av1dae/internal/metadata/metadata.go

package metadata

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

	"videnc-vibe/internal/logger"
	"videnc-vibe/pkg/types"
)

const (
	omdbAPIURL   = "https://www.omdbapi.com/"
	tvmazeAPIURL = "https://api.tvmaze.com"
)

var (
	imdbRegex      = regexp.MustCompile(`tt(\d+)`)
	tvmRegex       = regexp.MustCompile(`(?i)tvm(\d+)`)
	seRegex        = regexp.MustCompile(`(?i)s(\d+)e(\d+)`)
	mediaTypeRegex = regexp.MustCompile(`(?i)\b(dvd|bluray|webdl|tvrip)\b`)
)

type OMDbResponse struct {
	Title    string `json:"Title"`
	Year     string `json:"Released"`
	IMDBID   string `json:"imdbID"`
	Response string `json:"Response"`
	Error    string `json:"Error"`
}

type TVMazeShowResponse struct {
	Name   string `json:"name"`
	Extern struct {
		IMDB string `json:"imdb"`
	} `json:"externals"`
}

type TVMazeEpisode struct {
	Name    string `json:"name"`
	Season  int    `json:"season"`
	Number  int    `json:"number"`
	Airdate string `json:"airdate"`
}

type Client struct {
	omdbAPIKey string
	httpClient *http.Client
	log        *logger.Logger
}

func NewClient(apiKey string, log *logger.Logger) *Client {
	return &Client{
		omdbAPIKey: apiKey,
		httpClient: &http.Client{Timeout: 10 * time.Second},
		log:        log,
	}
}

// redactURL strips secret query parameters (e.g. apikey) before logging.
func redactURL(rawURL string) string {
	u, err := url.Parse(rawURL)
	if err != nil {
		return rawURL
	}
	q := u.Query()
	for _, k := range []string{"apikey", "api_key"} {
		if q.Has(k) {
			q.Set(k, "REDACTED")
		}
	}
	u.RawQuery = q.Encode()
	return u.String()
}

func (c *Client) logHTTP(label, rawURL string, status int, body []byte) {
	if c.log == nil {
		return
	}
	extra, _ := json.Marshal(struct {
		URL    string `json:"url"`
		Status int    `json:"status"`
		Body   string `json:"body"`
	}{
		URL:    redactURL(rawURL),
		Status: status,
		Body:   string(body),
	})
	c.log.Debug(label, "", string(extra))
}

func ParseFilename(filename string) (isSeries bool, imdbID, tvmazeID, season, episode string) {
	filename = strings.TrimSuffix(filename, ".mkv")

	seMatch := seRegex.FindStringSubmatch(filename)
	tvmMatch := tvmRegex.FindStringSubmatch(filename)
	imdbMatch := imdbRegex.FindStringSubmatch(filename)

	if seMatch != nil && tvmMatch != nil {
		isSeries = true
		season = fmt.Sprintf("%02d", parseInt(seMatch[1]))
		episode = fmt.Sprintf("%02d", parseInt(seMatch[2]))
		tvmazeID = tvmMatch[1]
		return
	}

	if imdbMatch != nil {
		imdbID = imdbMatch[0]
	}

	return
}

func parseInt(s string) int {
	var n int
	fmt.Sscanf(s, "%d", &n)
	return n
}

// ParseMediaType looks for a `.dvd.` / `.bluray.` / `.webdl.` / `.tvrip.`
// token (case-insensitive) in the filename and returns the matching MediaType.
// Returns an empty MediaType if no token is found, so the caller can fall back
// to the pixel-count heuristic.
func ParseMediaType(filename string) types.MediaType {
	m := mediaTypeRegex.FindStringSubmatch(strings.TrimSuffix(filename, ".mkv"))
	if m == nil {
		return ""
	}
	switch strings.ToLower(m[1]) {
	case "dvd":
		return types.MediaTypeDVD
	case "bluray":
		return types.MediaTypeBluRay
	case "webdl":
		return types.MediaTypeWebDL
	case "tvrip":
		return types.MediaTypeTVRip
	}
	return ""
}

func (c *Client) FetchMovieMetadata(ctx context.Context, imdbID string) (*types.Metadata, error) {
	reqURL := fmt.Sprintf("%s?i=%s&apikey=%s", omdbAPIURL, imdbID, c.omdbAPIKey)

	req, err := http.NewRequestWithContext(ctx, http.MethodGet, reqURL, nil)
	if err != nil {
		return nil, fmt.Errorf("building OMDb request: %w", err)
	}
	resp, err := c.httpClient.Do(req)
	if err != nil {
		return nil, fmt.Errorf("fetching OMDb: %w", err)
	}
	defer resp.Body.Close()

	body, err := io.ReadAll(io.LimitReader(resp.Body, 64*1024))
	if err != nil {
		return nil, fmt.Errorf("reading OMDb body: %w", err)
	}

	c.logHTTP("OMDb GET", reqURL, resp.StatusCode, body)

	if resp.StatusCode != http.StatusOK {
		return nil, fmt.Errorf("OMDb HTTP %d: %s", resp.StatusCode, snippet(body))
	}

	var result OMDbResponse
	if err := json.Unmarshal(body, &result); err != nil {
		return nil, fmt.Errorf("decoding OMDb response: %w (body: %s)", err, snippet(body))
	}

	if result.Response != "True" {
		if result.Error != "" {
			return nil, fmt.Errorf("OMDb error: %s", result.Error)
		}
		return nil, fmt.Errorf("OMDb non-True response: %s", snippet(body))
	}

	if result.Title == "" || result.IMDBID == "" {
		return nil, fmt.Errorf("OMDb response missing Title/imdbID (schema drift?): %s", snippet(body))
	}

	dateReleased := formatDate(result.Year)

	return &types.Metadata{
		Title:         result.Title,
		DateReleased:  dateReleased,
		IMDBID:        result.IMDBID,
		OriginalMedia: types.MediaTypeBluRay,
	}, nil
}

func (c *Client) FetchSeriesMetadata(ctx context.Context, tvmazeID, season, episode string) (*types.Metadata, error) {
	seasonNum, err := strconv.Atoi(season)
	if err != nil {
		return nil, fmt.Errorf("parsing season %q: %w", season, err)
	}
	episodeNum, err := strconv.Atoi(episode)
	if err != nil {
		return nil, fmt.Errorf("parsing episode %q: %w", episode, err)
	}

	epURL := fmt.Sprintf("%s/shows/%s/episodebynumber?season=%d&number=%d", tvmazeAPIURL, tvmazeID, seasonNum, episodeNum)
	var ep TVMazeEpisode
	if err := c.fetchTVMazeJSON(ctx, epURL, &ep); err != nil {
		return nil, fmt.Errorf("TVmaze episode %s S%dE%d: %w", tvmazeID, seasonNum, episodeNum, err)
	}

	showURL := fmt.Sprintf("%s/shows/%s", tvmazeAPIURL, tvmazeID)
	var show TVMazeShowResponse
	if err := c.fetchTVMazeJSON(ctx, showURL, &show); err != nil {
		return nil, fmt.Errorf("TVmaze show %s: %w", tvmazeID, err)
	}

	if ep.Name == "" || show.Name == "" {
		return nil, fmt.Errorf("TVmaze response missing episode/show name (schema drift?) for %s S%dE%d", tvmazeID, seasonNum, episodeNum)
	}

	return &types.Metadata{
		Title:         ep.Name,
		Collection:    show.Name,
		Season:        season,
		Episode:       episode,
		DateReleased:  formatDate(ep.Airdate),
		IMDBID:        show.Extern.IMDB,
		TVMazeID:      tvmazeID,
		IsSeries:      true,
		OriginalMedia: types.MediaTypeBluRay,
	}, nil
}

func (c *Client) fetchTVMazeJSON(ctx context.Context, reqURL string, v interface{}) error {
	req, err := http.NewRequestWithContext(ctx, http.MethodGet, reqURL, nil)
	if err != nil {
		return fmt.Errorf("build request: %w", err)
	}
	resp, err := c.httpClient.Do(req)
	if err != nil {
		return fmt.Errorf("GET: %w", err)
	}
	defer resp.Body.Close()

	body, err := io.ReadAll(io.LimitReader(resp.Body, 256*1024))
	if err != nil {
		return fmt.Errorf("read body: %w", err)
	}

	c.logHTTP("TVmaze GET", reqURL, resp.StatusCode, body)

	if resp.StatusCode != http.StatusOK {
		return fmt.Errorf("HTTP %d: %s", resp.StatusCode, snippet(body))
	}

	if err := json.Unmarshal(body, v); err != nil {
		return fmt.Errorf("decode: %w (body: %s)", err, snippet(body))
	}
	return nil
}

func snippet(body []byte) string {
	s := strings.TrimSpace(string(body))
	const max = 200
	if len(s) > max {
		s = s[:max] + "…"
	}
	return s
}

func formatDate(date string) string {
	if date == "" || date == "N/A" {
		return ""
	}
	t, err := time.Parse("02 Jan 2006", date)
	if err != nil {
		return date
	}
	return t.Format("2006-01-02")
}