While looking at the source code for recordings, I noticed that key ids(which are the recording names) are being placed directly on endpoint URLs without being escaped. This can cause errors when, for example, the recording name is itself a file path: this would lead to an HTTP call to /recordings/live/something/else/name, which is obviously wrong.
While looking at the source code for recordings, I noticed that key ids(which are the recording names) are being placed directly on endpoint URLs without being escaped. This can cause errors when, for example, the recording name is itself a file path: this would lead to an HTTP call to
/recordings/live/something/else/name
, which is obviously wrong.