This will prevent detecting remote changes done in the data folder /
root storage by default. In the rare cases where the data folder is
shared with other apps/users outside ownCloud and change detection is
needed, the admin will have to set the option explicitly from now on.
Note that this doesn't affect external storages which have their own
setting in the mount options.
* Log condition for log level increase based on conditions. Once one of these
conditions is met, the required log level is set to debug. This allows to
debug specific requests, users or apps
* Supported conditions (setting `log_condition` in `config.php`):
- `shared_secret`: if a request parameter with the name `log_secret` is set to
this value the condition is met
- `users`: if the current request is done by one of the specified users,
this condition is met
- `apps`: if the log message is invoked by one of the specified apps,
this condition is met
* fix unit test and add app log condition test
This Pull Request introduces a SabreDAV plugin that will block all older clients than 1.6.1 to connect and sync with the ownCloud instance.
This has multiple reasons:
1. Old ownCloud client versions before 1.6.0 are not properly working with sticky cookies for load balancers and thus generating sessions en masse
2. Old ownCloud client versions tend to be horrible buggy
In some cases we had in 80minutes about 10'000 sessions created by a single user. While this change set does not really "fix" the problem as 3rdparty legacy clients are affected as well, it is a good work-around and hopefully should force users to update their client
The first time we're asked to generate a preview we'll generate one of the maximum dimension indicated in the configuration and all future resizing requests will be done on that preview in order to not waste time converting the same file over and over.
One of the fixes required for #12465
Caches divided up into two groups: distributed and local. 'Low latency' is an
alias for local caches, while the standard `create()` call tries to get
distributed caches first, then local caches.
Memcache backend is set in `config.php`, with the keys `memcache.local` and
`memcache.distributed`. If not set, `memcache.distributed` defaults to the value
of `memcache.local`.
Doing this in the PHP code is not the right approach for multiple reasons:
1. A bug in the PHP code prevents them from being added to the response.
2. They are only added when something is served via PHP and not in other cases (that makes for example the newest IE UXSS which is not yet patched by Microsoft exploitable on ownCloud)
3. Some headers such as the Strict-Transport-Security might require custom modifications by administrators. This was not possible before and lead to buggy situations.
This pull request moves those headers out of the PHP code and adds a security check to the admin settings performed via JS.
This change allows AppFramework applications to specify a custom CSP header for example when the default policy is too strict. Furthermore this allows us to partially migrate away from CSS and allowed eval() in our JavaScript components.
Legacy ownCloud components will still use the previous policy. Application developers can use this as following in their controllers:
```php
$response = new TemplateResponse('activity', 'list', []);
$cspHelper = new ContentSecurityPolicyHelper();
$cspHelper->addAllowedScriptDomain('www.owncloud.org');
$response->addHeader('Content-Security-Policy', $cspHelper->getPolicy());
return $response;
```
Fixes https://github.com/owncloud/core/issues/11857 which is a pre-requisite for https://github.com/owncloud/core/issues/13458 and https://github.com/owncloud/core/issues/11925
This allows the directory where CSS/JS asset collections are
written to be changed, in case SERVERROOT is not writeable. Note
it does *not* allow the expected URL to be changed: whatever
directory is used, the server must be configured to serve it
at WEBROOT/assets. It may be possible to add another config
parameter to allow the admin to specify a custom asset URL,
but I thought I'd keep the first implementation simple.
LDAP User Cleanup
background job for user clean up
adjust user backend for clean up
register background job
remove dead code
dependency injection
make Helper non-static for proper testing
check whether it is OK to run clean up job. Do not forget to pass arguments.
use correct method to get the config from server
methods can be private, proper indirect testing is given
no automatic user deletion
make limit readable for test purposes
make method less complex
add first tests
let preferences accept limit and offset for getUsersForValue
DI via constructor does not work for background jobs
after detecting, now we have retrieving deleted users and their details
we need this method to be public for now
finalize export method, add missing getter
clean up namespaces and get rid of unnecessary files
helper is not static anymore
cleanup according to scrutinizer
add cli tool to show deleted users
uses are necessary after recent namespace change
also remove user from mappings table on deletion
add occ command to delete users
fix use statement
improve output
big fixes / improvements
PHP doc
return true in userExists early for cleaning up deleted users
bump version
control state and interval with one config.php setting, now ldapUserCleanupInterval. 0 will disable it. enabled by default.
improve doc
rename cli method to be consistent with others
introduce ldapUserCleanupInterval in sample config
don't show last login as unix epoche start when no login happend
less log output
consistent namespace for OfflineUser
rename GarbageCollector to DeletedUsersIndex and move it to user subdir
fix unit tests
add tests for deleteUser
more test adjustements
Conflicts:
apps/user_ldap/ajax/clearMappings.php
apps/user_ldap/appinfo/app.php
apps/user_ldap/lib/access.php
apps/user_ldap/lib/helper.php
apps/user_ldap/tests/helper.php
core/register_command.php
lib/private/preferences.php
lib/private/user.php
add ldap:check-user to check user existance on the fly
Conflicts:
apps/user_ldap/lib/helper.php
forgotten file
PHPdoc fixes, no code change
and don't forget to adjust tests
Workaround required for IIS setups running ownCloud to prevent dataloss.
Long-term solution would be to move some configuration settings to the database
This commit will make the supported DBs for installation configurable within config.php. By default the following databases are tested: "sqlite", "mysql", "pgsql". The reason behind this is that there might be instances where we want to prevent SQLite to be used by mistake.
To test this play around with the new configuration parameter "supportedDatabases".
Correctly use overwritemailurl value when generating absolute urls in CLI
Fix#11500
Rename the config to *cli
Add overwrite.cli.url to the sample config
Revert separator fix, fixes unit test
* use ' instead of " for config option
* place default parameters where useful into sample
* use proper comment block
* limit line size to 80 characters
Remove the second comment which could be confused with suggesting an invalid value: `// 100 MiB`.
To keep the easily readable example value, it has been moved into the comment header for the config item.
This adds some security utilities to core including:
- A library for basic crypto operations (e.g. to encrypt passwords)
- A better library for cryptographic actions which allows you to specify the charset
- A library for secure string comparisions
Remove .htaccess
Remove .htaccess
Fix typo
Add public API
Use timing constant comparision
Remove CBC constant
Adjust code
Remove confusing $this
Hopefully this will stop people from copying the sample config. I'm so annoyed by all those wrong bug reports...
Add some explanation about this switch
Move check to init
* Use "filesystem_cache_readonly" config setting, update comment in config.sample
* Use $this->cacheActive to cache config setting
* Add public Scanner::setCacheActive() to set $cacheActive programmatically
Some people believe that they should copy the sample config to the "real" config. I noticed this several times in IRC and on the bugtracker.
I guess this warning should be enough to avoid this in the future.
Users often ask in IRC or the forum how to add another domain.
Hopefully they will be able to find it out on their own if we have an example with two domains.
When using an external cache folder, it is automatically mounted in
FileSystem::initFileSystem so that any app can use it transparently
by creating a view on the "/$user/cache" directory.