which encoding should i used with fopen in matlab

the instruction :

fullpathr = strcat (exp_subfolder,tmf_file)
fidr = fopen(fullpathr,r,'ieee-le','UCS-2')

i want to open and read a file in matlab but always it return -1 and warning "the encoding UTF-16 is not supported"

please can any one help me in this

[Walter Roberson]

This is because MATLAB only officially supports UTF-8 encoding with 'fopen'.

Not exactly. MATLAB supports a long list of encodings, mostly ISO. However, it does not officially support UTF-16

Answer 1

[Saurabh]

Hi @rakia belgacem,

I understand you are encountering an issue opening a UCS‑2 (or UTF‑16) encoded file in MATLAB. This is because MATLAB only officially supports UTF-8 encoding with 'fopen'. While encodings like UCS-2, UTF-16LE are not officially supported.

To workaround this limitation:

Read raw bytes and Decode explicitly.

fid = fopen(fullpathr, 'rb');
fread(fid, 2, '*uint8');  % Skip BOM
bytes = fread(fid, 'uint8=>uint8')';
fclose(fid);
str = native2unicode(bytes, 'UTF-16LE');
data = textscan(str, '%s %f %f', 'Delimiter', ',', 'HeaderLines', 1);

This method reads raw bytes, manually decodes them using native2unicode, and then parses the resulting string.

• 'native2unicode' converts byte arrays to MATLAB character arrays based on the specified encoding (UTF-16LE, UTF-8, etc.)
• This approach handles files with 16-bit encoding reliably, avoiding issues from fopen’s limited encoding support .

To know more about 'native2unicode' refer to the following official MathWorks documentation:

https://www.mathworks.com/help/matlab/ref/native2unicode.html

I hope this helps in resolving your query.