How to use textscan with empty char fields

How can I use textscan with files that include these kind of lines, with empty chars?

m1 = 'HELLO,42.53,A,B,1,C';  % Only this works fine
m2 = 'HELLO,42.53,A,,1,C';
m3 = 'HELLO,42.53,A,B,1,';
m4 = 'HELLO,42.53,,,1,C';

pattern = 'HELLO%f%c%c%f%c';
delimiter = ',';

[C1, pos1] = textscan(m1, pattern, 'Delimiter', delimiter)

C1 = 1×5 cell array
    {[42.5300]}    {'A'}    {'B'}    {[1]}    {'C'}

pos1 = 19

[C2, pos2] = textscan(m2, pattern, 'Delimiter', delimiter)

C2 = 1×5 cell array
    {[42.5300]}    {'A'}    {','}    {[1]}    {'C'}

pos2 = 18

[C3, pos3] = textscan(m3, pattern, 'Delimiter', delimiter)

C3 = 1×5 cell array
    {[42.5300]}    {'A'}    {'B'}    {[1]}    {0×1 char}

pos3 = 18

[C4, pos4] = textscan(m4, pattern, 'Delimiter', delimiter)

C4 = 1×5 cell array
    {[42.5300]}    {','}    {'1'}    {0×1 double}    {0×1 char}

pos4 = 16

[Adam Danz]

Since this doesn't answer the question, How can I use textscan ..., I'll leave a comment rather than an answer.

A workaround is to use strsplit and then some post-processing to convert chars to numbers.

m1 = 'HELLO,42.53,A,B,1,C';  
m2 = 'HELLO,42.53,A,,1,C';
m3 = 'HELLO,42.53,A,B,1,';
m4 = 'HELLO,42.53,,,1,C';

cleanStrFcn = @(s)[str2double(s{2}),s(3:end)];
s1 = cleanStrFcn(strsplit(m1,',','CollapseDelimiters',false))

s1 = 1×5 cell array
    {[42.5300]}    {'A'}    {'B'}    {'1'}    {'C'}

s2 = cleanStrFcn(strsplit(m2,',','CollapseDelimiters',false))

s2 = 1×5 cell array
    {[42.5300]}    {'A'}    {0×0 char}    {'1'}    {'C'}

s3 = cleanStrFcn(strsplit(m3,',','CollapseDelimiters',false))

s3 = 1×5 cell array
    {[42.5300]}    {'A'}    {'B'}    {'1'}    {0×0 char}

s4 = cleanStrFcn(strsplit(m4,',','CollapseDelimiters',false))

s4 = 1×5 cell array
    {[42.5300]}    {0×0 char}    {0×0 char}    {'1'}    {'C'}

Answer 1

[Jeremy Hughes]

The format specifier %c always reads one character, no matter what that character is. What you likely want it %s, which reads all the characters up to the next matching delimiter.

[Stephen23]

str = sprintf('%s\n', 'HELLO,42.53,A,B,1,C', 'HELLO,42.53,A,,1,C', 'HELLO,42.53,A,B,1,', 'HELLO,42.53,,,1,C')

str = 
    'HELLO,42.53,A,B,1,C
     HELLO,42.53,A,,1,C
     HELLO,42.53,A,B,1,
     HELLO,42.53,,,1,C
     '

out = textscan(str, 'HELLO,%f%s%s%f%s', 'Delimiter',',');
out{:}

ans = 4×1

42.5300 42.5300 42.5300 42.5300

ans = 4×1 cell array
    {'A'     }
    {'A'     }
    {'A'     }
    {0×0 char}

ans = 4×1 cell array
    {'B'     }
    {0×0 char}
    {'B'     }
    {0×0 char}

ans = 4×1

1 1 1 1

ans = 4×1 cell array
    {'C'     }
    {'C'     }
    {0×0 char}
    {'C'     }

[Joan Vazquez]

Thanks Jeremy and thanks Stephen for the examples