[feature] Media attachment placeholders (#2331)

* [feature] Use placeholders for unknown media types

* fix read of underreported small files

* switch to reduce nesting

* simplify cleanup
This commit is contained in:
tobi 2023-11-10 19:29:26 +01:00 committed by GitHub
commit ba9d6b467a
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
41 changed files with 1472 additions and 841 deletions

View file

@ -622,21 +622,16 @@ func (d *Dereferencer) fetchRemoteAccountAvatar(ctx context.Context, tsport tran
processing, ok := d.derefAvatars[latestAcc.AvatarRemoteURL]
if !ok {
var err error
// Set the media data function to dereference avatar from URI.
data := func(ctx context.Context) (io.ReadCloser, int64, error) {
return tsport.DereferenceMedia(ctx, avatarURI)
}
// Create new media processing request from the media manager instance.
processing, err = d.mediaManager.PreProcessMedia(ctx, data, latestAcc.ID, &media.AdditionalMediaInfo{
processing = d.mediaManager.PreProcessMedia(data, latestAcc.ID, &media.AdditionalMediaInfo{
Avatar: func() *bool { v := true; return &v }(),
RemoteURL: &latestAcc.AvatarRemoteURL,
})
if err != nil {
return gtserror.Newf("error preprocessing media for attachment %s: %w", latestAcc.AvatarRemoteURL, err)
}
// Store media in map to mark as processing.
d.derefAvatars[latestAcc.AvatarRemoteURL] = processing
@ -713,21 +708,16 @@ func (d *Dereferencer) fetchRemoteAccountHeader(ctx context.Context, tsport tran
processing, ok := d.derefHeaders[latestAcc.HeaderRemoteURL]
if !ok {
var err error
// Set the media data function to dereference avatar from URI.
data := func(ctx context.Context) (io.ReadCloser, int64, error) {
return tsport.DereferenceMedia(ctx, headerURI)
}
// Create new media processing request from the media manager instance.
processing, err = d.mediaManager.PreProcessMedia(ctx, data, latestAcc.ID, &media.AdditionalMediaInfo{
processing = d.mediaManager.PreProcessMedia(data, latestAcc.ID, &media.AdditionalMediaInfo{
Header: func() *bool { v := true; return &v }(),
RemoteURL: &latestAcc.HeaderRemoteURL,
})
if err != nil {
return gtserror.Newf("error preprocessing media for attachment %s: %w", latestAcc.HeaderRemoteURL, err)
}
// Store media in map to mark as processing.
d.derefHeaders[latestAcc.HeaderRemoteURL] = processing

View file

@ -1,54 +0,0 @@
// GoToSocial
// Copyright (C) GoToSocial Authors admin@gotosocial.org
// SPDX-License-Identifier: AGPL-3.0-or-later
//
// This program is free software: you can redistribute it and/or modify
// it under the terms of the GNU Affero General Public License as published by
// the Free Software Foundation, either version 3 of the License, or
// (at your option) any later version.
//
// This program is distributed in the hope that it will be useful,
// but WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
// GNU Affero General Public License for more details.
//
// You should have received a copy of the GNU Affero General Public License
// along with this program. If not, see <http://www.gnu.org/licenses/>.
package dereferencing
import (
"context"
"fmt"
"io"
"net/url"
"github.com/superseriousbusiness/gotosocial/internal/media"
)
func (d *Dereferencer) GetRemoteMedia(ctx context.Context, requestingUsername string, accountID string, remoteURL string, ai *media.AdditionalMediaInfo) (*media.ProcessingMedia, error) {
if accountID == "" {
return nil, fmt.Errorf("GetRemoteMedia: account ID was empty")
}
t, err := d.transportController.NewTransportForUsername(ctx, requestingUsername)
if err != nil {
return nil, fmt.Errorf("GetRemoteMedia: error creating transport: %s", err)
}
derefURI, err := url.Parse(remoteURL)
if err != nil {
return nil, fmt.Errorf("GetRemoteMedia: error parsing url: %s", err)
}
dataFunc := func(innerCtx context.Context) (io.ReadCloser, int64, error) {
return t.DereferenceMedia(innerCtx, derefURI)
}
processingMedia, err := d.mediaManager.ProcessMedia(ctx, dataFunc, accountID, ai)
if err != nil {
return nil, fmt.Errorf("GetRemoteMedia: error processing attachment: %s", err)
}
return processingMedia, nil
}

View file

@ -1,161 +0,0 @@
// GoToSocial
// Copyright (C) GoToSocial Authors admin@gotosocial.org
// SPDX-License-Identifier: AGPL-3.0-or-later
//
// This program is free software: you can redistribute it and/or modify
// it under the terms of the GNU Affero General Public License as published by
// the Free Software Foundation, either version 3 of the License, or
// (at your option) any later version.
//
// This program is distributed in the hope that it will be useful,
// but WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
// GNU Affero General Public License for more details.
//
// You should have received a copy of the GNU Affero General Public License
// along with this program. If not, see <http://www.gnu.org/licenses/>.
package dereferencing_test
import (
"context"
"testing"
"time"
"github.com/stretchr/testify/suite"
"github.com/superseriousbusiness/gotosocial/internal/gtsmodel"
"github.com/superseriousbusiness/gotosocial/internal/media"
)
type AttachmentTestSuite struct {
DereferencerStandardTestSuite
}
func (suite *AttachmentTestSuite) TestDereferenceAttachmentBlocking() {
ctx := context.Background()
fetchingAccount := suite.testAccounts["local_account_1"]
attachmentOwner := "01FENS9F666SEQ6TYQWEEY78GM"
attachmentStatus := "01FENS9NTTVNEX1YZV7GB63MT8"
attachmentContentType := "image/jpeg"
attachmentURL := "https://s3-us-west-2.amazonaws.com/plushcity/media_attachments/files/106/867/380/219/163/828/original/88e8758c5f011439.jpg"
attachmentDescription := "It's a cute plushie."
attachmentBlurhash := "LtQ9yKi__4%g%MRjWCt7%hozM_az"
media, err := suite.dereferencer.GetRemoteMedia(ctx, fetchingAccount.Username, attachmentOwner, attachmentURL, &media.AdditionalMediaInfo{
StatusID: &attachmentStatus,
RemoteURL: &attachmentURL,
Description: &attachmentDescription,
Blurhash: &attachmentBlurhash,
})
suite.NoError(err)
// make a blocking call to load the attachment from the in-process media
attachment, err := media.LoadAttachment(ctx)
suite.NoError(err)
suite.NotNil(attachment)
suite.Equal(attachmentOwner, attachment.AccountID)
suite.Equal(attachmentStatus, attachment.StatusID)
suite.Equal(attachmentURL, attachment.RemoteURL)
suite.NotEmpty(attachment.URL)
suite.NotEmpty(attachment.Blurhash)
suite.NotEmpty(attachment.ID)
suite.NotEmpty(attachment.CreatedAt)
suite.NotEmpty(attachment.UpdatedAt)
suite.EqualValues(1.3365462, attachment.FileMeta.Original.Aspect)
suite.Equal(2071680, attachment.FileMeta.Original.Size)
suite.Equal(1245, attachment.FileMeta.Original.Height)
suite.Equal(1664, attachment.FileMeta.Original.Width)
suite.Equal(attachmentBlurhash, attachment.Blurhash)
suite.Equal(gtsmodel.ProcessingStatusProcessed, attachment.Processing)
suite.NotEmpty(attachment.File.Path)
suite.Equal(attachmentContentType, attachment.File.ContentType)
suite.Equal(attachmentDescription, attachment.Description)
suite.NotEmpty(attachment.Thumbnail.Path)
suite.NotEmpty(attachment.Type)
// attachment should also now be in the database
dbAttachment, err := suite.db.GetAttachmentByID(context.Background(), attachment.ID)
suite.NoError(err)
suite.NotNil(dbAttachment)
suite.Equal(attachmentOwner, dbAttachment.AccountID)
suite.Equal(attachmentStatus, dbAttachment.StatusID)
suite.Equal(attachmentURL, dbAttachment.RemoteURL)
suite.NotEmpty(dbAttachment.URL)
suite.NotEmpty(dbAttachment.Blurhash)
suite.NotEmpty(dbAttachment.ID)
suite.NotEmpty(dbAttachment.CreatedAt)
suite.NotEmpty(dbAttachment.UpdatedAt)
suite.EqualValues(1.3365462, dbAttachment.FileMeta.Original.Aspect)
suite.Equal(2071680, dbAttachment.FileMeta.Original.Size)
suite.Equal(1245, dbAttachment.FileMeta.Original.Height)
suite.Equal(1664, dbAttachment.FileMeta.Original.Width)
suite.Equal(attachmentBlurhash, dbAttachment.Blurhash)
suite.Equal(gtsmodel.ProcessingStatusProcessed, dbAttachment.Processing)
suite.NotEmpty(dbAttachment.File.Path)
suite.Equal(attachmentContentType, dbAttachment.File.ContentType)
suite.Equal(attachmentDescription, dbAttachment.Description)
suite.NotEmpty(dbAttachment.Thumbnail.Path)
suite.NotEmpty(dbAttachment.Type)
}
func (suite *AttachmentTestSuite) TestDereferenceAttachmentAsync() {
ctx := context.Background()
fetchingAccount := suite.testAccounts["local_account_1"]
attachmentOwner := "01FENS9F666SEQ6TYQWEEY78GM"
attachmentStatus := "01FENS9NTTVNEX1YZV7GB63MT8"
attachmentContentType := "image/jpeg"
attachmentURL := "https://s3-us-west-2.amazonaws.com/plushcity/media_attachments/files/106/867/380/219/163/828/original/88e8758c5f011439.jpg"
attachmentDescription := "It's a cute plushie."
attachmentBlurhash := "LtQ9yKi__4%g%MRjWCt7%hozM_az"
processingMedia, err := suite.dereferencer.GetRemoteMedia(ctx, fetchingAccount.Username, attachmentOwner, attachmentURL, &media.AdditionalMediaInfo{
StatusID: &attachmentStatus,
RemoteURL: &attachmentURL,
Description: &attachmentDescription,
Blurhash: &attachmentBlurhash,
})
suite.NoError(err)
attachmentID := processingMedia.AttachmentID()
time.Sleep(time.Second * 3)
// now get the attachment from the database
attachment, err := suite.db.GetAttachmentByID(ctx, attachmentID)
suite.NoError(err)
suite.NotNil(attachment)
suite.Equal(attachmentOwner, attachment.AccountID)
suite.Equal(attachmentStatus, attachment.StatusID)
suite.Equal(attachmentURL, attachment.RemoteURL)
suite.NotEmpty(attachment.URL)
suite.NotEmpty(attachment.Blurhash)
suite.NotEmpty(attachment.ID)
suite.NotEmpty(attachment.CreatedAt)
suite.NotEmpty(attachment.UpdatedAt)
suite.EqualValues(1.3365462, attachment.FileMeta.Original.Aspect)
suite.Equal(2071680, attachment.FileMeta.Original.Size)
suite.Equal(1245, attachment.FileMeta.Original.Height)
suite.Equal(1664, attachment.FileMeta.Original.Width)
suite.Equal(attachmentBlurhash, attachment.Blurhash)
suite.Equal(gtsmodel.ProcessingStatusProcessed, attachment.Processing)
suite.NotEmpty(attachment.File.Path)
suite.Equal(attachmentContentType, attachment.File.ContentType)
suite.Equal(attachmentDescription, attachment.Description)
suite.NotEmpty(attachment.Thumbnail.Path)
suite.NotEmpty(attachment.Type)
}
func TestAttachmentTestSuite(t *testing.T) {
suite.Run(t, new(AttachmentTestSuite))
}

View file

@ -789,7 +789,7 @@ func (d *Dereferencer) fetchStatusAttachments(ctx context.Context, tsport transp
for i := range status.Attachments {
attachment := status.Attachments[i]
// Look for existing media attachment with remoet URL first.
// Look for existing media attachment with remote URL first.
existing, ok := existing.GetAttachmentByRemoteURL(attachment.RemoteURL)
if ok && existing.ID != "" && *existing.Cached {
status.Attachments[i] = existing
@ -804,25 +804,33 @@ func (d *Dereferencer) fetchStatusAttachments(ctx context.Context, tsport transp
continue
}
// Start pre-processing remote media at remote URL.
processing, err := d.mediaManager.PreProcessMedia(ctx, func(ctx context.Context) (io.ReadCloser, int64, error) {
data := func(ctx context.Context) (io.ReadCloser, int64, error) {
return tsport.DereferenceMedia(ctx, remoteURL)
}, status.AccountID, &media.AdditionalMediaInfo{
}
ai := &media.AdditionalMediaInfo{
StatusID: &status.ID,
RemoteURL: &attachment.RemoteURL,
Description: &attachment.Description,
Blurhash: &attachment.Blurhash,
})
if err != nil {
log.Errorf(ctx, "error processing attachment: %v", err)
continue
}
// Start pre-processing remote media at remote URL.
processing := d.mediaManager.PreProcessMedia(data, status.AccountID, ai)
// Force attachment loading *right now*.
attachment, err = processing.LoadAttachment(ctx)
if err != nil {
log.Errorf(ctx, "error loading attachment: %v", err)
continue
if attachment == nil {
// Totally failed to load;
// bail on this attachment.
log.Errorf(ctx, "error loading attachment: %v", err)
continue
}
// Partially loaded. Keep as
// placeholder and try again later.
log.Warnf(ctx, "partially loaded attachment: %v", err)
}
// Set the *new* attachment and ID.
@ -832,8 +840,7 @@ func (d *Dereferencer) fetchStatusAttachments(ctx context.Context, tsport transp
for i := 0; i < len(status.AttachmentIDs); {
if status.AttachmentIDs[i] == "" {
// This is a failed attachment population, this may
// be due to us not currently supporting a media type.
// Remove totally failed attachment populations
copy(status.Attachments[i:], status.Attachments[i+1:])
copy(status.AttachmentIDs[i:], status.AttachmentIDs[i+1:])
status.Attachments = status.Attachments[:len(status.Attachments)-1]