For #14179
By default responses should have the strictest (and simplest) CSP
possible. Only template responses should require an actual CSP.
Signed-off-by: Roeland Jago Douma <roeland@famdouma.nl>
This for now uses the jsNonce. That way we can easily backport it.
For 17 I will fix it properly.
Signed-off-by: Roeland Jago Douma <roeland@famdouma.nl>
This can be used by pages that do not have the full Nextcloud UI.
So notifications etc do not load there.
Signed-off-by: Roeland Jago Douma <roeland@famdouma.nl>
Fixes#13662
This will fire of an event after a Template Response has been returned.
There is an event for the generic loading and one when logged in. So
apps can chose to load only on loged in pages.
This is a more generic approach than the files app event. As some things
we might want to load on other pages as well besides the files app.
Signed-off-by: Roeland Jago Douma <roeland@famdouma.nl>
When id column has no autoincrement flag query for lastInsertId fails
on postgres because no value has been generated. Call lastInsertId only
if id is null.
Signed-off-by: Daniel Kesselberg <mail@danielkesselberg.de>
This way code using the DB mappers can have try catch blocks on this
type of exceptions if they do not care if there was non or to many.
Signed-off-by: Roeland Jago Douma <roeland@famdouma.nl>
This allows elegant upserts where the entity ID is provided (e.g. by an
external system) and when that data is fed into our database multiple
times.
Signed-off-by: Christoph Wurst <christoph@winzerhof-wurst.at>
Fixes#11035
Since the child-src directive is deprecated (we should kill it at some
point) we need to have the proper worker-src available
Signed-off-by: Roeland Jago Douma <roeland@famdouma.nl>
Now this is in core so the basics (that 99% of the app will want to
use) looks always the same.
Signed-off-by: Roeland Jago Douma <roeland@famdouma.nl>
* Deprecate our default CSP
* Add strict CSP that is always our strictest setting
* Add strict eval CSP (disable unsafe-eval)
* Add strict inline CSP (disables inline styles)
This is just to move forward and have a incremental improvement of our
CSP
Signed-off-by: Roeland Jago Douma <roeland@famdouma.nl>
This avoids having to do it at all the places we want cached responses.
We can't inject the ITimeFactor without breaking public API.
However we can perfectly overwrite the service (resulting in the same
testable effect).
Signed-off-by: Roeland Jago Douma <roeland@famdouma.nl>
Roeland Jago Douma
Marius David Wieschollek
Joas Schilling
Daniel Kesselberg
Morris Jobke
Jakob Sack
Christoph Wurst
Julius Härtl
Roeland Jago Douma