Normalize url

src/media/fetch.go

@@ -6,6 +6,7 @@ import (
 	"fmt"
 	"github.com/dustin/go-humanize"
 	"html/template"
+	"media-roller/src/utils"
 	"net/http"
 	"path/filepath"
 	"regexp"
@@ -94,6 +95,8 @@ func getMediaResults(url string) ([]Media, string, error) {
 		return nil, "", errors.New("missing URL")
 	}
 
+	url = utils.NormalizeUrl(url)
+
 	// NOTE: This system is for a simple use case, meant to run at home. This is not a great design for a robust system.
 	// We are hashing the URL here and writing files to disk to a consistent directory based on the ID. You can imagine
 	// concurrent users would break this for the same URL. That's fine given this is for a simple home system.
@@ -128,7 +131,7 @@ func downloadMedia(url string) (string, string, error) {
 	cmd := exec.Command("yt-dlp",
 		"--format", "bestvideo[ext=mp4]+bestaudio[ext=m4a]/best[ext=mp4]/best",
 		"--merge-output-format", "mp4",
-	        "--trim-filenames", "200",
+		"--trim-filenames", "200",
 		"--restrict-filenames",
 		"--write-info-json",
 		"--verbose",

src/utils/urls.go

@@ -0,0 +1,18 @@
+package utils
+
+import "strings"
+
+func NormalizeUrl(url string) string {
+	url = strings.TrimSpace(url)
+	parts := strings.Split(url, " ")
+
+	for _, part := range parts {
+		// Take the firs string that looks like a URL.
+		// TODO: We could try to parse the url, but will save that for later
+		if strings.HasPrefix(part, "http") || strings.HasPrefix(part, "www") {
+			return part
+		}
+	}
+
+	return url
+}

src/utils/urls_test.go

@@ -0,0 +1,21 @@
+package utils
+
+import "testing"
+
+func TestNormalizeUrl(t *testing.T) {
+	tests := []struct {
+		url  string
+		want string
+	}{
+		{url: "example.com", want: "example.com"},
+		{url: "https://example.com", want: "https://example.com"},
+		{url: "https://example.com this is an example", want: "https://example.com"},
+	}
+	for _, tt := range tests {
+		t.Run(tt.url, func(t *testing.T) {
+			if got := NormalizeUrl(tt.url); got != tt.want {
+				t.Errorf("NormalizeUrl() = %v, want %v", got, tt.want)
+			}
+		})
+	}
+}