some smb servers are very insistent in reporting that the root of the share is readonly, even if it isn't.
This works around the problem by adding a hidden option to overwrite the permissions of the root of the share.
This can be enabled using
```bash
occ files_external:config <mount id> root_force_writable true
```
where you can find your mount id using
```bash
occ files_external:list
```
Signed-off-by: Robin Appelman <robin@icewind.nl>
This makes sure that a possible ForbiddenException is properly passed
through the storage as a ForbiddenException and can be catched when
trying to fetch the quota info of a parent folder
Signed-off-by: Julius Härtl <jus@bitgrid.net>
It seems that in some recent upgrade here. Not always the entries we
expect are returned. So we should first check if they exist. As to not
spam the log.
Signed-off-by: Roeland Jago Douma <roeland@famdouma.nl>
if a directory is empty, only the 'marker' object `/` exists.
since not all s3 implementations return just the prefix when listing objects by prefix,
when listing objects by the folder prefix, nothing will be returned if the directory is empty.
by not including the trailing slash in the prefix, the folder marked will always be returned if it exists
Signed-off-by: Robin Appelman <robin@icewind.nl>
Since we try to do range requests this will fail hard.
However since empty files are not that interesting to read anyways we
just read from an emptry memory stream.
Signed-off-by: Roeland Jago Douma <roeland@famdouma.nl>
the php ftp streamwrapper doesn't handle hashes correctly and will break when it tries to enter a path containing a hash.
By filtering out paths containing a hash we can at least stop the external storage from breaking completely
Signed-off-by: Robin Appelman <robin@icewind.nl>
since updating the mtime requires doing a copy this can have a significant overhead and it messes with versioning
Signed-off-by: Robin Appelman <robin@icewind.nl>
If a file or folder in a directory doesn't have read permissions they
will not be shown
Note that enabling this option incurs a performance penalty additional
requests need to be made to get all the acl.
Additionally the acl resolving logic is fairly primitive at the moment
and might not work correctly in all setups (it should error to showing
the entry)
Signed-off-by: Robin Appelman <robin@icewind.nl>
Currently you need to use `opendir` and then call `getMetadata` for
every file, which adds overhead because most storage backends already
get the metadata when doing the `opendir`.
While storagebackends can (and do) use caching to relief this problem,
this adds cache invalidation dificulties and only a limited number of
items are generally cached (to prevent memory usage exploding when
scanning large storages)
With this new methods storage backends can use the child metadata they
got from listing the folder to return metadata without having to keep
seperate caches.
Signed-off-by: Robin Appelman <robin@icewind.nl>
Robin Appelman
Morris Jobke
Roeland Jago Douma
Julius Härtl
Joas Schilling
Vincent Petry
Roeland Jago Douma
Christoph Wurst
Daniel Kesselberg
Julien Lutran