metadata-agregator/internal/provider/musicbrainz/client.go

package musicbrainz

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

	"github.com/rs/zerolog"
	"golang.org/x/time/rate"

	"github.com/metadata-agregator/internal/metrics"
)

const (
	baseURL   = "https://musicbrainz.org/ws/2"
	userAgent = "MetadataAggregator/0.1.0 (https://github.com/metadata-agregator)"
)

type client struct {
	http    *http.Client
	limiter *rate.Limiter
}

func newClient() *client {
	return &client{
		http: &http.Client{
			Timeout: 30 * time.Second,
		},
		limiter: rate.NewLimiter(rate.Every(time.Second), 1),
	}
}

func (c *client) get(ctx context.Context, endpoint string, params url.Values) ([]byte, error) {
	log := zerolog.Ctx(ctx)

	log.Trace().Str("endpoint", endpoint).Msg("waiting for rate limiter")
	if err := c.limiter.Wait(ctx); err != nil {
		log.Debug().Err(err).Msg("rate limiter interrupted")
		return nil, fmt.Errorf("rate limiter: %w", err)
	}

	if params == nil {
		params = url.Values{}
	}
	params.Set("fmt", "json")

	reqURL := fmt.Sprintf("%s/%s?%s", baseURL, endpoint, params.Encode())

	req, err := http.NewRequestWithContext(ctx, http.MethodGet, reqURL, nil)
	if err != nil {
		return nil, fmt.Errorf("create request: %w", err)
	}

	req.Header.Set("User-Agent", userAgent)
	req.Header.Set("Accept", "application/json")

	start := time.Now()
	log.Trace().Str("url", reqURL).Msg("sending HTTP request")

	resp, err := c.http.Do(req)
	if err != nil {
		log.Debug().Err(err).Str("endpoint", endpoint).Dur("duration", time.Since(start)).Msg("HTTP request failed")
		metrics.ProviderRequests.WithLabelValues("musicbrainz", endpoint, "error").Inc()
		return nil, fmt.Errorf("do request: %w", err)
	}
	defer resp.Body.Close()

	duration := time.Since(start)
	metrics.ProviderLatency.WithLabelValues("musicbrainz", endpoint).Observe(duration.Seconds())

	if resp.StatusCode == http.StatusNotFound {
		log.Debug().Str("endpoint", endpoint).Int("status", resp.StatusCode).Dur("duration", duration).Msg("HTTP 404")
		metrics.ProviderRequests.WithLabelValues("musicbrainz", endpoint, "not_found").Inc()
		return nil, ErrNotFound
	}

	if resp.StatusCode == http.StatusServiceUnavailable {
		log.Warn().Str("endpoint", endpoint).Dur("duration", duration).Msg("HTTP 503 rate limited by provider")
		metrics.ProviderRequests.WithLabelValues("musicbrainz", endpoint, "rate_limited").Inc()
		return nil, ErrRateLimited
	}

	if resp.StatusCode != http.StatusOK {
		body, _ := io.ReadAll(resp.Body)
		log.Warn().Str("endpoint", endpoint).Int("status", resp.StatusCode).Str("body", string(body)).Dur("duration", duration).Msg("unexpected HTTP status")
		metrics.ProviderRequests.WithLabelValues("musicbrainz", endpoint, fmt.Sprintf("%d", resp.StatusCode)).Inc()
		return nil, fmt.Errorf("unexpected status %d: %s", resp.StatusCode, string(body))
	}

	log.Trace().Str("endpoint", endpoint).Int("status", resp.StatusCode).Dur("duration", duration).Msg("HTTP request succeeded")
	metrics.ProviderRequests.WithLabelValues("musicbrainz", endpoint, "ok").Inc()

	return io.ReadAll(resp.Body)
}

func (c *client) lookup(ctx context.Context, entity, id string, inc []string) ([]byte, error) {
	zerolog.Ctx(ctx).Debug().Str("entity", entity).Str("id", id).Strs("includes", inc).Msg("provider lookup")

	params := url.Values{}
	if len(inc) > 0 {
		incStr := ""
		for i, v := range inc {
			if i > 0 {
				incStr += "+"
			}
			incStr += v
		}
		params.Set("inc", incStr)
	}

	return c.get(ctx, fmt.Sprintf("%s/%s", entity, id), params)
}

func (c *client) browse(ctx context.Context, entity, linkedEntity, linkedID string, limit, offset int, inc []string) ([]byte, error) {
	zerolog.Ctx(ctx).Debug().
		Str("entity", entity).
		Str("linked_entity", linkedEntity).
		Str("linked_id", linkedID).
		Int("limit", limit).
		Int("offset", offset).
		Msg("provider browse")

	params := url.Values{}
	params.Set(linkedEntity, linkedID)
	params.Set("limit", fmt.Sprintf("%d", limit))
	params.Set("offset", fmt.Sprintf("%d", offset))

	if len(inc) > 0 {
		incStr := ""
		for i, v := range inc {
			if i > 0 {
				incStr += "+"
			}
			incStr += v
		}
		params.Set("inc", incStr)
	}

	return c.get(ctx, entity, params)
}

func (c *client) search(ctx context.Context, entity, query string, limit, offset int) ([]byte, error) {
	zerolog.Ctx(ctx).Debug().Str("entity", entity).Str("query", query).Int("limit", limit).Int("offset", offset).Msg("provider search")

	params := url.Values{}
	params.Set("query", query)
	params.Set("limit", fmt.Sprintf("%d", limit))
	params.Set("offset", fmt.Sprintf("%d", offset))

	return c.get(ctx, entity, params)
}

func decode[T any](data []byte) (*T, error) {
	var result T
	if err := json.Unmarshal(data, &result); err != nil {
		return nil, fmt.Errorf("decode: %w", err)
	}
	return &result, nil
}