feat(search): extract more data from tika

go.mod

@@ -22,7 +22,6 @@ require (
 	github.com/ggwhite/go-masker v1.1.0
 	github.com/go-chi/chi/v5 v5.3.0
 	github.com/go-chi/render v1.0.3
-	github.com/go-jose/go-jose/v3 v3.0.5
 	github.com/go-ldap/ldap/v3 v3.4.13
 	github.com/go-ldap/ldif v0.0.0-20200320164324-fd88d9b715b3
 	github.com/go-micro/plugins/v4/client/grpc v1.2.1
@@ -63,7 +62,7 @@ require (
 	github.com/onsi/gomega v1.42.1
 	github.com/open-policy-agent/opa v1.17.1
 	github.com/opencloud-eu/icap-client v0.0.0-20250930132611-28a2afe62d89
-	github.com/opencloud-eu/libre-graph-api-go v1.0.8-0.20260310090739-853d972b282d
+	github.com/opencloud-eu/libre-graph-api-go v1.0.8-0.20260711144435-70a1ac98fd3c
 	github.com/opencloud-eu/reva/v2 v2.46.4-0.20260625152426-8cff2a7032ec
 	github.com/opensearch-project/opensearch-go/v4 v4.6.0
 	github.com/orcaman/concurrent-map v1.0.0
@@ -208,6 +207,7 @@ require (
 	github.com/go-git/go-billy/v5 v5.9.0 // indirect
 	github.com/go-git/go-git/v5 v5.19.1 // indirect
 	github.com/go-ini/ini v1.67.0 // indirect
+	github.com/go-jose/go-jose/v3 v3.0.5 // indirect
 	github.com/go-jose/go-jose/v4 v4.1.4 // indirect
 	github.com/go-kit/log v0.2.1 // indirect
 	github.com/go-logfmt/logfmt v0.5.1 // indirect

go.sum

@@ -947,8 +947,8 @@ github.com/opencloud-eu/go-micro-plugins/v4/store/nats-js-kv v0.0.0-202505121527
 github.com/opencloud-eu/go-micro-plugins/v4/store/nats-js-kv v0.0.0-20250512152754-23325793059a/go.mod h1:pjcozWijkNPbEtX5SIQaxEW/h8VAVZYTLx+70bmB3LY=
 github.com/opencloud-eu/icap-client v0.0.0-20250930132611-28a2afe62d89 h1:W1ms+lP5lUUIzjRGDg93WrQfZJZCaV1ZP3KeyXi8bzY=
 github.com/opencloud-eu/icap-client v0.0.0-20250930132611-28a2afe62d89/go.mod h1:vigJkNss1N2QEceCuNw/ullDehncuJNFB6mEnzfq9UI=
-github.com/opencloud-eu/libre-graph-api-go v1.0.8-0.20260310090739-853d972b282d h1:JcqGDiyrcaQwVyV861TUyQgO7uEmsjkhfm7aQd84dOw=
-github.com/opencloud-eu/libre-graph-api-go v1.0.8-0.20260310090739-853d972b282d/go.mod h1:pzatilMEHZFT3qV7C/X3MqOa3NlRQuYhlRhZTL+hN6Q=
+github.com/opencloud-eu/libre-graph-api-go v1.0.8-0.20260711144435-70a1ac98fd3c h1:uIDNX0ipAwHT6dVgGxjbAOQ3GC9X77Tq3y8mDXJEXRU=
+github.com/opencloud-eu/libre-graph-api-go v1.0.8-0.20260711144435-70a1ac98fd3c/go.mod h1:lTM8JeGblNpoMySTW7Lui2+c5TTLI95mwxtdUIHHrhU=
 github.com/opencloud-eu/reva/v2 v2.46.4-0.20260625152426-8cff2a7032ec h1:kL9P87cVpe8GKvo5vzLpJstpiVXFvkZn2+Dco53nfuM=
 github.com/opencloud-eu/reva/v2 v2.46.4-0.20260625152426-8cff2a7032ec/go.mod h1:l4CG5u8tdWZj6IyavXO9xDdnNTV85oaFudGwEDPvsnk=
 github.com/opencloud-eu/secure v0.0.0-20260312082735-b6f5cb2244e4 h1:l2oB/RctH+t8r7QBj5p8thfEHCM/jF35aAY3WQ3hADI=

services/collaboration/pkg/connector/fileconnector.go

@@ -1578,8 +1578,6 @@ func (f *FileConnector) GetAvatar(ctx context.Context, userID string) (*Connecto
 		}
 		return nil, NewConnectorError(http.StatusBadGateway, "failed to fetch avatar")
 	}
-	defer photoFile.Close()
-
 	data, err := io.ReadAll(photoFile)
 	if err != nil {
 		logger.Error().Err(err).Msg("GetAvatar: failed to read avatar body")

services/search/pkg/content/tika.go

@@ -3,16 +3,12 @@ package content
 import (
 	"context"
 	"fmt"
-	"math"
-	"strconv"
 	"strings"
-	"time"
 
 	gateway "github.com/cs3org/go-cs3apis/cs3/gateway/v1beta1"
 	provider "github.com/cs3org/go-cs3apis/cs3/storage/provider/v1beta1"
 	"github.com/google/go-tika/tika"
 	"github.com/opencloud-eu/reva/v2/pkg/rgrpc/todo/pool"
-	libregraph "github.com/opencloud-eu/libre-graph-api-go"
 
 	"github.com/opencloud-eu/opencloud/pkg/log"
 	"github.com/opencloud-eu/opencloud/services/search/pkg/config"
@@ -94,10 +90,7 @@ func (t Tika) Extract(ctx context.Context, ri *provider.ResourceInfo) (Document,
 		doc.Location = t.getLocation(meta)
 		doc.Image = t.getImage(meta)
 		doc.Photo = t.getPhoto(meta)
-
-		if contentType, err := getFirstValue(meta, "Content-Type"); err == nil && strings.HasPrefix(contentType, "audio/") {
-			doc.Audio = t.getAudio(meta)
-		}
+		doc.Audio = t.getAudio(meta)
 	}
 
 	if langCode, _ := t.tika.LanguageString(ctx, doc.Content); langCode != "" && t.CleanStopWords {
@@ -106,199 +99,3 @@ func (t Tika) Extract(ctx context.Context, ri *provider.ResourceInfo) (Document,
 
 	return doc, nil
 }
-
-func (t Tika) getImage(meta map[string][]string) *libregraph.Image {
-	var image *libregraph.Image
-	initImage := func() {
-		if image == nil {
-			image = libregraph.NewImage()
-		}
-	}
-
-	if v, err := getFirstValue(meta, "tiff:ImageWidth"); err == nil {
-		if i, err := strconv.ParseInt(v, 0, 32); err == nil {
-			initImage()
-			image.SetWidth(int32(i))
-		}
-	}
-
-	if v, err := getFirstValue(meta, "tiff:ImageLength"); err == nil {
-		if i, err := strconv.ParseInt(v, 0, 32); err == nil {
-			initImage()
-			image.SetHeight(int32(i))
-		}
-	}
-
-	return image
-}
-
-func (t Tika) getLocation(meta map[string][]string) *libregraph.GeoCoordinates {
-	var location *libregraph.GeoCoordinates
-	initLocation := func() {
-		if location == nil {
-			location = libregraph.NewGeoCoordinates()
-		}
-	}
-
-	// TODO: location.Altitute: transform the following data to … feet above sea level.
-	// "GPS:GPS Altitude":                          []string{"227.4 metres"},
-	// "GPS:GPS Altitude Ref":                      []string{"Sea level"},
-
-	if v, err := getFirstValue(meta, "geo:lat"); err == nil {
-		if i, err := strconv.ParseFloat(v, 64); err == nil {
-			initLocation()
-			location.SetLatitude(i)
-		}
-	}
-
-	if v, err := getFirstValue(meta, "geo:long"); err == nil {
-		if i, err := strconv.ParseFloat(v, 64); err == nil {
-			initLocation()
-			location.SetLongitude(i)
-		}
-	}
-
-	return location
-}
-
-func (t Tika) getPhoto(meta map[string][]string) *libregraph.Photo {
-	var photo *libregraph.Photo
-	initPhoto := func() {
-		if photo == nil {
-			photo = libregraph.NewPhoto()
-		}
-	}
-
-	if v, err := getFirstValue(meta, "tiff:Make"); err == nil {
-		initPhoto()
-		photo.SetCameraMake(v)
-	}
-
-	if v, err := getFirstValue(meta, "tiff:Model"); err == nil {
-		initPhoto()
-		photo.SetCameraModel(v)
-	}
-
-	if v, err := getFirstValue(meta, "exif:FNumber"); err == nil {
-		if i, err := strconv.ParseFloat(v, 64); err == nil {
-			initPhoto()
-			photo.SetFNumber(i)
-		}
-	}
-
-	if v, err := getFirstValue(meta, "exif:FocalLength"); err == nil {
-		if i, err := strconv.ParseFloat(v, 64); err == nil {
-			initPhoto()
-			photo.SetFocalLength(i)
-		}
-	}
-
-	if v, err := getFirstValue(meta, "Base ISO"); err == nil {
-		if i, err := strconv.ParseInt(v, 0, 32); err == nil {
-			initPhoto()
-			photo.SetIso(int32(i))
-		}
-	}
-
-	if v, err := getFirstValue(meta, "tiff:Orientation"); err == nil {
-		if i, err := strconv.ParseInt(v, 0, 32); err == nil {
-			initPhoto()
-			photo.SetOrientation(int32(i))
-		}
-	}
-
-	if v, err := getFirstValue(meta, "exif:DateTimeOriginal"); err == nil {
-		layout := "2006-01-02T15:04:05"
-		if t, err := time.Parse(layout, v); err == nil {
-			initPhoto()
-			photo.SetTakenDateTime(t)
-		}
-	}
-
-	if v, err := getFirstValue(meta, "exif:ExposureTime"); err == nil {
-		if i, err := strconv.ParseFloat(v, 64); err == nil {
-			initPhoto()
-			photo.SetExposureNumerator(1)
-			photo.SetExposureDenominator(math.Round(1 / i))
-		}
-	}
-
-	return photo
-}
-
-func (t Tika) getAudio(meta map[string][]string) *libregraph.Audio {
-	var audio *libregraph.Audio
-	initAudio := func() {
-		if audio == nil {
-			audio = libregraph.NewAudio()
-		}
-	}
-
-	if v, err := getFirstValue(meta, "xmpDM:album"); err == nil {
-		initAudio()
-		audio.SetAlbum(v)
-	}
-
-	if v, err := getFirstValue(meta, "xmpDM:albumArtist"); err == nil {
-		initAudio()
-		audio.SetAlbumArtist(v)
-	}
-
-	if v, err := getFirstValue(meta, "xmpDM:artist"); err == nil {
-		initAudio()
-		audio.SetArtist(v)
-	}
-
-	// TODO: audio.Bitrate: not provided by tika
-	// TODO: audio.Composers: not provided by tika
-	// TODO: audio.Copyright: not provided by tika for audio files?
-
-	if v, err := getFirstValue(meta, "xmpDM:discNumber"); err == nil {
-		if i, err := strconv.ParseInt(v, 10, 32); err == nil {
-			initAudio()
-			audio.SetDisc(int32(i))
-		}
-
-	}
-
-	//  TODO: audio.DiscCount: not provided by tika
-
-	if v, err := getFirstValue(meta, "xmpDM:duration"); err == nil {
-		// Tika emits fractional seconds.
-		if f, err := strconv.ParseFloat(v, 64); err == nil {
-			initAudio()
-			audio.SetDuration(int64(math.Round(f * 1000)))
-		}
-	}
-
-	if v, err := getFirstValue(meta, "xmpDM:genre"); err == nil {
-		initAudio()
-		audio.SetGenre(v)
-	}
-
-	// TODO: audio.HasDrm: not provided by tika
-	// TODO: audio.IsVariableBitrate: not provided by tika
-
-	if v, err := getFirstValue(meta, "dc:title"); err == nil {
-		initAudio()
-		audio.SetTitle(v)
-	}
-
-	if v, err := getFirstValue(meta, "xmpDM:trackNumber"); err == nil {
-		if i, err := strconv.ParseInt(v, 10, 32); err == nil {
-			initAudio()
-			audio.SetTrack(int32(i))
-		}
-	}
-
-	// TODO: audio.TrackCount: not provided by tika
-
-	if v, err := getFirstValue(meta, "xmpDM:releaseDate"); err == nil {
-		if i, err := strconv.ParseInt(v, 10, 32); err == nil {
-			initAudio()
-			audio.SetYear(int32(i))
-		}
-	}
-
-	return audio
-}

services/search/pkg/content/tika_audio.go

@@ -0,0 +1,120 @@
+package content
+
+import (
+	"math"
+	"strconv"
+
+	libregraph "github.com/opencloud-eu/libre-graph-api-go"
+)
+
+func (t Tika) getAudio(meta map[string][]string) *libregraph.Audio {
+	var audio *libregraph.Audio
+	initAudio := func() {
+		if audio == nil {
+			audio = libregraph.NewAudio()
+		}
+	}
+
+	if v, err := getFirstValue(meta, "xmpDM:album"); err == nil {
+		initAudio()
+		audio.SetAlbum(v)
+	}
+
+	if v, err := getFirstValue(meta, "xmpDM:albumArtist"); err == nil {
+		initAudio()
+		audio.SetAlbumArtist(v)
+	}
+
+	if v, err := getFirstValue(meta, "xmpDM:artist"); err == nil {
+		initAudio()
+		audio.SetArtist(v)
+	}
+
+	if v, err := getFirstValue(meta, "audio:bitrate"); err == nil {
+		// tika emits bits per second, graph wants kbps
+		if bps, err := strconv.ParseInt(v, 10, 64); err == nil {
+			initAudio()
+			audio.SetBitrate(bps / 1000)
+		}
+	}
+
+	if v, err := getFirstValue(meta, "xmpDM:composer"); err == nil {
+		initAudio()
+		audio.SetComposers(v)
+	}
+
+	if v, err := getFirstValue(meta, "xmpDM:copyright"); err == nil {
+		initAudio()
+		audio.SetCopyright(v)
+	}
+
+	if v, err := getFirstValue(meta, "xmpDM:discNumber"); err == nil {
+		if i, err := strconv.ParseInt(v, 10, 32); err == nil {
+			initAudio()
+			audio.SetDisc(int32(i))
+		}
+
+	}
+
+	if v, err := getFirstValue(meta, "audio:disc-count"); err == nil {
+		if i, err := strconv.ParseInt(v, 10, 32); err == nil {
+			initAudio()
+			audio.SetDiscCount(int32(i))
+		}
+	}
+
+	if v, err := getFirstValue(meta, "xmpDM:duration"); err == nil {
+		// Tika emits fractional seconds.
+		if f, err := strconv.ParseFloat(v, 64); err == nil {
+			initAudio()
+			audio.SetDuration(int64(math.Round(f * 1000)))
+		}
+	}
+
+	if v, err := getFirstValue(meta, "xmpDM:genre"); err == nil {
+		initAudio()
+		audio.SetGenre(v)
+	}
+
+	if v, err := getFirstValue(meta, "audio:has-drm"); err == nil {
+		if b, err := strconv.ParseBool(v); err == nil {
+			initAudio()
+			audio.SetHasDrm(b)
+		}
+	}
+
+	if v, err := getFirstValue(meta, "audio:is-variable-bitrate"); err == nil {
+		if b, err := strconv.ParseBool(v); err == nil {
+			initAudio()
+			audio.SetIsVariableBitrate(b)
+		}
+	}
+
+	if v, err := getFirstValue(meta, "dc:title"); err == nil {
+		initAudio()
+		audio.SetTitle(v)
+	}
+
+	if v, err := getFirstValue(meta, "xmpDM:trackNumber"); err == nil {
+		if i, err := strconv.ParseInt(v, 10, 32); err == nil {
+			initAudio()
+			audio.SetTrack(int32(i))
+		}
+	}
+
+	if v, err := getFirstValue(meta, "audio:track-count"); err == nil {
+		if i, err := strconv.ParseInt(v, 10, 32); err == nil {
+			initAudio()
+			audio.SetTrackCount(int32(i))
+		}
+	}
+
+	if v, err := getFirstValue(meta, "xmpDM:releaseDate"); err == nil {
+		if i, err := strconv.ParseInt(v, 10, 32); err == nil {
+			initAudio()
+			audio.SetYear(int32(i))
+		}
+	}
+
+	return audio
+}