Merge branch 'main' into patch-4

Dockerfile

@@ -1,20 +1,231 @@
 # syntax=docker/dockerfile:1
 # check=error=true
 
-FROM debian:bookworm-slim
+ARG FFMPEG_DATE="2025-01-10-19-43"
-
+ARG FFMPEG_VERSION="N-118280-g5cd49e1bfd"
-ARG TARGETARCH
-ARG TARGETPLATFORM
 
 ARG S6_VERSION="3.2.0.2"
+
 ARG SHA256_S6_AMD64="59289456ab1761e277bd456a95e737c06b03ede99158beb24f12b165a904f478"
 ARG SHA256_S6_ARM64="8b22a2eaca4bf0b27a43d36e65c89d2701738f628d1abd0cea5569619f66f785"
 ARG SHA256_S6_NOARCH="6dbcde158a3e78b9bb141d7bcb5ccb421e563523babbe2c64470e76f4fd02dae"
 
-ARG FFMPEG_DATE="autobuild-2024-12-24-14-15"
+ARG ALPINE_VERSION="latest"
-ARG FFMPEG_VERSION="N-118163-g954d55c2a4"
+ARG FFMPEG_PREFIX_FILE="ffmpeg-${FFMPEG_VERSION}"
-ARG SHA256_FFMPEG_AMD64="798a7e5a0724139e6bb70df8921522b23be27028f9f551dfa83c305ec4ffaf3a"
+ARG FFMPEG_SUFFIX_FILE=".tar.xz"
-ARG SHA256_FFMPEG_ARM64="c3e6cc0fec42cc7e3804014fbb02c1384a1a31ef13f6f9a36121f2e1216240c0"
+
+ARG FFMPEG_CHECKSUM_ALGORITHM="sha256"
+ARG S6_CHECKSUM_ALGORITHM="sha256"
+
+FROM alpine:${ALPINE_VERSION} AS ffmpeg-download
+ARG FFMPEG_DATE
+ARG FFMPEG_VERSION
+ARG FFMPEG_PREFIX_FILE
+ARG FFMPEG_SUFFIX_FILE
+ARG SHA256_FFMPEG_AMD64
+ARG SHA256_FFMPEG_ARM64
+ARG FFMPEG_CHECKSUM_ALGORITHM
+ARG CHECKSUM_ALGORITHM="${FFMPEG_CHECKSUM_ALGORITHM}"
+ARG FFMPEG_CHECKSUM_AMD64="${SHA256_FFMPEG_AMD64}"
+ARG FFMPEG_CHECKSUM_ARM64="${SHA256_FFMPEG_ARM64}"
+
+ARG FFMPEG_FILE_SUMS="checksums.${CHECKSUM_ALGORITHM}"
+ARG FFMPEG_URL="https://github.com/yt-dlp/FFmpeg-Builds/releases/download/autobuild-${FFMPEG_DATE}"
+
+ARG DESTDIR="/downloaded"
+ARG TARGETARCH
+ADD "${FFMPEG_URL}/${FFMPEG_FILE_SUMS}" "${DESTDIR}/"
+RUN set -eu ; \
+    apk --no-cache --no-progress add cmd:aria2c cmd:awk ; \
+\
+    aria2c_options() { \
+        algorithm="${CHECKSUM_ALGORITHM%[0-9]??}" ; \
+        bytes="${CHECKSUM_ALGORITHM#${algorithm}}" ; \
+        hash="$( awk -v fn="${1##*/}" '$0 ~ fn"$" { print $1; exit; }' "${DESTDIR}/${FFMPEG_FILE_SUMS}" )" ; \
+\
+        printf -- '\t%s\n' \
+          'allow-overwrite=true' \
+          'always-resume=false' \
+          'check-integrity=true' \
+          "checksum=${algorithm}-${bytes}=${hash}" \
+          'max-connection-per-server=2' \
+; \
+        printf -- '\n' ; \
+    } ; \
+\
+    decide_arch() { \
+        case "${TARGETARCH}" in \
+            (amd64) printf -- 'linux64' ;; \
+            (arm64) printf -- 'linuxarm64' ;; \
+        esac ; \
+    } ; \
+\
+    FFMPEG_ARCH="$(decide_arch)" ; \
+    FFMPEG_PREFIX_FILE="$( printf -- '%s' "${FFMPEG_PREFIX_FILE}" | cut -d '-' -f 1,2 )" ; \
+    for url in $(awk ' \
+      $2 ~ /^[*]?'"${FFMPEG_PREFIX_FILE}"'/ && /-'"${FFMPEG_ARCH}"'-/ { $1=""; print; } \
+      ' "${DESTDIR}/${FFMPEG_FILE_SUMS}") ; \
+    do \
+        url="${FFMPEG_URL}/${url# }" ; \
+        printf -- '%s\n' "${url}" ; \
+        aria2c_options "${url}" ; \
+        printf -- '\n' ; \
+    done > /tmp/downloads ; \
+    unset -v url ; \
+\
+    aria2c --no-conf=true \
+      --dir /downloaded \
+      --lowest-speed-limit='16K' \
+      --show-console-readout=false \
+      --summary-interval=0 \
+      --input-file /tmp/downloads ; \
+\
+    apk --no-cache --no-progress add "cmd:${CHECKSUM_ALGORITHM}sum" ; \
+\    
+    decide_expected() { \
+        case "${TARGETARCH}" in \
+            (amd64) printf -- '%s' "${FFMPEG_CHECKSUM_AMD64}" ;; \
+            (arm64) printf -- '%s' "${FFMPEG_CHECKSUM_ARM64}" ;; \
+        esac ; \
+    } ; \
+\
+    FFMPEG_HASH="$(decide_expected)" ; \
+\
+    cd "${DESTDIR}" ; \
+    if [ -n "${FFMPEG_HASH}" ] ; \
+    then \
+        printf -- '%s *%s\n' "${FFMPEG_HASH}" "${FFMPEG_PREFIX_FILE}"*-"${FFMPEG_ARCH}"-*"${FFMPEG_SUFFIX_FILE}" >> /tmp/SUMS ; \
+        "${CHECKSUM_ALGORITHM}sum" --check --warn --strict /tmp/SUMS || exit ; \
+    fi ; \
+    "${CHECKSUM_ALGORITHM}sum" --check --warn --strict --ignore-missing "${DESTDIR}/${FFMPEG_FILE_SUMS}" ; \
+\
+    mkdir -v -p "/verified/${TARGETARCH}" ; \
+    ln -v "${FFMPEG_PREFIX_FILE}"*-"${FFMPEG_ARCH}"-*"${FFMPEG_SUFFIX_FILE}" "/verified/${TARGETARCH}/" ; \
+    rm -rf "${DESTDIR}" ;
+
+FROM alpine:${ALPINE_VERSION} AS ffmpeg-extracted
+COPY --from=ffmpeg-download /verified /verified
+
+ARG FFMPEG_PREFIX_FILE
+ARG FFMPEG_SUFFIX_FILE
+ARG TARGETARCH
+RUN set -eux ; \
+    mkdir -v /extracted ; \
+    cd /extracted ; \
+    ln -s "/verified/${TARGETARCH}"/"${FFMPEG_PREFIX_FILE}"*"${FFMPEG_SUFFIX_FILE}" "/tmp/ffmpeg${FFMPEG_SUFFIX_FILE}" ; \
+    tar -tf "/tmp/ffmpeg${FFMPEG_SUFFIX_FILE}" | grep '/bin/\(ffmpeg\|ffprobe\)' > /tmp/files ; \
+    tar -xop \
+      --strip-components=2 \
+      -f "/tmp/ffmpeg${FFMPEG_SUFFIX_FILE}" \
+      -T /tmp/files ; \
+\
+    ls -AlR /extracted ;
+
+FROM scratch AS ffmpeg
+COPY --from=ffmpeg-extracted /extracted /usr/local/bin/
+
+FROM alpine:${ALPINE_VERSION} AS s6-overlay-download
+ARG S6_VERSION
+ARG SHA256_S6_AMD64
+ARG SHA256_S6_ARM64
+ARG SHA256_S6_NOARCH
+
+ARG DESTDIR="/downloaded"
+ARG S6_CHECKSUM_ALGORITHM
+ARG CHECKSUM_ALGORITHM="${S6_CHECKSUM_ALGORITHM}"
+
+ARG S6_CHECKSUM_AMD64="${CHECKSUM_ALGORITHM}:${SHA256_S6_AMD64}"
+ARG S6_CHECKSUM_ARM64="${CHECKSUM_ALGORITHM}:${SHA256_S6_ARM64}"
+ARG S6_CHECKSUM_NOARCH="${CHECKSUM_ALGORITHM}:${SHA256_S6_NOARCH}"
+
+ARG S6_OVERLAY_URL="https://github.com/just-containers/s6-overlay/releases/download/v${S6_VERSION}"
+ARG S6_PREFIX_FILE="s6-overlay-"
+ARG S6_SUFFIX_FILE=".tar.xz"
+
+ARG S6_FILE_AMD64="${S6_PREFIX_FILE}x86_64${S6_SUFFIX_FILE}"
+ARG S6_FILE_ARM64="${S6_PREFIX_FILE}aarch64${S6_SUFFIX_FILE}"
+ARG S6_FILE_NOARCH="${S6_PREFIX_FILE}noarch${S6_SUFFIX_FILE}"
+
+ADD "${S6_OVERLAY_URL}/${S6_FILE_AMD64}.${CHECKSUM_ALGORITHM}" "${DESTDIR}/"
+ADD "${S6_OVERLAY_URL}/${S6_FILE_ARM64}.${CHECKSUM_ALGORITHM}" "${DESTDIR}/"
+ADD "${S6_OVERLAY_URL}/${S6_FILE_NOARCH}.${CHECKSUM_ALGORITHM}" "${DESTDIR}/"
+
+##ADD --checksum="${S6_CHECKSUM_AMD64}" "${S6_OVERLAY_URL}/${S6_FILE_AMD64}" "${DESTDIR}/"
+##ADD --checksum="${S6_CHECKSUM_ARM64}" "${S6_OVERLAY_URL}/${S6_FILE_ARM64}" "${DESTDIR}/"
+##ADD --checksum="${S6_CHECKSUM_NOARCH}" "${S6_OVERLAY_URL}/${S6_FILE_NOARCH}" "${DESTDIR}/"
+
+# --checksum wasn't recognized, so use busybox to check the sums instead
+ADD "${S6_OVERLAY_URL}/${S6_FILE_AMD64}" "${DESTDIR}/"
+RUN set -eu ; checksum="${S6_CHECKSUM_AMD64}" ; file="${S6_FILE_AMD64}" ; cd "${DESTDIR}/" && \
+    printf -- '%s *%s\n' "$(printf -- '%s' "${checksum}" | cut -d : -f 2-)" "${file}" | "${CHECKSUM_ALGORITHM}sum" -cw
+
+ADD "${S6_OVERLAY_URL}/${S6_FILE_ARM64}" "${DESTDIR}/"
+RUN set -eu ; checksum="${S6_CHECKSUM_ARM64}" ; file="${S6_FILE_ARM64}" ; cd "${DESTDIR}/" && \
+    printf -- '%s *%s\n' "$(printf -- '%s' "${checksum}" | cut -d : -f 2-)" "${file}" | "${CHECKSUM_ALGORITHM}sum" -cw
+
+ADD "${S6_OVERLAY_URL}/${S6_FILE_NOARCH}" "${DESTDIR}/"
+RUN set -eu ; checksum="${S6_CHECKSUM_NOARCH}" ; file="${S6_FILE_NOARCH}" ; cd "${DESTDIR}/" && \
+    printf -- '%s *%s\n' "$(printf -- '%s' "${checksum}" | cut -d : -f 2-)" "${file}" | "${CHECKSUM_ALGORITHM}sum" -cw
+
+FROM alpine:${ALPINE_VERSION} AS s6-overlay-extracted
+COPY --from=s6-overlay-download /downloaded /downloaded
+
+ARG S6_CHECKSUM_ALGORITHM
+ARG CHECKSUM_ALGORITHM="${S6_CHECKSUM_ALGORITHM}"
+
+ARG TARGETARCH
+
+RUN set -eu ; \
+\
+    decide_arch() { \
+      local arg1 ; \
+      arg1="${1:-$(uname -m)}" ; \
+\
+      case "${arg1}" in \
+        (amd64) printf -- 'x86_64' ;; \
+        (arm64) printf -- 'aarch64' ;; \
+        (armv7l) printf -- 'arm' ;; \
+        (*) printf -- '%s' "${arg1}" ;; \
+      esac ; \
+      unset -v arg1 ; \
+    } ; \
+\
+    apk --no-cache --no-progress add "cmd:${CHECKSUM_ALGORITHM}sum" ; \
+    mkdir -v /verified ; \
+    cd /downloaded ; \
+    for f in *.sha256 ; \
+    do \
+      "${CHECKSUM_ALGORITHM}sum" --check --warn --strict "${f}" || exit ; \
+      ln -v "${f%.sha256}" /verified/ || exit ; \
+    done ; \
+    unset -v f ; \
+\
+    S6_ARCH="$(decide_arch "${TARGETARCH}")" ; \
+    set -x ; \
+    mkdir -v /s6-overlay-rootfs ; \
+    cd /s6-overlay-rootfs ; \
+    for f in /verified/*.tar* ; \
+    do \
+      case "${f}" in \
+        (*-noarch.tar*|*-"${S6_ARCH}".tar*) \
+          tar -xpf "${f}" || exit ;; \
+      esac ; \
+    done ; \
+    set +x ; \
+    unset -v f ;
+
+FROM scratch AS s6-overlay
+COPY --from=s6-overlay-extracted /s6-overlay-rootfs /
+
+FROM debian:bookworm-slim AS tubesync
+
+ARG TARGETARCH
+ARG TARGETPLATFORM
+
+ARG S6_VERSION
+
+ARG FFMPEG_DATE
+ARG FFMPEG_VERSION
 
 ENV S6_VERSION="${S6_VERSION}" \
   FFMPEG_DATE="${FFMPEG_DATE}" \
@@ -29,89 +240,20 @@ ENV DEBIAN_FRONTEND="noninteractive" \
   S6_CMD_WAIT_FOR_SERVICES_MAXTIME="0"
 
 # Install third party software
+COPY --from=s6-overlay / /
+COPY --from=ffmpeg /usr/local/bin/ /usr/local/bin/
+
 # Reminder: the SHELL handles all variables
-RUN decide_arch() { \
+RUN set -x && \
-      case "${TARGETARCH:=amd64}" in \
-        (arm64) printf -- 'aarch64' ;; \
-        (*) printf -- '%s' "${TARGETARCH}" ;; \
-      esac ; \
-    } && \
-    decide_expected() { \
-      case "${1}" in \
-        (ffmpeg) case "${2}" in \
-            (amd64) printf -- '%s' "${SHA256_FFMPEG_AMD64}" ;; \
-            (arm64) printf -- '%s' "${SHA256_FFMPEG_ARM64}" ;; \
-          esac ;; \
-        (s6) case "${2}" in \
-            (amd64) printf -- '%s' "${SHA256_S6_AMD64}" ;; \
-            (arm64) printf -- '%s' "${SHA256_S6_ARM64}" ;; \
-            (noarch) printf -- '%s' "${SHA256_S6_NOARCH}" ;; \
-          esac ;; \
-      esac ; \
-    } && \
-    decide_url() { \
-      case "${1}" in \
-        (ffmpeg) printf -- \
-          'https://github.com/yt-dlp/FFmpeg-Builds/releases/download/%s/ffmpeg-%s-linux%s-gpl%s.tar.xz' \
-          "${FFMPEG_DATE}" \
-          "${FFMPEG_VERSION}" \
-          "$(case "${2}" in \
-            (amd64) printf -- '64' ;; \
-            (*) printf -- '%s' "${2}" ;; \
-          esac)" \
-          "$(case "${FFMPEG_VERSION%%-*}" in \
-            (n*) printf -- '-%s\n' "${FFMPEG_VERSION#n}" | cut -d '-' -f 1,2 ;; \
-            (*) printf -- '' ;; \
-          esac)" ;; \
-        (s6) printf -- \
-          'https://github.com/just-containers/s6-overlay/releases/download/v%s/s6-overlay-%s.tar.xz' \
-          "${S6_VERSION}" \
-          "$(case "${2}" in \
-            (amd64) printf -- 'x86_64' ;; \
-            (arm64) printf -- 'aarch64' ;; \
-            (*) printf -- '%s' "${2}" ;; \
-          esac)" ;; \
-      esac ; \
-    } && \
-    verify_download() { \
-      while [ $# -ge 2 ] ; do \
-        sha256sum "${2}" ; \
-        printf -- '%s  %s\n' "${1}" "${2}" | sha256sum -c || return ; \
-        shift ; shift ; \
-      done ; \
-    } && \
-    download_expected_file() { \
-      local arg1 expected file url ; \
-      arg1="$(printf -- '%s\n' "${1}" | awk '{print toupper($0);}')" ; \
-      expected="$(decide_expected "${1}" "${2}")" ; \
-      file="${3}" ; \
-      url="$(decide_url "${1}" "${2}")" ; \
-      printf -- '%s\n' \
-        "Building for arch: ${2}|${ARCH}, downloading ${arg1} from: ${url}, expecting ${arg1} SHA256: ${expected}" && \
-      rm -rf "${file}" && \
-      curl --disable --output "${file}" --clobber --location --no-progress-meter --url "${url}" && \
-      verify_download "${expected}" "${file}" ; \
-    } && \
-  export ARCH="$(decide_arch)" && \
-  set -x && \
   apt-get update && \
   apt-get -y --no-install-recommends install locales && \
   printf -- "en_US.UTF-8 UTF-8\n" > /etc/locale.gen && \
   locale-gen en_US.UTF-8 && \
   # Install required distro packages
   apt-get -y --no-install-recommends install curl ca-certificates file binutils xz-utils && \
-  # Install s6
+  # Installed s6 (using COPY earlier)
-  _file="/tmp/s6-overlay-noarch.tar.xz" && \
-  download_expected_file s6 noarch "${_file}" && \
-  tar -C / -xpf "${_file}" && rm -f "${_file}" && \
-  _file="/tmp/s6-overlay-${ARCH}.tar.xz" && \
-  download_expected_file s6 "${TARGETARCH}" "${_file}" && \
-  tar -C / -xpf "${_file}" && rm -f "${_file}" && \
   file -L /command/s6-overlay-suexec && \
-  # Install ffmpeg
+  # Installed ffmpeg (using COPY earlier)
-  _file="/tmp/ffmpeg-${ARCH}.tar.xz" && \
-  download_expected_file ffmpeg "${TARGETARCH}" "${_file}" && \
-  tar -xvvpf "${_file}" --strip-components=2 --no-anchored -C /usr/local/bin/ "ffmpeg" "ffprobe" && rm -f "${_file}" && \
   /usr/local/bin/ffmpeg -version && \
   file /usr/local/bin/ff* && \
   # Clean up
@@ -157,7 +299,9 @@ ENV PIP_NO_COMPILE=1 \
 WORKDIR /app
 
 # Set up the app
-RUN set -x && \
+#BuildKit#RUN --mount=type=bind,source=Pipfile,target=/app/Pipfile \
+RUN \
+  set -x && \
   apt-get update && \
   # Install required build packages
   apt-get -y --no-install-recommends install \

tubesync/sync/models.py

@@ -19,7 +19,7 @@ from common.utils import clean_filename, clean_emoji
 from .youtube import (get_media_info as get_youtube_media_info,
                       download_media as download_youtube_media,
                       get_channel_image_info as get_youtube_channel_image_info)
-from .utils import seconds_to_timestr, parse_media_format
+from .utils import seconds_to_timestr, parse_media_format, filter_response
 from .matching import (get_best_combined_format, get_best_audio_format,
                        get_best_video_format)
 from .mediaservers import PlexMediaServer
@@ -589,6 +589,7 @@ class Source(models.Model):
             'key': 'SoMeUnIqUiD',
             'format': '-'.join(fmt),
             'playlist_title': 'Some Playlist Title',
+            'video_order': '01',
             'ext': self.extension,
             'resolution': self.source_resolution if self.source_resolution else '',
             'height': '720' if self.source_resolution else '',
@@ -1128,6 +1129,7 @@ class Media(models.Model):
             'key': self.key,
             'format': '-'.join(display_format['format']),
             'playlist_title': self.playlist_title,
+            'video_order': self.get_episode_str(True),
             'ext': self.source.extension,
             'resolution': display_format['resolution'],
             'height': display_format['height'],
@@ -1143,8 +1145,39 @@ class Media(models.Model):
     def has_metadata(self):
         return self.metadata is not None
 
+
+    @property
+    def reduce_data(self):
+        try:
+            from common.logger import log
+            from common.utils import json_serial
+
+            old_mdl = len(self.metadata or "")
+            data = json.loads(self.metadata or "")
+            compact_json = json.dumps(data, separators=(',', ':'), default=json_serial)
+            
+            filtered_data = filter_response(data, True)
+            filtered_json = json.dumps(filtered_data, separators=(',', ':'), default=json_serial)
+        except Exception as e:
+            log.exception('reduce_data: %s', e)
+        else:
+            # log the results of filtering / compacting on metadata size
+            new_mdl = len(compact_json)
+            if old_mdl > new_mdl:
+                delta = old_mdl - new_mdl
+                log.info(f'{self.key}: metadata compacted by {delta:,} characters ({old_mdl:,} -> {new_mdl:,})')
+            new_mdl = len(filtered_json)
+            if old_mdl > new_mdl:
+                delta = old_mdl - new_mdl
+                log.info(f'{self.key}: metadata reduced by {delta:,} characters ({old_mdl:,} -> {new_mdl:,})')
+                if getattr(settings, 'SHRINK_OLD_MEDIA_METADATA', False):
+                    self.metadata = filtered_json
+
+
     @property
     def loaded_metadata(self):
+        if getattr(settings, 'SHRINK_OLD_MEDIA_METADATA', False):
+            self.reduce_data
         try:
             data = json.loads(self.metadata)
             if not isinstance(data, dict):
@@ -1263,8 +1296,7 @@ class Media(models.Model):
 
     @property
     def directory_path(self):
-        dirname = self.source.directory_path / self.filename
+        return self.filepath.parent
-        return dirname.parent
 
     @property
     def filepath(self):
@@ -1373,8 +1405,7 @@ class Media(models.Model):
         nfo.append(season)
         # episode = number of video in the year
         episode = nfo.makeelement('episode', {})
-        episode_number = self.calculate_episode_number()
+        episode.text = self.get_episode_str()
-        episode.text = str(episode_number) if episode_number else ''
         episode.tail = '\n  '
         nfo.append(episode)
         # ratings = media metadata youtube rating
@@ -1524,6 +1555,16 @@ class Media(models.Model):
                 return position_counter
             position_counter += 1
 
+    def get_episode_str(self, use_padding=False):
+        episode_number = self.calculate_episode_number()
+        if not episode_number:
+            return ''
+
+        if use_padding:
+            return f'{episode_number:02}'
+        
+        return str(episode_number)
+
 
 class MediaServer(models.Model):
     '''

tubesync/sync/tasks.py

@@ -26,7 +26,7 @@ from common.errors import NoMediaException, DownloadFailedException
 from common.utils import json_serial
 from .models import Source, Media, MediaServer
 from .utils import (get_remote_image, resize_image_to_height, delete_file,
-                    write_text_file)
+                    write_text_file, filter_response)
 from .filtering import filter_media
 
 
@@ -304,7 +304,10 @@ def download_media_metadata(media_id):
         return
     source = media.source
     metadata = media.index_metadata()
-    media.metadata = json.dumps(metadata, default=json_serial)
+    response = metadata
+    if getattr(settings, 'SHRINK_NEW_MEDIA_METADATA', False):
+        response = filter_response(metadata, True)
+    media.metadata = json.dumps(response, separators=(',', ':'), default=json_serial)
     upload_date = media.upload_date
     # Media must have a valid upload date
     if upload_date:

tubesync/sync/templates/sync/_mediaformatvars.html

@@ -73,6 +73,11 @@
       <td>Playlist title of media, if it's in a playlist</td>
       <td>Some Playlist</td>
     </tr>
+    <tr>
+      <td>{video_order}</td>
+      <td>Episode order in playlist, if in playlist <sub><sup>(can cause issues if playlist is changed after adding)</sup></sub></td>
+      <td>01</td>
+    </tr>
     <tr>
       <td>{ext}</td>
       <td>File extension</td>

tubesync/sync/tests.py

@@ -18,6 +18,7 @@ from background_task.models import Task
 from .models import Source, Media
 from .tasks import cleanup_old_media
 from .filtering import filter_media
+from .utils import filter_response
 
 
 class FrontEndTestCase(TestCase):
@@ -1709,6 +1710,84 @@ class FormatMatchingTestCase(TestCase):
                     f'expected {expected_match_result}')
 
 
+class ResponseFilteringTestCase(TestCase):
+
+    def setUp(self):
+        # Disable general logging for test case
+        logging.disable(logging.CRITICAL)
+        # Add a test source
+        self.source = Source.objects.create(
+            source_type=Source.SOURCE_TYPE_YOUTUBE_CHANNEL,
+            key='testkey',
+            name='testname',
+            directory='testdirectory',
+            index_schedule=3600,
+            delete_old_media=False,
+            days_to_keep=14,
+            source_resolution=Source.SOURCE_RESOLUTION_1080P,
+            source_vcodec=Source.SOURCE_VCODEC_VP9,
+            source_acodec=Source.SOURCE_ACODEC_OPUS,
+            prefer_60fps=False,
+            prefer_hdr=False,
+            fallback=Source.FALLBACK_FAIL
+        )
+        # Add some media
+        self.media = Media.objects.create(
+            key='mediakey',
+            source=self.source,
+            metadata='{}'
+        )
+
+    def test_metadata_20230629(self):
+        self.media.metadata = all_test_metadata['20230629']
+        self.media.save()
+
+        unfiltered = self.media.loaded_metadata
+        filtered = filter_response(self.media.loaded_metadata)
+        self.assertIn('formats', unfiltered.keys())
+        self.assertIn('formats', filtered.keys())
+        # filtered 'downloader_options'
+        self.assertIn('downloader_options', unfiltered['formats'][10].keys())
+        self.assertNotIn('downloader_options', filtered['formats'][10].keys())
+        # filtered 'http_headers'
+        self.assertIn('http_headers', unfiltered['formats'][0].keys())
+        self.assertNotIn('http_headers', filtered['formats'][0].keys())
+        # did not lose any formats
+        self.assertEqual(48, len(unfiltered['formats']))
+        self.assertEqual(48, len(filtered['formats']))
+        self.assertEqual(len(unfiltered['formats']), len(filtered['formats']))
+        # did not remove everything with url
+        self.assertIn('original_url', unfiltered.keys())
+        self.assertIn('original_url', filtered.keys())
+        self.assertEqual(unfiltered['original_url'], filtered['original_url'])
+        # did reduce the size of the metadata
+        self.assertTrue(len(str(filtered)) < len(str(unfiltered)))
+
+        url_keys = []
+        for format in unfiltered['formats']:
+            for key in format.keys():
+                if 'url' in key:
+                    url_keys.append((format['format_id'], key, format[key],))
+        unfiltered_url_keys = url_keys
+        self.assertEqual(63, len(unfiltered_url_keys), msg=str(unfiltered_url_keys))
+
+        url_keys = []
+        for format in filtered['formats']:
+            for key in format.keys():
+                if 'url' in key:
+                    url_keys.append((format['format_id'], key, format[key],))
+        filtered_url_keys = url_keys
+        self.assertEqual(3, len(filtered_url_keys), msg=str(filtered_url_keys))
+
+        url_keys = []
+        for lang_code, captions in filtered['automatic_captions'].items():
+            for caption in captions:
+                for key in caption.keys():
+                    if 'url' in key:
+                        url_keys.append((lang_code, caption['ext'], caption[key],))
+        self.assertEqual(0, len(url_keys), msg=str(url_keys))
+
+
 class TasksTestCase(TestCase):
 
     def setUp(self):

tubesync/sync/utils.py

@@ -1,6 +1,7 @@
 import os
 import re
 import math
+from copy import deepcopy
 from operator import itemgetter
 from pathlib import Path
 from tempfile import NamedTemporaryFile
@@ -171,6 +172,95 @@ def normalize_codec(codec_str):
     return result
 
 
+def _url_keys(arg_dict, filter_func):
+    result = {}
+    for key in arg_dict.keys():
+        if 'url' in key:
+            result.update(
+                {key: filter_func(key=key, url=arg_dict[key])}
+            )
+    return result
+
+
+def _drop_url_keys(arg_dict, key, filter_func):
+    if key in arg_dict.keys():
+        for val_dict in arg_dict[key]:
+            for url_key, remove in _url_keys(val_dict, filter_func).items():
+                if remove is True:
+                    del val_dict[url_key]
+
+
+def filter_response(arg_dict, copy_arg=False):
+    '''
+        Clean up the response so as to not store useless metadata in the database.
+    '''
+    response_dict = arg_dict
+    # raise an exception for an unexpected argument type
+    if not isinstance(response_dict, dict):
+        raise TypeError(f'response_dict must be a dict, got "{type(response_dict)}"')
+
+    if copy_arg:
+        response_dict = deepcopy(arg_dict)
+
+    # optimize the empty case
+    if not response_dict:
+        return response_dict
+
+    # beginning of formats cleanup {{{
+    # drop urls that expire, or restrict IPs
+    def drop_format_url(**kwargs):
+        url = kwargs['url']
+        return (
+            url
+            and '://' in url
+            and (
+                '/ip/' in url
+                or 'ip=' in url
+                or '/expire/' in url
+                or 'expire=' in url
+            )
+        )
+
+    # these format keys are not useful to us
+    drop_keys = frozenset((
+        'downloader_options',
+        'fragments',
+        'http_headers',
+        '__needs_testing',
+        '__working',
+    ))
+    for key in frozenset(('formats', 'requested_formats',)):
+        _drop_url_keys(response_dict, key, drop_format_url)
+        if key in response_dict.keys():
+            for format in response_dict[key]:
+                for drop_key in drop_keys:
+                    if drop_key in format.keys():
+                        del format[drop_key]
+    # end of formats cleanup }}}
+
+    # beginning of subtitles cleanup {{{
+    # drop urls that expire
+    def drop_subtitles_url(**kwargs):
+        url = kwargs['url']
+        return (
+            url
+            and '://' in url
+            and (
+                '/expire/' in url
+                or '&expire=' in url
+            )
+        )
+
+    for key in frozenset(('subtitles', 'automatic_captions',)):
+        if key in response_dict.keys():
+            key_dict = response_dict[key]
+            for lang_code in key_dict:
+                _drop_url_keys(key_dict, lang_code, drop_subtitles_url)
+    # end of subtitles cleanup }}}
+
+    return response_dict
+
+
 def parse_media_format(format_dict):
     '''
         This parser primarily adapts the format dict returned by youtube-dl into a