Reading greek letters!

I need to read text from an EXCEL file that will include greek letters in some cells. While I can transform the letters (e.g., spelling out "alpha" in place of the actual letter), I need to preserve the content. Currently, when I read the file using readcell, I get the greek letters in the text, but as soon as I try to do anything with them they get substituted with question marks. This is problematic, as I need to make decisions based on the text, and can't do that if every theta, eta, and nu is just going to become identical ?'s.

So, any ideas?

EDIT: For clarification on "...but as soon as I try to do anything..."

Below, the word tau, by itself, is the greek letter itself

%%%%%%%%%%%%%%%%%%%%%%

>>Text = readcell(File,'Sheet',Sheet);

>>VN = char(Text(2,1));

VN =

tau

>>switch VN

case tau

A = 'tau';

otherwise

A = 'ERROR';

end

>>A

A =

'ERROR';

%%%%%%%%%%%%%%%%%%%%%%%

Please note that when I call VN, I see the letter tau on the screen. When I save it in an .mat file and reopen, VN = '?'.

In essense, I need a way to do the reverse of what char(945) would do - where char(945) takes a MATLAB-manipulatable value and produces a greek letter, I need something that will take a greek letter and produce a MATLAB manipulative value (like Unicode or '\alpha').

[Adam Danz]

"..but as soon as I try to do anything...

???

[Adam Danz]

If you're just referencing single greek letters, you could use the ASCII values for those characters.

For example, the ASCII value for tau is

value = double('τ')  % 964, lower case
value = double(upper('τ'))  % 932, upper case

So when you read in the letters, you can instantly convert them to their ASCII values.

VN = double(Text{2,1}); 

and then in your switch statement,

switch VN
    case [932, 964] % for upper and lower cases
        % code
    otherwise
        error()
end

I'm writing this as a comment rather than an answer because I'm unsure of how you're actually using the text and the switch case.

[Jeremy Hughes]

To be pedantic, ASCII is only the [0 127] range of characters. Beyond 127, everything is "extended ascii" which means nothing by itself. You need to know what encoding the characters are in. <end pendantry>

In this case, the TAU character is the UTF-16 representation (which is how MATLAB stores characters). The "?" showing up in display can mean one of two things:

The character isn't recognized from the encoding it's stored in the file as, or the FONT you're using doesn't have the character available.

To tell which, use double(TAU_CHAR) should show you the UTF-16 value(s) for the character:

If it's

ans = 26, then MATLAB has replaced the character, as it didn't know what to do with it. Otherwise, if it there's a number other than 26, the data is there, but the FONT doesn't have an entry for that UTF-16 character.

[Stephen Hall]

Adam, your comment is sufficient for an answer - I did not know that double would give the code for the letter. It does, and my problem is now resolved.

Thanks, both!

[Adam Danz]

@Jeremy Hughes, thanks for the clarification; pedantry welcomed!

@Stephen Hall, I'll copy an improved version of my comment to the answers section so your question appears as answered.

Answer 1

[Adam Danz]

[continued from comment section under the answer]

If you're just referencing single greek letters, you could use the double() value for those characters.

For example, the double() value for tau is

value = double('τ')  % 964, lower case
value = double(upper('τ'))  % 932, upper case

So when you read in the letters, you can instantly convert them to their numeric values.

VN = double(Text{2,1}); 

and then in your switch statement,

switch VN
    case [932, 964] % for upper and lower cases
        % code
    otherwise
        error()
end