Files
metadata-agregator/internal/provider/musicbrainz/client.go
T
2026-07-12 10:16:54 +02:00

232 lines
6.5 KiB
Go

package musicbrainz
import (
"context"
"encoding/json"
"fmt"
"io"
"net/http"
"net/url"
"strings"
"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)"
maxRetries = 3
baseBackoff = 2 * time.Second
)
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) {
var lastErr error
for attempt := 0; attempt <= maxRetries; attempt++ {
if err := c.waitForRateLimit(ctx); err != nil {
return nil, err
}
req, err := c.buildRequest(ctx, endpoint, params)
if err != nil {
return nil, err
}
data, err := c.executeAndRead(ctx, req, endpoint)
if err == nil {
return data, nil
}
if err != ErrRateLimited {
return nil, err
}
lastErr = err
if attempt < maxRetries {
backoff := baseBackoff * time.Duration(1<<attempt)
zerolog.Ctx(ctx).Info().
Int("attempt", attempt+1).
Int("max_retries", maxRetries).
Dur("backoff", backoff).
Msg("rate limited, waiting before retry")
select {
case <-ctx.Done():
return nil, ctx.Err()
case <-time.After(backoff):
}
}
}
return nil, lastErr
}
func (c *client) waitForRateLimit(ctx context.Context) error {
log := zerolog.Ctx(ctx)
log.Trace().Msg("waiting for rate limiter")
if err := c.limiter.Wait(ctx); err != nil {
log.Debug().Err(err).Msg("rate limiter interrupted")
return fmt.Errorf("rate limiter: %w", err)
}
return nil
}
func (c *client) buildRequest(ctx context.Context, endpoint string, params url.Values) (*http.Request, error) {
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")
return req, nil
}
func (c *client) executeAndRead(ctx context.Context, req *http.Request, endpoint string) ([]byte, error) {
log := zerolog.Ctx(ctx)
start := time.Now()
log.Trace().Str("url", req.URL.String()).Msg("sending HTTP request")
resp, err := c.http.Do(req)
if err != nil {
duration := time.Since(start)
log.Debug().Err(err).Str("endpoint", endpoint).Dur("duration", duration).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())
return c.readResponse(ctx, resp, endpoint, duration)
}
func (c *client) readResponse(ctx context.Context, resp *http.Response, endpoint string, duration time.Duration) ([]byte, error) {
log := zerolog.Ctx(ctx)
switch resp.StatusCode {
case http.StatusOK:
log.Trace().Str("endpoint", endpoint).Dur("duration", duration).Msg("HTTP request succeeded")
metrics.ProviderRequests.WithLabelValues("musicbrainz", endpoint, "ok").Inc()
return io.ReadAll(resp.Body)
case http.StatusBadRequest:
body, _ := io.ReadAll(resp.Body)
log.Warn().Str("endpoint", endpoint).Str("body", string(body)).Dur("duration", duration).Msg("HTTP 400 bad request")
metrics.ProviderRequests.WithLabelValues("musicbrainz", endpoint, "bad_request").Inc()
return nil, fmt.Errorf("%w: %s", ErrBadRequest, extractErrorMessage(body))
case http.StatusNotFound:
log.Debug().Str("endpoint", endpoint).Dur("duration", duration).Msg("HTTP 404")
metrics.ProviderRequests.WithLabelValues("musicbrainz", endpoint, "not_found").Inc()
return nil, ErrNotFound
case 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
default:
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))
}
}
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 joined := joinIncludes(inc); joined != "" {
params.Set("inc", joined)
}
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 joined := joinIncludes(inc); joined != "" {
params.Set("inc", joined)
}
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 joinIncludes(inc []string) string {
if len(inc) == 0 {
return ""
}
return strings.Join(inc, "+")
}
func extractErrorMessage(body []byte) string {
var errResp struct {
Error string `json:"error"`
}
if json.Unmarshal(body, &errResp) == nil && errResp.Error != "" {
return errResp.Error
}
return string(body)
}
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
}