Skip to content

feat(search): extract more data from tika #3198

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Draft
dschmidt wants to merge 3 commits into from
Draft

feat(search): extract more data from tika #3198

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
2edb893
refactor(search): split the tika extractor and its tests into per-fac…
dschmidt Jul 29, 2026
1514910
test(search): pin the acceptance tika image to the latest 4.0.0 snapshot
dschmidt Jul 29, 2026
35231b1
feat(search): populate audio and gps-altitude facets from tika, drop …
dschmidt Jul 29, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
205 changes: 1 addition & 204 deletions services/search/pkg/content/tika.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -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"
Expand Down Expand Up @@ -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 {
Expand All @@ -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
}
120 changes: 120 additions & 0 deletions services/search/pkg/content/tika_audio.go
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -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
}
Loading