topic Re: Redirect SHELL GETCHAR to custom implementation missing? in MCUXpresso SDK

Redirect SHELL GETCHAR to custom implementation missing?

robertoquilez — Fri, 11 Jan 2019 09:08:14 GMT

In the previous version, a nice feature that now is missing was to redirect the GETCHAR (SHELLReveiveDataCallBack) to a custom implementation. For instance, this enabled to implement a GETCHAR version with a SHELL freertos task in SUSPENDED STATE while waiting for a character.

Is there a way to redirect the GETCHAR to a custom application as before with the new SDK shell version?

By the way, the SHELL in SDK 2.5.0 documentation (MCUXpresso SDK API Reference Manual_MK66F18.pdf) does not match the source code.

Documentation refers to :

SHELL_Init(&user_context, SHELL_SendDataCallback, SHELL_ReceiveDataCallback, "SHELL>> ");

When the SDK source code:

shell_status_t SHELL_Init(shell_handle_t shellHandle, serial_handle_t serialHandle, char *prompt);

Re: Redirect SHELL GETCHAR to custom implementation missing?

jorge_a_vazquez — Tue, 15 Jan 2019 21:43:49 GMT

Hi Roberto Quilez

The implementation in the SDK 2.4 version uses callbacks for the "send" and the "receive" functions. Now in SDK 2.5, those parts are managed in the serial_manager component, so if you want to change this, you can do it in the SerialManager_StartReading function in the serial_manager.c/.h

Regarding the documentation, thanks for pointing out this, I will report this for future releases.

Hope this information helps
Have a great day,
Jorge Alcala

-----------------------------------------------------------------------------------------------------------------------
Note: If this post answers your question, please click the Correct Answer button. Thank you!
-----------------------------------------------------------------------------------------------------------------------

Re: Redirect SHELL GETCHAR to custom implementation missing?

robertoquilez — Wed, 16 Jan 2019 08:38:13 GMT

Thank you very much, Joge, for your rapid response.

From my personal use case and personal point of view, I think that it is a pity to have to modify the source code of one of the SDK components. Every time I update the SDK, I might loose my modifications and I will have to take care of including them manually.

I think the previous implementation was more convenient as it was part of the shell initialization settings. Actually I have the impression that it is much more complex (and limited) that the previous one.