feat: add youtube cmd (video -> mapping_table_row)

```bash
dbctl youtube --youtube-link https://www.youtube.com/watch?v=rtuJqQDWmIA >> ../server/internal/data/mapping_table.pbtxt
```

which will append one row to mapping_table.pbtxt

```prototext
rows:  {
  pr_id:  327
  paper_arxiv_ids:  "2010.14701"
  paper_arxiv_ids:  "2001.08361"
  paper_arxiv_ids:  "2106.09488"
  youtube_video_id:  "2irqTp-3hQ0"
}
```

dbctl/cmd/youtube.go

@@ -0,0 +1,118 @@
+package cmd
+
+import (
+	"fmt"
+	"time"
+
+	"github.com/codingpot/pr12er/dbctl/internal/transform"
+	"github.com/codingpot/pr12er/server/pkg/pr12er"
+	log "github.com/sirupsen/logrus"
+	"github.com/spf13/cobra"
+	"github.com/spf13/viper"
+	"golang.org/x/time/rate"
+	"google.golang.org/protobuf/encoding/prototext"
+)
+
+const envNameYouTubeLink = "YOUTUBE_LINK"
+
+// youtubeCmd represents the youtube command
+var youtubeCmd = &cobra.Command{
+	Use:   "youtube",
+	Short: "Generate a single GenMapping to STDOUT",
+	Long: `You can run this file and redirect to the mapping file.
+
+For example,
+	go run main.go youtube --youtube-link https://www.youtube.com/watch?v=rtuJqQDWmIA >> ../server/internal/data/mapping_table.pbtxt
+`,
+	RunE: func(cmd *cobra.Command, args []string) error {
+		apiKey := viper.GetString(envNameYouTubeAPIKey)
+		youTubeLink := viper.GetString(envNameYouTubeLink)
+		cx := viper.GetString(envNameCx)
+
+		log.WithFields(log.Fields{
+			envNameYouTubeAPIKey: apiKey,
+			envNameYouTubeLink:   youTubeLink,
+			envNameCx:            cx,
+		}).Info("binding variables")
+
+		rateLimiter := rate.NewLimiter(rate.Every(time.Second), 1)
+
+		youtubeID, err := transform.ExtractYouTubeID(youTubeLink)
+		if err != nil {
+			return err
+		}
+
+		client, err := fetcherClient(apiKey)
+		if err != nil {
+			return err
+		}
+
+		// Get video metadata.
+		youTubeVideos, err := client.FetchMultiYouTubeVideoByIDs([]string{youtubeID})
+		if err != nil {
+			return err
+		}
+
+		if len(youTubeVideos) != 1 {
+			return fmt.Errorf("expected one video but got %v", youTubeVideos)
+		}
+
+		vid := youTubeVideos[0]
+
+		// Get PR ID from video.
+		prID, err := transform.ExtractPRID(vid.GetVideoTitle())
+		if err != nil {
+			return err
+		}
+
+		var paperIDs []string
+
+		// Get paper IDs.
+		if cx == "" {
+			paperIDs, err = transform.ExtractPaperIDs(vid.GetVideoTitle())
+		} else {
+			paperIDs, err = transform.
+				ExtractPaperIDsViaProgrammableSearch(
+					vid.GetVideoTitle(),
+					cx,
+					apiKey,
+					rateLimiter)
+		}
+		if err != nil {
+			return err
+		}
+
+		data := &pr12er.MappingTable{
+			Rows: []*pr12er.MappingTableRow{
+				{
+					PrId:           prID,
+					PaperArxivIds:  paperIDs,
+					YoutubeVideoId: vid.GetVideoId(),
+				},
+			},
+		}
+
+		bs, err := prototext.MarshalOptions{
+			Multiline: true,
+			Indent:    "  ",
+		}.Marshal(data)
+		if err != nil {
+			return err
+		}
+
+		// Print to stdout
+		fmt.Print(string(bs))
+
+		return nil
+	},
+}
+
+func init() {
+	rootCmd.AddCommand(youtubeCmd)
+	youtubeCmd.Flags().String("youtube-api-key", "", "YouTube API Key (required)")
+	_ = viper.BindPFlag(envNameYouTubeAPIKey, youtubeCmd.Flag("youtube-api-key"))
+	youtubeCmd.Flags().String("youtube-link", "", "YouTube Link to fetch")
+	_ = viper.BindPFlag(envNameYouTubeLink, youtubeCmd.Flag("youtube-link"))
+	youtubeCmd.Flags().String("cx", "", "Search Engine ID from https://programmablesearchengine.google.com/ If this is not set, it will use googlesearch (free)")
+	_ = viper.BindPFlag(envNameCx, youtubeCmd.Flag("cx"))
+}

dbctl/internal/transform/transform.go

@@ -4,6 +4,7 @@ package transform
 import (
 	"context"
 	"fmt"
+	"net/url"
 	"regexp"
 	"strconv"
 	"strings"
@@ -171,3 +172,45 @@ func ExtractPRID(title string) (int32, error) {
 	}
 	return int32(atoi), nil
 }
+
+type ErrNotYouTubeLink struct {
+	url string
+}
+
+func (e ErrNotYouTubeLink) Error() string {
+	return fmt.Sprintf("no valid YouTubeID is found in %s", e.url)
+}
+
+// ExtractYouTubeID extracts videoID from YouTube link
+//
+// For example,
+//
+// https://www.youtube.com/watch?v=rtuJqQDWmIA => rtuJqQDWmIA
+// https://youtube.com/watch?v=rtuJqQDWmIA => rtuJqQDWmIA
+// https://youtu.be/rtuJqQDWmIA => rtuJqQDWmIA
+func ExtractYouTubeID(link string) (string, error) {
+	parse, err := url.Parse(link)
+	if err != nil {
+		return "", err
+	}
+
+	errNotYouTubeLink := ErrNotYouTubeLink{link}
+
+	if strings.Contains(parse.Hostname(), "youtube") {
+		youtubeID := parse.Query().Get("v")
+		if youtubeID == "" {
+			return "", errNotYouTubeLink
+		}
+		return youtubeID, nil
+	}
+
+	if strings.Contains(parse.Hostname(), "youtu.be") {
+		youtubeID := strings.TrimPrefix(parse.Path, "/")
+		if youtubeID == "" {
+			return "", errNotYouTubeLink
+		}
+		return youtubeID, nil
+	}
+
+	return "", errNotYouTubeLink
+}

dbctl/internal/transform/transform_test.go

@@ -203,3 +203,59 @@ func TestExtractArxivIDFromURL(t *testing.T) {
 		})
 	}
 }
+
+func TestExtractYouTubeID(t *testing.T) {
+	type args struct {
+		link string
+	}
+	tests := []struct {
+		name    string
+		args    args
+		want    string
+		wantErr bool
+	}{
+		{
+			name: "https://www.youtube.com/watch?v=rtuJqQDWmIA => rtuJqQDWmIA",
+			args: args{
+				link: "https://www.youtube.com/watch?v=rtuJqQDWmIA",
+			},
+			want:    "rtuJqQDWmIA",
+			wantErr: false,
+		},
+		{
+			name: "https://youtube.com/watch?v=rtuJqQDWmIA => rtuJqQDWmIA",
+			args: args{
+				link: "https://youtube.com/watch?v=rtuJqQDWmIA",
+			},
+			want:    "rtuJqQDWmIA",
+			wantErr: false,
+		},
+		{
+			name: "https://youtu.be/rtuJqQDWmIA => rtuJqQDWmIA",
+			args: args{
+				link: "https://youtu.be/rtuJqQDWmIA",
+			},
+			want:    "rtuJqQDWmIA",
+			wantErr: false,
+		},
+		{
+			name: "invalid url returns an error",
+			args: args{
+				link: "www.gooogle.com",
+			},
+			want:    "",
+			wantErr: true,
+		},
+	}
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			got, err := ExtractYouTubeID(tt.args.link)
+			if tt.wantErr {
+				assert.Error(t, err)
+			} else {
+				assert.NoError(t, err)
+				assert.Equal(t, tt.want, got)
+			}
+		})
+	}
+}