Add an admin api to delete local media. (#8519)

Related to: #6459, #3479

Add `DELETE /_synapse/admin/v1/media/<server_name>/<media_id>` to delete
a single file from server.

changelog.d/8519.feature

+Add an admin api to delete a single file or files were not used for a defined time from server. Contributed by @dklimpel.
\ No newline at end of file

docs/admin_api/media_admin_api.md

@@ -100,3 +100,82 @@ Response:
   "num_quarantined": 10  # The number of media items successfully quarantined
 }
 ```
+
+# Delete local media
+This API deletes the *local* media from the disk of your own server.
+This includes any local thumbnails and copies of media downloaded from
+remote homeservers.
+This API will not affect media that has been uploaded to external
+media repositories (e.g https://github.com/turt2live/matrix-media-repo/).
+See also [purge_remote_media.rst](purge_remote_media.rst).
+
+## Delete a specific local media
+Delete a specific `media_id`.
+
+Request:
+
+```
+DELETE /_synapse/admin/v1/media/<server_name>/<media_id>
+
+{}
+```
+
+URL Parameters
+
+* `server_name`: string - The name of your local server (e.g `matrix.org`)
+* `media_id`: string - The ID of the media (e.g `abcdefghijklmnopqrstuvwx`)
+
+Response:
+
+```json
+    {
+       "deleted_media": [
+          "abcdefghijklmnopqrstuvwx"
+       ],
+       "total": 1
+    }
+```
+
+The following fields are returned in the JSON response body:
+
+* `deleted_media`: an array of strings - List of deleted `media_id`
+* `total`: integer - Total number of deleted `media_id`
+
+## Delete local media by date or size
+
+Request:
+
+```
+POST /_synapse/admin/v1/media/<server_name>/delete?before_ts=<before_ts>
+
+{}
+```
+
+URL Parameters
+
+* `server_name`: string - The name of your local server (e.g `matrix.org`).
+* `before_ts`: string representing a positive integer - Unix timestamp in ms.
+Files that were last used before this timestamp will be deleted. It is the timestamp of
+last access and not the timestamp creation. 
+* `size_gt`: Optional - string representing a positive integer - Size of the media in bytes.
+Files that are larger will be deleted. Defaults to `0`.
+* `keep_profiles`: Optional - string representing a boolean - Switch to also delete files
+that are still used in image data (e.g user profile, room avatar).
+If `false` these files will be deleted. Defaults to `true`.
+
+Response:
+
+```json
+    {
+       "deleted_media": [
+          "abcdefghijklmnopqrstuvwx",
+          "abcdefghijklmnopqrstuvwz"
+       ],
+       "total": 2
+    }
+```
+
+The following fields are returned in the JSON response body:
+
+* `deleted_media`: an array of strings - List of deleted `media_id`
+* `total`: integer - Total number of deleted `media_id`

synapse/rest/admin/media.py

@@ -16,9 +16,10 @@
 
 import logging
 
-from synapse.api.errors import AuthError
-from synapse.http.servlet import RestServlet, parse_integer
+from synapse.api.errors import AuthError, Codes, NotFoundError, SynapseError
+from synapse.http.servlet import RestServlet, parse_boolean, parse_integer
 from synapse.rest.admin._base import (
+    admin_patterns,
     assert_requester_is_admin,
     assert_user_is_admin,
     historical_admin_path_patterns,
@@ -150,6 +151,80 @@ class PurgeMediaCacheRestServlet(RestServlet):
         return 200, ret
 
 
+class DeleteMediaByID(RestServlet):
+    """Delete local media by a given ID. Removes it from this server.
+    """
+
+    PATTERNS = admin_patterns("/media/(?P<server_name>[^/]+)/(?P<media_id>[^/]+)")
+
+    def __init__(self, hs):
+        self.store = hs.get_datastore()
+        self.auth = hs.get_auth()
+        self.server_name = hs.hostname
+        self.media_repository = hs.get_media_repository()
+
+    async def on_DELETE(self, request, server_name: str, media_id: str):
+        await assert_requester_is_admin(self.auth, request)
+
+        if self.server_name != server_name:
+            raise SynapseError(400, "Can only delete local media")
+
+        if await self.store.get_local_media(media_id) is None:
+            raise NotFoundError("Unknown media")
+
+        logging.info("Deleting local media by ID: %s", media_id)
+
+        deleted_media, total = await self.media_repository.delete_local_media(media_id)
+        return 200, {"deleted_media": deleted_media, "total": total}
+
+
+class DeleteMediaByDateSize(RestServlet):
+    """Delete local media and local copies of remote media by
+    timestamp and size.
+    """
+
+    PATTERNS = admin_patterns("/media/(?P<server_name>[^/]+)/delete")
+
+    def __init__(self, hs):
+        self.store = hs.get_datastore()
+        self.auth = hs.get_auth()
+        self.server_name = hs.hostname
+        self.media_repository = hs.get_media_repository()
+
+    async def on_POST(self, request, server_name: str):
+        await assert_requester_is_admin(self.auth, request)
+
+        before_ts = parse_integer(request, "before_ts", required=True)
+        size_gt = parse_integer(request, "size_gt", default=0)
+        keep_profiles = parse_boolean(request, "keep_profiles", default=True)
+
+        if before_ts < 0:
+            raise SynapseError(
+                400,
+                "Query parameter before_ts must be a string representing a positive integer.",
+                errcode=Codes.INVALID_PARAM,
+            )
+        if size_gt < 0:
+            raise SynapseError(
+                400,
+                "Query parameter size_gt must be a string representing a positive integer.",
+                errcode=Codes.INVALID_PARAM,
+            )
+
+        if self.server_name != server_name:
+            raise SynapseError(400, "Can only delete local media")
+
+        logging.info(
+            "Deleting local media by timestamp: %s, size larger than: %s, keep profile media: %s"
+            % (before_ts, size_gt, keep_profiles)
+        )
+
+        deleted_media, total = await self.media_repository.delete_old_local_media(
+            before_ts, size_gt, keep_profiles
+        )
+        return 200, {"deleted_media": deleted_media, "total": total}
+
+
 def register_servlets_for_media_repo(hs, http_server):
     """
     Media repo specific APIs.
@@ -159,3 +234,5 @@ def register_servlets_for_media_repo(hs, http_server):
     QuarantineMediaByID(hs).register(http_server)
     QuarantineMediaByUser(hs).register(http_server)
     ListMediaInRoom(hs).register(http_server)
+    DeleteMediaByID(hs).register(http_server)
+    DeleteMediaByDateSize(hs).register(http_server)

synapse/rest/media/v1/filepath.py

@@ -69,6 +69,23 @@ class MediaFilePaths:
 
     local_media_thumbnail = _wrap_in_base_path(local_media_thumbnail_rel)
 
+    def local_media_thumbnail_dir(self, media_id: str) -> str:
+        """
+        Retrieve the local store path of thumbnails of a given media_id
+
+        Args:
+            media_id: The media ID to query.
+        Returns:
+            Path of local_thumbnails from media_id
+        """
+        return os.path.join(
+            self.base_path,
+            "local_thumbnails",
+            media_id[0:2],
+            media_id[2:4],
+            media_id[4:],
+        )
+
     def remote_media_filepath_rel(self, server_name, file_id):
         return os.path.join(
             "remote_content", server_name, file_id[0:2], file_id[2:4], file_id[4:]

synapse/rest/media/v1/media_repository.py

@@ -18,7 +18,7 @@ import errno
 import logging
 import os
 import shutil
-from typing import IO, Dict, Optional, Tuple
+from typing import IO, Dict, List, Optional, Tuple
 
 import twisted.internet.error
 import twisted.web.http
@@ -767,6 +767,76 @@ class MediaRepository:
 
         return {"deleted": deleted}
 
+    async def delete_local_media(self, media_id: str) -> Tuple[List[str], int]:
+        """
+        Delete the given local or remote media ID from this server
+
+        Args:
+            media_id: The media ID to delete.
+        Returns:
+            A tuple of (list of deleted media IDs, total deleted media IDs).
+        """
+        return await self._remove_local_media_from_disk([media_id])
+
+    async def delete_old_local_media(
+        self, before_ts: int, size_gt: int = 0, keep_profiles: bool = True,
+    ) -> Tuple[List[str], int]:
+        """
+        Delete local or remote media from this server by size and timestamp. Removes
+        media files, any thumbnails and cached URLs.
+
+        Args:
+            before_ts: Unix timestamp in ms.
+                       Files that were last used before this timestamp will be deleted
+            size_gt: Size of the media in bytes. Files that are larger will be deleted
+            keep_profiles: Switch to delete also files that are still used in image data
+                           (e.g user profile, room avatar)
+                           If false these files will be deleted
+        Returns:
+            A tuple of (list of deleted media IDs, total deleted media IDs).
+        """
+        old_media = await self.store.get_local_media_before(
+            before_ts, size_gt, keep_profiles,
+        )
+        return await self._remove_local_media_from_disk(old_media)
+
+    async def _remove_local_media_from_disk(
+        self, media_ids: List[str]
+    ) -> Tuple[List[str], int]:
+        """
+        Delete local or remote media from this server. Removes media files,
+        any thumbnails and cached URLs.
+
+        Args:
+            media_ids: List of media_id to delete
+        Returns:
+            A tuple of (list of deleted media IDs, total deleted media IDs).
+        """
+        removed_media = []
+        for media_id in media_ids:
+            logger.info("Deleting media with ID '%s'", media_id)
+            full_path = self.filepaths.local_media_filepath(media_id)
+            try:
+                os.remove(full_path)
+            except OSError as e:
+                logger.warning("Failed to remove file: %r: %s", full_path, e)
+                if e.errno == errno.ENOENT:
+                    pass
+                else:
+                    continue
+
+            thumbnail_dir = self.filepaths.local_media_thumbnail_dir(media_id)
+            shutil.rmtree(thumbnail_dir, ignore_errors=True)
+
+            await self.store.delete_remote_media(self.server_name, media_id)
+
+            await self.store.delete_url_cache((media_id,))
+            await self.store.delete_url_cache_media((media_id,))
+
+            removed_media.append(media_id)
+
+        return removed_media, len(removed_media)
+
 
 class MediaRepositoryResource(Resource):
     """File uploading and downloading.

synapse/storage/databases/main/media_repository.py

@@ -93,6 +93,7 @@ class MediaRepositoryStore(MediaRepositoryBackgroundUpdateStore):
 
     def __init__(self, database: DatabasePool, db_conn, hs):
         super().__init__(database, db_conn, hs)
+        self.server_name = hs.hostname
 
     async def get_local_media(self, media_id: str) -> Optional[Dict[str, Any]]:
         """Get the metadata for a local piece of media
@@ -115,6 +116,58 @@ class MediaRepositoryStore(MediaRepositoryBackgroundUpdateStore):
             desc="get_local_media",
         )
 
+    async def get_local_media_before(
+        self, before_ts: int, size_gt: int, keep_profiles: bool,
+    ) -> Optional[List[str]]:
+
+        # to find files that have never been accessed (last_access_ts IS NULL)
+        # compare with `created_ts`
+        sql = """
+            SELECT media_id
+            FROM local_media_repository AS lmr
+            WHERE
+                ( last_access_ts < ?
+                OR ( created_ts < ? AND last_access_ts IS NULL ) )
+                AND media_length > ?
+        """
+
+        if keep_profiles:
+            sql_keep = """
+                AND (
+                    NOT EXISTS
+                        (SELECT 1
+                         FROM profiles
+                         WHERE profiles.avatar_url = '{media_prefix}' || lmr.media_id)
+                    AND NOT EXISTS
+                        (SELECT 1
+                         FROM groups
+                         WHERE groups.avatar_url = '{media_prefix}' || lmr.media_id)
+                    AND NOT EXISTS
+                        (SELECT 1
+                         FROM room_memberships
+                         WHERE room_memberships.avatar_url = '{media_prefix}' || lmr.media_id)
+                    AND NOT EXISTS
+                        (SELECT 1
+                         FROM user_directory
+                         WHERE user_directory.avatar_url = '{media_prefix}' || lmr.media_id)
+                    AND NOT EXISTS
+                        (SELECT 1
+                         FROM room_stats_state
+                         WHERE room_stats_state.avatar = '{media_prefix}' || lmr.media_id)
+                )
+            """.format(
+                media_prefix="mxc://%s/" % (self.server_name,),
+            )
+            sql += sql_keep
+
+        def _get_local_media_before_txn(txn):
+            txn.execute(sql, (before_ts, before_ts, size_gt))
+            return [row[0] for row in txn]
+
+        return await self.db_pool.runInteraction(
+            "get_local_media_before", _get_local_media_before_txn
+        )
+
     async def store_local_media(
         self,
         media_id,