Improve large library storage backends and scans

This commit is contained in:
ShukeBta
2026-06-15 18:16:34 +08:00
parent c9e6adf041
commit a5bf4bfdd4
36 changed files with 2340 additions and 93 deletions
+212
View File
@@ -0,0 +1,212 @@
package repository
import (
"bytes"
"context"
"encoding/json"
"fmt"
"io"
"net/http"
"net/url"
"strings"
"time"
"github.com/ShukeBta/MediaStationGo/internal/config"
"github.com/ShukeBta/MediaStationGo/internal/model"
)
type OpenSearchMediaBackend struct {
baseURL string
index string
username string
password string
client *http.Client
}
func NewOpenSearchMediaBackend(cfg config.SearchConfig) *OpenSearchMediaBackend {
if strings.TrimSpace(cfg.Backend) != "opensearch" || strings.TrimSpace(cfg.OpenSearchURL) == "" {
return nil
}
index := strings.TrimSpace(cfg.Index)
if index == "" {
index = "mediastation_media"
}
return &OpenSearchMediaBackend{
baseURL: strings.TrimRight(strings.TrimSpace(cfg.OpenSearchURL), "/"),
index: index,
username: strings.TrimSpace(cfg.Username),
password: cfg.Password,
client: &http.Client{Timeout: 4 * time.Second},
}
}
func (b *OpenSearchMediaBackend) SearchMediaIDs(ctx context.Context, query string, offset, limit int, filter MediaQueryFilter) ([]string, int64, error) {
if b == nil || b.client == nil || b.baseURL == "" || b.index == "" {
return nil, 0, fmt.Errorf("opensearch backend not configured")
}
if limit <= 0 {
limit = 50
}
must := []any{
map[string]any{
"multi_match": map[string]any{
"query": query,
"fields": []string{"title^4", "original_name^3", "genres^2", "path"},
"type": "best_fields",
"operator": "and",
"fuzziness": "AUTO",
},
},
}
filters := []any{
map[string]any{"term": map[string]any{"deleted": false}},
}
if !filter.IncludeNSFW {
filters = append(filters, map[string]any{"term": map[string]any{"nsfw": false}})
}
if len(filter.AllowedLibraryIDs) > 0 {
filters = append(filters, map[string]any{"terms": map[string]any{"library_id": filter.AllowedLibraryIDs}})
}
if len(filter.HiddenLibraryIDs) > 0 {
filters = append(filters, map[string]any{"bool": map[string]any{
"must_not": []any{map[string]any{"terms": map[string]any{"library_id": filter.HiddenLibraryIDs}}},
}})
}
body := map[string]any{
"from": offset,
"size": limit,
"_source": []string{
"id",
},
"query": map[string]any{
"bool": map[string]any{
"must": must,
"filter": filters,
},
},
}
var resp struct {
Hits struct {
Total any `json:"total"`
Hits []struct {
ID string `json:"_id"`
Source struct {
ID string `json:"id"`
} `json:"_source"`
} `json:"hits"`
} `json:"hits"`
}
if err := b.doJSON(ctx, http.MethodPost, "/"+url.PathEscape(b.index)+"/_search", body, &resp); err != nil {
return nil, 0, err
}
ids := make([]string, 0, len(resp.Hits.Hits))
for _, hit := range resp.Hits.Hits {
id := strings.TrimSpace(hit.Source.ID)
if id == "" {
id = strings.TrimSpace(hit.ID)
}
if id != "" {
ids = append(ids, id)
}
}
return ids, openSearchTotal(resp.Hits.Total), nil
}
func (b *OpenSearchMediaBackend) EnsureIndex(ctx context.Context) error {
if err := b.do(ctx, http.MethodHead, "/"+url.PathEscape(b.index), nil, "", nil); err == nil {
return nil
}
mapping := map[string]any{
"mappings": map[string]any{
"properties": map[string]any{
"id": map[string]any{"type": "keyword"},
"library_id": map[string]any{"type": "keyword"},
"title": map[string]any{"type": "text"},
"original_name": map[string]any{"type": "text"},
"path": map[string]any{"type": "text"},
"genres": map[string]any{"type": "text"},
"nsfw": map[string]any{"type": "boolean"},
"deleted": map[string]any{"type": "boolean"},
"created_at": map[string]any{"type": "date"},
},
},
}
return b.doJSON(ctx, http.MethodPut, "/"+url.PathEscape(b.index), mapping, nil)
}
func (b *OpenSearchMediaBackend) IndexMedia(ctx context.Context, rows []model.Media) error {
if len(rows) == 0 {
return nil
}
var bulk bytes.Buffer
enc := json.NewEncoder(&bulk)
for _, row := range rows {
if err := enc.Encode(map[string]any{"index": map[string]any{"_index": b.index, "_id": row.ID}}); err != nil {
return err
}
if err := enc.Encode(map[string]any{
"id": row.ID,
"library_id": row.LibraryID,
"title": row.Title,
"original_name": row.OriginalName,
"path": row.Path,
"genres": row.Genres,
"nsfw": row.NSFW,
"deleted": row.DeletedAt.Valid,
"created_at": row.CreatedAt,
}); err != nil {
return err
}
}
return b.do(ctx, http.MethodPost, "/_bulk", &bulk, "application/x-ndjson", nil)
}
func (b *OpenSearchMediaBackend) doJSON(ctx context.Context, method, path string, body any, out any) error {
var reader io.Reader
if body != nil {
raw, err := json.Marshal(body)
if err != nil {
return err
}
reader = bytes.NewReader(raw)
}
return b.do(ctx, method, path, reader, "application/json", out)
}
func (b *OpenSearchMediaBackend) do(ctx context.Context, method, path string, body io.Reader, contentType string, out any) error {
req, err := http.NewRequestWithContext(ctx, method, b.baseURL+path, body)
if err != nil {
return err
}
if contentType != "" {
req.Header.Set("Content-Type", contentType)
}
if b.username != "" {
req.SetBasicAuth(b.username, b.password)
}
resp, err := b.client.Do(req)
if err != nil {
return err
}
defer resp.Body.Close()
if resp.StatusCode >= 400 {
return fmt.Errorf("opensearch %s %s returned %d", method, path, resp.StatusCode)
}
if out == nil {
_, _ = io.Copy(io.Discard, resp.Body)
return nil
}
return json.NewDecoder(resp.Body).Decode(out)
}
func openSearchTotal(value any) int64 {
switch v := value.(type) {
case float64:
return int64(v)
case map[string]any:
if n, ok := v["value"].(float64); ok {
return int64(n)
}
}
return 0
}
+57
View File
@@ -0,0 +1,57 @@
package repository
import (
"encoding/json"
"net/http"
"net/http/httptest"
"testing"
"github.com/ShukeBta/MediaStationGo/internal/config"
)
func TestOpenSearchMediaBackendSearchesIDs(t *testing.T) {
var gotPath string
upstream := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
gotPath = r.URL.Path
if r.Method != http.MethodPost {
t.Fatalf("method = %s, want POST", r.Method)
}
var body map[string]any
if err := json.NewDecoder(r.Body).Decode(&body); err != nil {
t.Fatal(err)
}
if body["from"].(float64) != 5 || body["size"].(float64) != 10 {
t.Fatalf("paging body = %#v", body)
}
_ = json.NewEncoder(w).Encode(map[string]any{
"hits": map[string]any{
"total": map[string]any{"value": 2},
"hits": []any{
map[string]any{"_id": "m-1", "_source": map[string]any{"id": "m-1"}},
map[string]any{"_id": "m-2", "_source": map[string]any{"id": "m-2"}},
},
},
})
}))
defer upstream.Close()
backend := NewOpenSearchMediaBackend(config.SearchConfig{
Backend: "opensearch",
OpenSearchURL: upstream.URL,
Index: "media-test",
})
ids, total, err := backend.SearchMediaIDs(t.Context(), "流浪地球", 5, 10, MediaQueryFilter{
IncludeNSFW: false,
AllowedLibraryIDs: []string{"lib-1"},
HiddenLibraryIDs: []string{"adult"},
})
if err != nil {
t.Fatal(err)
}
if gotPath != "/media-test/_search" {
t.Fatalf("path = %q", gotPath)
}
if total != 2 || len(ids) != 2 || ids[0] != "m-1" || ids[1] != "m-2" {
t.Fatalf("ids=%#v total=%d", ids, total)
}
}
+102
View File
@@ -274,6 +274,23 @@ type MediaRepository struct {
searchIndexOnce sync.Once
searchIndexAvailable bool
searchBackend MediaSearchBackend
}
type MediaSearchBackend interface {
SearchMediaIDs(ctx context.Context, query string, offset, limit int, filter MediaQueryFilter) ([]string, int64, error)
}
type MediaSearchSyncBackend interface {
MediaSearchBackend
EnsureIndex(ctx context.Context) error
IndexMedia(ctx context.Context, rows []model.Media) error
}
func (r *MediaRepository) SetSearchBackend(backend MediaSearchBackend) {
if r != nil {
r.searchBackend = backend
}
}
// MediaQueryFilter is applied to user-facing media queries so NSFW items and
@@ -323,6 +340,7 @@ func (r *MediaRepository) upsert(ctx context.Context, m *model.Media) error {
m.ScrapeStatus = "pending"
}
if createErr := r.db.WithContext(ctx).Create(m).Error; createErr == nil {
r.indexMediaBestEffort(ctx, *m)
return nil
} else if retryErr := r.db.WithContext(ctx).Unscoped().Where("path = ?", m.Path).First(&existing).Error; retryErr != nil {
return createErr
@@ -433,9 +451,20 @@ func (r *MediaRepository) upsert(ctx context.Context, m *model.Media) error {
}
// 回写 ID / 不可变字段,让 caller 拿到完整的现有行。
*m = existing
if fresh, err := r.FindByID(ctx, existing.ID); err == nil && fresh != nil {
r.indexMediaBestEffort(ctx, *fresh)
}
return nil
}
func (r *MediaRepository) indexMediaBestEffort(ctx context.Context, media model.Media) {
backend, ok := r.searchBackend.(MediaSearchSyncBackend)
if !ok {
return
}
_ = backend.IndexMedia(ctx, []model.Media{media})
}
func setIfChanged[T comparable](updates map[string]any, key string, current, next T) {
if current != next {
updates[key] = next
@@ -500,6 +529,11 @@ func (r *MediaRepository) SearchFilteredPage(ctx context.Context, query string,
if limit <= 0 {
limit = 50
}
if query != "" && r.searchBackend != nil {
if items, total, ok := r.searchFilteredBackend(ctx, query, offset, limit, filter); ok {
return items, total, nil
}
}
if query != "" {
if items, total, ok := r.searchFilteredFTS(ctx, query, offset, limit, filter); ok {
if total > 0 {
@@ -510,6 +544,36 @@ func (r *MediaRepository) SearchFilteredPage(ctx context.Context, query string,
return r.searchFilteredLIKE(ctx, query, offset, limit, filter)
}
func (r *MediaRepository) searchFilteredBackend(ctx context.Context, query string, offset, limit int, filter MediaQueryFilter) ([]model.Media, int64, bool) {
ids, total, err := r.searchBackend.SearchMediaIDs(ctx, query, offset, limit, filter)
if err != nil {
return nil, 0, false
}
if len(ids) == 0 {
return []model.Media{}, total, true
}
var rows []model.Media
q := r.db.WithContext(ctx).Model(&model.Media{}).Where("id IN ?", ids)
q = applyMediaQueryFilter(q, filter)
if err := q.Find(&rows).Error; err != nil {
return nil, 0, false
}
byID := make(map[string]model.Media, len(rows))
for _, row := range rows {
byID[row.ID] = row
}
items := make([]model.Media, 0, len(ids))
for _, id := range ids {
if row, ok := byID[id]; ok {
items = append(items, row)
}
}
if len(items) == 0 && total > 0 {
return nil, 0, false
}
return items, total, true
}
func (r *MediaRepository) searchFilteredFTS(ctx context.Context, query string, offset, limit int, filter MediaQueryFilter) ([]model.Media, int64, bool) {
if !r.searchIndexEnabled(ctx) {
return nil, 0, false
@@ -630,6 +694,9 @@ func escapeLike(value string) string {
}
func (r *MediaRepository) BackfillSearchIndex(ctx context.Context, batchLimit int) (int64, error) {
if backend, ok := r.searchBackend.(MediaSearchSyncBackend); ok {
return r.backfillExternalSearchIndex(ctx, backend, batchLimit)
}
if batchLimit <= 0 {
batchLimit = 1000
}
@@ -654,10 +721,45 @@ LIMIT ?
return res.RowsAffected, res.Error
}
func (r *MediaRepository) backfillExternalSearchIndex(ctx context.Context, backend MediaSearchSyncBackend, batchLimit int) (int64, error) {
if batchLimit <= 0 {
batchLimit = 1000
}
if err := backend.EnsureIndex(ctx); err != nil {
return 0, err
}
var lastID string
for {
var rows []model.Media
q := r.db.WithContext(ctx).
Model(&model.Media{}).
Where("deleted_at IS NULL")
if lastID != "" {
q = q.Where("id > ?", lastID)
}
if err := q.Order("id ASC").Limit(batchLimit).Find(&rows).Error; err != nil {
return 0, err
}
if len(rows) == 0 {
return 0, nil
}
if err := backend.IndexMedia(ctx, rows); err != nil {
return 0, err
}
lastID = rows[len(rows)-1].ID
if len(rows) < batchLimit {
return 0, nil
}
}
}
func (r *MediaRepository) searchIndexEnabled(ctx context.Context) bool {
if r == nil || r.db == nil {
return false
}
if r.db.Dialector == nil || r.db.Dialector.Name() != "sqlite" {
return false
}
r.searchIndexOnce.Do(func() {
var count int64
err := r.db.WithContext(ctx).
+54
View File
@@ -1,6 +1,8 @@
package repository
import (
"context"
"errors"
"testing"
"time"
@@ -70,6 +72,58 @@ func TestMediaUpsertSkipsUnchangedExistingRow(t *testing.T) {
}
}
type fakeMediaSearchBackend struct {
ids []string
err error
}
func (f fakeMediaSearchBackend) SearchMediaIDs(context.Context, string, int, int, MediaQueryFilter) ([]string, int64, error) {
if f.err != nil {
return nil, 0, f.err
}
return append([]string(nil), f.ids...), int64(len(f.ids)), nil
}
func TestMediaSearchUsesExternalBackendAndFallsBack(t *testing.T) {
db, err := gorm.Open(sqlite.Open(":memory:"), &gorm.Config{})
if err != nil {
t.Fatal(err)
}
if err := database.AutoMigrate(db); err != nil {
t.Fatalf("migrate: %v", err)
}
repos := New(db)
lib := model.Library{Name: "Movies", Path: "/media/movie", Type: "movie", Enabled: true}
if err := repos.Library.Create(t.Context(), &lib); err != nil {
t.Fatal(err)
}
for _, row := range []model.Media{
{Base: model.Base{ID: "m-1"}, LibraryID: lib.ID, Title: "Alpha", Path: "/media/a.mkv"},
{Base: model.Base{ID: "m-2"}, LibraryID: lib.ID, Title: "Beta", Path: "/media/b.mkv"},
} {
if err := repos.DB.Create(&row).Error; err != nil {
t.Fatal(err)
}
}
repos.Media.SetSearchBackend(fakeMediaSearchBackend{ids: []string{"m-2", "m-1"}})
items, total, err := repos.Media.SearchFilteredPage(t.Context(), "anything", 0, 10, MediaQueryFilter{IncludeNSFW: true})
if err != nil {
t.Fatal(err)
}
if total != 2 || len(items) != 2 || items[0].ID != "m-2" || items[1].ID != "m-1" {
t.Fatalf("external search result total=%d items=%#v", total, items)
}
repos.Media.SetSearchBackend(fakeMediaSearchBackend{err: errors.New("opensearch down")})
items, total, err = repos.Media.SearchFilteredPage(t.Context(), "Alpha", 0, 10, MediaQueryFilter{IncludeNSFW: true})
if err != nil {
t.Fatal(err)
}
if total != 1 || len(items) != 1 || items[0].ID != "m-1" {
t.Fatalf("fallback result total=%d items=%#v", total, items)
}
}
func TestMediaSearchFilteredSupportsChineseFuzzyTerms(t *testing.T) {
db, err := gorm.Open(sqlite.Open(":memory:"), &gorm.Config{})
if err != nil {