lookbook/internal/embed/detect.go

package embed

import (
	"context"
	"encoding/json"
	"fmt"
	"net/http"
	"net/url"
	"regexp"
	"strings"
	"time"
)

// Provider identifies the video hosting platform.
type Provider string

const (
	ProviderYouTube Provider = "youtube"
	ProviderVimeo   Provider = "vimeo"
	ProviderTwitter Provider = "twitter"
)

// VideoInfo contains information about an embedded video.
type VideoInfo struct {
	Provider      Provider
	VideoID       string
	Title         string
	Description   string
	ThumbnailURL  string   // First/primary thumbnail (for backward compatibility)
	ThumbnailURLs []string // All thumbnail URLs (for multi-image tweets)
	EmbedHTML     string
	VideoURL      string // Direct video URL (for Twitter videos)
}

var (
	youtubeRegex = regexp.MustCompile(`(?:youtube\.com/(?:watch\?v=|embed/|v/|shorts/)|youtu\.be/)([a-zA-Z0-9_-]{11})`)
	vimeoRegex   = regexp.MustCompile(`(?:vimeo\.com/(?:video/)?|player\.vimeo\.com/video/)(\d+)`)
	twitterRegex = regexp.MustCompile(`(?:twitter\.com|x\.com)/([^/]+)/status/(\d+)`)
	tcoRegex     = regexp.MustCompile(`\s*https://t\.co/\S+`)
)

// Detect checks if a URL is a YouTube, Vimeo, or Twitter/X post and returns its info.
func Detect(ctx context.Context, targetURL string) (*VideoInfo, error) {
	// Try YouTube
	if matches := youtubeRegex.FindStringSubmatch(targetURL); len(matches) > 1 {
		return fetchYouTube(ctx, matches[1])
	}

	// Try Vimeo
	if matches := vimeoRegex.FindStringSubmatch(targetURL); len(matches) > 1 {
		return fetchVimeo(ctx, matches[1])
	}

	// Try Twitter/X
	if matches := twitterRegex.FindStringSubmatch(targetURL); len(matches) > 2 {
		return fetchTwitter(ctx, matches[2], targetURL)
	}

	return nil, nil // Not a recognized embed
}

func fetchYouTube(ctx context.Context, videoID string) (*VideoInfo, error) {
	// YouTube thumbnails are available without API
	thumbnailURL := fmt.Sprintf("https://img.youtube.com/vi/%s/maxresdefault.jpg", videoID)

	// Try to get metadata via oEmbed
	oembedURL := fmt.Sprintf("https://www.youtube.com/oembed?url=%s&format=json",
		url.QueryEscape("https://www.youtube.com/watch?v="+videoID))

	var title string
	if meta, err := fetchOEmbed(ctx, oembedURL); err == nil {
		title = meta.Title
	}

	embedHTML := fmt.Sprintf(
		`<iframe width="560" height="315" src="https://www.youtube.com/embed/%s" frameborder="0" allowfullscreen></iframe>`,
		videoID,
	)

	return &VideoInfo{
		Provider:     ProviderYouTube,
		VideoID:      videoID,
		Title:        title,
		ThumbnailURL: thumbnailURL,
		EmbedHTML:    embedHTML,
	}, nil
}

func fetchVimeo(ctx context.Context, videoID string) (*VideoInfo, error) {
	oembedURL := fmt.Sprintf("https://vimeo.com/api/oembed.json?url=%s",
		url.QueryEscape("https://vimeo.com/"+videoID))

	meta, err := fetchOEmbed(ctx, oembedURL)
	if err != nil {
		return nil, fmt.Errorf("vimeo oembed: %w", err)
	}

	embedHTML := fmt.Sprintf(
		`<iframe src="https://player.vimeo.com/video/%s" width="640" height="360" frameborder="0" allowfullscreen></iframe>`,
		videoID,
	)

	return &VideoInfo{
		Provider:     ProviderVimeo,
		VideoID:      videoID,
		Title:        meta.Title,
		Description:  meta.Description,
		ThumbnailURL: meta.ThumbnailURL,
		EmbedHTML:    embedHTML,
	}, nil
}

// twitterSyndicationResponse represents the Twitter syndication API response
type twitterSyndicationResponse struct {
	Text string `json:"text"`
	User struct {
		Name       string `json:"name"`
		ScreenName string `json:"screen_name"`
	} `json:"user"`
	Photos []struct {
		URL    string `json:"url"`
		Width  int    `json:"width"`
		Height int    `json:"height"`
	} `json:"photos"`
	MediaDetails []struct {
		MediaURLHTTPS string `json:"media_url_https"`
		Type          string `json:"type"`
		VideoInfo     struct {
			Variants []struct {
				ContentType string `json:"content_type"`
				URL         string `json:"url"`
				Bitrate     int    `json:"bitrate,omitempty"`
			} `json:"variants"`
		} `json:"video_info"`
	} `json:"mediaDetails"`
	Video struct {
		Poster string `json:"poster"`
	} `json:"video"`
}

func fetchTwitter(ctx context.Context, tweetID string, originalURL string) (*VideoInfo, error) {
	apiURL := fmt.Sprintf("https://cdn.syndication.twimg.com/tweet-result?id=%s&token=0", tweetID)

	req, err := http.NewRequestWithContext(ctx, "GET", apiURL, nil)
	if err != nil {
		return nil, err
	}
	req.Header.Set("User-Agent", "Mozilla/5.0 (compatible; Lookbook/1.0)")

	client := &http.Client{Timeout: 10 * time.Second}
	resp, err := client.Do(req)
	if err != nil {
		return nil, fmt.Errorf("twitter syndication: %w", err)
	}
	defer resp.Body.Close()

	if resp.StatusCode != http.StatusOK {
		return nil, fmt.Errorf("twitter syndication status: %d", resp.StatusCode)
	}

	var tweet twitterSyndicationResponse
	if err := json.NewDecoder(resp.Body).Decode(&tweet); err != nil {
		return nil, fmt.Errorf("twitter syndication decode: %w", err)
	}

	// Find thumbnail and video URL from media
	var thumbnailURL, videoURL string
	var thumbnailURLs []string
	if len(tweet.Photos) > 0 {
		thumbnailURL = tweet.Photos[0].URL
		// Collect all photo URLs for multi-image tweets
		for _, photo := range tweet.Photos {
			thumbnailURLs = append(thumbnailURLs, photo.URL)
		}
	} else if len(tweet.MediaDetails) > 0 {
		media := tweet.MediaDetails[0]
		thumbnailURL = media.MediaURLHTTPS

		// Extract video URL - find highest bitrate MP4
		if media.Type == "video" || media.Type == "animated_gif" {
			var bestBitrate int
			for _, v := range media.VideoInfo.Variants {
				if v.ContentType == "video/mp4" && v.Bitrate >= bestBitrate {
					bestBitrate = v.Bitrate
					videoURL = v.URL
				}
			}
		}
	}
	if thumbnailURL == "" && tweet.Video.Poster != "" {
		thumbnailURL = tweet.Video.Poster
	}

	// Build embed HTML using Twitter's embed widget
	embedHTML := fmt.Sprintf(
		`<blockquote class="twitter-tweet"><a href="%s"></a></blockquote><script async src="https://platform.twitter.com/widgets.js"></script>`,
		originalURL,
	)

	title := fmt.Sprintf("@%s", tweet.User.ScreenName)
	if tweet.User.Name != "" {
		title = fmt.Sprintf("%s (@%s)", tweet.User.Name, tweet.User.ScreenName)
	}

	// Clean up tweet text - remove trailing t.co URLs
	description := tcoRegex.ReplaceAllString(tweet.Text, "")
	description = strings.TrimSpace(description)

	return &VideoInfo{
		Provider:      ProviderTwitter,
		VideoID:       tweetID,
		Title:         title,
		Description:   description,
		ThumbnailURL:  thumbnailURL,
		ThumbnailURLs: thumbnailURLs,
		VideoURL:      videoURL,
		EmbedHTML:     embedHTML,
	}, nil
}

type oembedResponse struct {
	Title        string `json:"title"`
	Description  string `json:"description"`
	ThumbnailURL string `json:"thumbnail_url"`
}

func fetchOEmbed(ctx context.Context, oembedURL string) (*oembedResponse, error) {
	req, err := http.NewRequestWithContext(ctx, "GET", oembedURL, nil)
	if err != nil {
		return nil, err
	}

	client := &http.Client{Timeout: 10 * time.Second}
	resp, err := client.Do(req)
	if err != nil {
		return nil, err
	}
	defer resp.Body.Close()

	if resp.StatusCode != http.StatusOK {
		return nil, fmt.Errorf("oembed status: %d", resp.StatusCode)
	}

	var meta oembedResponse
	if err := json.NewDecoder(resp.Body).Decode(&meta); err != nil {
		return nil, err
	}

	return &meta, nil
}

// DownloadThumbnail downloads the thumbnail image for a video.
func DownloadThumbnail(ctx context.Context, thumbnailURL string) ([]byte, string, error) {
	req, err := http.NewRequestWithContext(ctx, "GET", thumbnailURL, nil)
	if err != nil {
		return nil, "", err
	}
	req.Header.Set("User-Agent", "Mozilla/5.0 (compatible; Lookbook/1.0)")

	client := &http.Client{Timeout: 30 * time.Second}
	resp, err := client.Do(req)
	if err != nil {
		return nil, "", err
	}
	defer resp.Body.Close()

	if resp.StatusCode != http.StatusOK {
		return nil, "", fmt.Errorf("thumbnail status: %d", resp.StatusCode)
	}

	contentType := resp.Header.Get("Content-Type")

	data := make([]byte, 0, 1<<20) // 1MB initial capacity
	buf := make([]byte, 32*1024)
	for {
		n, err := resp.Body.Read(buf)
		if n > 0 {
			data = append(data, buf[:n]...)
		}
		if err != nil {
			break
		}
	}

	return data, contentType, nil
}