How to move data from cells in a cell in a destinct column (importing CoDeSys variables)

Hi All

I have a challenge of reading PLC (CoDeSys IEC 61131) variables from a text file and putting them into an array(:,4). Here is an example of the formatting:

VAR_GLOBAL PERSISTENT
// Default Interfaces
	g_stCAL										: ST_GLOBAL_CAL;													// Calibration interface
	g_stCFG										: ST_GLOBAL_CFG := 10;	//test	 for dimensions											// Configuration interface
	g_stDAT										: ST_GLOBAL_DAT;													// Statistical data interface
	g_stOPT										: ST_GLOBAL_OPT;													// Option interface
	g_stSET										: ST_GLOBAL_SET;													// Settings interface

	g_nLifetimePlcRun							: ULINT;															// [ms] Total time the PLC is running during it's life time
END_VAR

The variables are grouped within VAR_xxx and END_VAR. The formating is like <var_name>:<data_type> := <initial_value>; // <comment> Within the file there are multiple groups of variable types. So VAR_GLOBAL PERSISTENT / VAR_GLOBAL CONSTANT / VAR_GLOBAL but in my case they can all be added to the same array. The following I already have:

01:	    fileID = fopen(iFile,'r');
02:	    TestStruct = fscanf(fileID,'%c');
03:	    [VarIdxStart,VarIdxStop] = regexp(TestStruct,'VAR_.+?\r\n')			% Find locations of 'VAR_xxx'
04:	    [EndIdxStart,EndIdxStop] = regexp(TestStruct,'END_VAR')				% Find locations of 'END_VAR'

05: 	Result = cell(1,4);         % Initialise 'name' 'datatype' 'initial value' 'comment'
06:	    for idx = 1 : max(size(VarIdxStart))
07:		    Variables = splitlines(extractBetween(TestStruct,VarIdxStop(idx),(EndIdxStart(idx)-1)));			% Get variables and put into list
08:		    Variables = Variables(~cellfun('isempty',Variables));												% Remove empty cells
09:		    Variables = Variables(cellfun(@(tmp) contains(tmp,{':' ';'}),Variables));							% Remove cell not containing variable declarations
10:     initIdx = cellfun(@(tmp) contains(tmp,{':='}),Variables);           % Indication that cell contains initialisation values
11:     Variables = cellfun(@(tmp) split(tmp,{':=' ':' ';'})', Variables, 'UniformOutput', false);			% Split on init value, colon or semicolon
12:		    Variables = cat(1,Variables{:});																	% Move data into same level
13:		    Result = cat(1,Result,Variables);			% Add found variables to 

My problem lies in the 2nd variable 'g_stCFG'. For testing I have added an initialisation value (in this case not appropriate, but it is for testing). When the initialisation value is NOT present then the code produces the desired information, but in an array(:,3).

So I added line '10:' to have an index to the and added the ':=' condition to the split of line '11:' so the result looks like:

Variables(6x1)cell
'1x3 cell'                => g_stCAL                ST_GLOBAL_CAL       // Calibration interface
'1x4 cell'                => g_stCFG                ST_GLOBAL_CFG       10      //test	 for dimensions
'1x3 cell'                => g_stDAT		        ST_GLOBAL_DAT		// Statistical data interface
'1x3 cell'                ...
'1x3 cell'                ...
'1x3 cell'                => g_nLifetimePlcRun	    ULINT	            // [ms] Total time the PLC is running during it's life time

This causes that line '12:' results in a crash as the size of the cells are not equal. So how to 'extend' all entries that have a '1x3 cell' and preferely that the comment moves to the 4th row. Or does anybody have a quicker way of importing the data? Thanks for any help.

Ludo

Answer 1

[Mathieu NOE]

hello

maybe this ?

where the lines in your file do not have initial values , I inserted a zero in 3rd position ; this is done in these two extra lines of code :

         %%%% my mod %%%
         mm = max(cellfun(@numel, Variables));
        Variables = cellfun(@(tmp) [tmp(:,1:2) zeros(1, mm - numel(tmp)) tmp(:,3:end)], Variables, 'un', 0); % introduce a zero in 3rd position if needed
        %%%%%%%%%%%%%%%

full code

iFile = 'test.txt';   % my test file 

fileID = fopen(iFile,'r');
TestStruct = fscanf(fileID,'%c');
[VarIdxStart,VarIdxStop] = regexp(TestStruct,'VAR_.+?\r\n');			% Find locations of 'VAR_xxx'
[EndIdxStart,EndIdxStop] = regexp(TestStruct,'END_VAR');				% Find locations of 'END_VAR'

% Result = cell(1,4);         % Initialise 'name' 'datatype' 'initial value' 'comment'
Result = {'name','datatype','initial value','comment'};         % Initialise 'name' 'datatype' 'initial value' 'comment'


    for idx = 1 : max(size(VarIdxStart))
	    Variables = splitlines(extractBetween(TestStruct,VarIdxStop(idx),(EndIdxStart(idx)-1)));			% Get variables and put into list
	    Variables = Variables(~cellfun('isempty',Variables));												% Remove empty cells
	    Variables = Variables(cellfun(@(tmp) contains(tmp,{':' ';'}),Variables));							% Remove cell not containing variable declarations
         initIdx = cellfun(@(tmp) contains(tmp,{':='}),Variables);           % Indication that cell contains initialisation values
         Variables = cellfun(@(tmp) split(tmp,{':=' ':' ';'})', Variables,'UniformOutput' , false);			% Split on init value, colon or semicolon
	    
         %%%% my mod %%%
         mm = max(cellfun(@numel, Variables));
        Variables = cellfun(@(tmp) [tmp(:,1:2) zeros(1, mm - numel(tmp)) tmp(:,3:end)], Variables, 'un', 0); % introduce a zero in 3rd position if needed
        %%%%%%%%%%%%%%%
         
         Variables = cat(1,Variables{:});																	% Move data into same level
	    Result = cat(1,Result,Variables);			% Add found variables to 
    end

my result so far :

[Ludo Houben]

Hi Mathieu

This works great, thanks :-)

[Mathieu NOE]

Hi Ludo

glad I could help !

minor thing, but I just figured out that maybe you wanted my inserted zeros as string and not numerical values

in that case simply replace

Variables = cellfun(@(tmp) [tmp(:,1:2) zeros(1, mm - numel(tmp)) tmp(:,3:end)], Variables, 'un', 0); % introduce a zero in 3rd position if needed

with

Variables = cellfun(@(tmp) [tmp(:,1:2) num2str(zeros(1, mm - numel(tmp))) tmp(:,3:end)], Variables, 'un', 0); % introduce a zero in 3rd position if needed

[Ludo Houben]

Yes, I wanted that but I already solved the 'STRING' part myself. I solved it with 'repmat'

cellfun(@(tmp) [tmp(:,1:2) repmat("",1, mm - numel(tmp)) tmp(:,3:end)], Variables, 'un', 0);

You can add any data in the field to add in that way. In my case an empty STRING, but whatever the user wants.

Cheers

Ludo

[Mathieu NOE]

Hi Ludo

ok , sounds good

have a nice day