[bugfix] better identify raw file when exporting file

[Edouard2laire]

Hello,

This fix a bug where some raw file would not be identified as raw file during export.

[rcassani]

Hi @Edouard2laire, how did you get to have a file with such a name?
Raw files should have _0raw in their names

[Edouard2laire]

it was created by #717

i copied what we did for the synchronisation to get filename (https://github.com/brainstorm-tools/brainstorm3/blob/master/toolbox/process/functions/process_sync_recordings.m#L255 ) :

OutputFile = bst_process('GetNewFilename', bst_fileparts(sNewStudy.FileName), 'data_raw_combned');

It is recognize everywhere else in brainstorm as raw file despite not having _0raw

[rcassani]

i copied what we did for the synchronisation to get filename (https://github.com/brainstorm-tools/brainstorm3/blob/master/toolbox/process/functions/process_sync_recordings.m#L255 ) :

Yeah, that's not right. It should have been _0raw

It is recognize everywhere else in brainstorm as raw file despite not having _0raw

It does work for most of the cases as different checks for 'raw' type just find the raw string, or check if the loaded F is a struct. However, it may have unexpected behaviours, e.g., setting a bad channel for a _raw file will not change the ChannelFlag inside F:

brainstorm3/toolbox/sensors/panel_channel_editor.m

Line 722 in ec632b9

isRaw = (length(DataFile) > 9) && ~isempty(strfind(DataFile, 'data_0raw'));

, and in other places that do check for _0raw

I'll push the correction on process_sync_recordings.m directly on the master

[Edouard2laire]

Is there a way I can fix the file already created without having to recompute ?

[rcassani]

Apologies for the inconveniences.
With Brainstorm close, you could renaming the file to add the zero and reloading the Protocol