
 		res, err := markup.RenderString(markup.NewTestRenderContext().WithRelativePath("a.md"), input)
 		assert.NoError(t, err)
-		assert.Equal(t, strings.TrimSpace(expected), strings.TrimSpace(res))
+		assert.Equal(t, strings.TrimSpace(expected), strings.TrimSpace(res), "input: %s", input)
 	}
-	// Text that should be turned into email link
 
+	// Text that should be turned into email link
 	test(
 		"[email protected]",
 		`<p><a href="mailto:[email protected]" rel="nofollow">[email protected]</a></p>`)
@@ -260,28 +260,48 @@ func TestRender_email(t *testing.T) {
 <a href="mailto:[email protected]" rel="nofollow">[email protected]</a>?
 <a href="mailto:[email protected]" rel="nofollow">[email protected]</a>!</p>`)
 
+	// match GitHub behavior
+	test("email@[email protected]", `<p>email@<a href="mailto:[email protected]" rel="nofollow">[email protected]</a></p>`)
+
+	// match GitHub behavior
+	test(`"[email protected]"`, `<p>&#34;<a href="mailto:[email protected]" rel="nofollow">[email protected]</a>&#34;</p>`)
+
 	// Test that should *not* be turned into email links
-	test(
-		"\"[email protected]\"",
-		`<p>&#34;[email protected]&#34;</p>`)
 	test(
 		"/home/gitea/mailstore/info@gitea/com",
 		`<p>/home/gitea/mailstore/info@gitea/com</p>`)
 	test(
 		"[email protected]:go-gitea/gitea.git",
 		`<p>[email protected]:go-gitea/gitea.git</p>`)
+	test(
+		"https://foo:[email protected]",
+		`<p><a href="https://foo:[email protected]" rel="nofollow">https://foo:[email protected]</a></p>`)
 	test(
 		"gitea@3",
 		`<p>gitea@3</p>`)
 	test(
 		"[email protected]",
 		`<p>[email protected]</p>`)
-	test(
-		"email@[email protected]",
-		`<p>email@[email protected]</p>`)
 	test(
 		"[email protected]",
 		`<p>[email protected]</p>`)
+
+	cases := []struct {
+		input, expected string
+	}{
+		// match GitHub behavior
+		{"[email protected]", `<p>?<a href="mailto:[email protected]" rel="nofollow">[email protected]</a></p>`},
+		{"*[email protected]", `<p>*<a href="mailto:[email protected]" rel="nofollow">[email protected]</a></p>`},
+		{"[email protected]", `<p>~<a href="mailto:[email protected]" rel="nofollow">[email protected]</a></p>`},
+
+		// the following cases don't match GitHub behavior, but they are valid email addresses ...
+		// maybe we should reduce the candidate characters for the "name" part in the future
+		{"a*[email protected]", `<p><a href="mailto:a*[email protected]" rel="nofollow">a*[email protected]</a></p>`},
+		{"[email protected]", `<p><a href="mailto:[email protected]" rel="nofollow">[email protected]</a></p>`},
+	}
+	for _, c := range cases {
+		test(c.input, c.expected)
+	}
 }
 
 func TestRender_emoji(t *testing.T) {