From 1481c3785637ad284d025bca80b69d243cc639ab Mon Sep 17 00:00:00 2001
From: Yota Toyama <raviqqe@gmail.com>
Date: Sat, 18 Nov 2017 01:25:49 +0900
Subject: [PATCH] Check if href is a valid URL

---
 main.go      |  8 +++++++-
 main_test.go | 10 ++++++++++
 2 files changed, 17 insertions(+), 1 deletion(-)

diff --git a/main.go b/main.go
index dc1baaf..36ee2cb 100644
--- a/main.go
+++ b/main.go
@@ -4,6 +4,7 @@ import (
 	"fmt"
 	"io/ioutil"
 	"net/http"
+	"net/url"
 	"os"
 	"strings"
 
@@ -61,7 +62,7 @@ func extractUrls(n *html.Node) map[string]bool {
 
 		if n.Type == html.ElementNode && n.Data == "a" {
 			for _, a := range n.Attr {
-				if a.Key == "href" {
+				if a.Key == "href" && isUrl(a.Val) {
 					ss[a.Val] = true
 					break
 				}
@@ -76,6 +77,11 @@ func extractUrls(n *html.Node) map[string]bool {
 	return ss
 }
 
+func isUrl(s string) bool {
+	u, err := url.Parse(s)
+	return err == nil && (u.Scheme == "http" || u.Scheme == "https")
+}
+
 func getArgs() map[string]interface{} {
 	usage := `Link checker for Markdown and HTML
 
diff --git a/main_test.go b/main_test.go
index 30ec969..fc934b7 100644
--- a/main_test.go
+++ b/main_test.go
@@ -47,3 +47,13 @@ func TestUrlParse(t *testing.T) {
 	assert.Equal(t, nil, err)
 	assert.Equal(t, "", u.Scheme)
 }
+
+func TestIsUrl(t *testing.T) {
+	for _, s := range []string{"http://google.com", "https://google.com"} {
+		assert.True(t, isUrl(s))
+	}
+
+	for _, s := range []string{"", "file-path"} {
+		assert.False(t, isUrl(s))
+	}
+}