The session configuration process consists of four steps: beginConfig, addInput, addOutput, and commitConfig. The calling sequence of these steps cannot be changed, and all four steps must be configured successfully for the camera to work properly. Possible causes include the following:
beginConfig fails.
addInput fails.
addOutput fails.
commitConfig fails.
Problem Analysis Steps
Step 1: beginConfig may fail for the following reasons: calling this API while the session is in commitConfig; a null pointer exception when the session object does not exist.
Step 2: addInput can only be called after the session calls beginConfig. It cannot be modified after commitConfig. Each session can be configured with only one input, and a null check is performed.
Step 3: addOutput can only be configured when the session is in the beginConfig state, after addInput, and before commitConfig. It checks the current profile to verify whether it matches the input. You can use ValidateOutputProfile to view the corresponding output profile content. A null check is performed.
Step 4: After the preceding configurations are confirmed, call commitConfig to submit the configuration to the underlying layer and connect the channels of the input stream and output stream, so that images can be returned to the app through the channels.
Related log search keywords: beginConfig|addInput|addOutput|commitConfig
If beginConfig fails, check whether the session has been released or whether the session is currently performing a commitConfig operation.
Before calling the addInput API, you can call the canAddInput API to verify whether the current session can add an input.
Before calling the addOutput API, you can call the canAddOutput API to verify whether the current session can add an output.
When calling commitConfig, ensure that none of the preceding steps have failed and that the session has not been released or used for other operations during this process.
For normal session configuration, see Photo Capture (ArkTS).