如何使用 textscan 读取这个 txt 文件 matlab/octave

Question

How to use textscan to reading this txt file matlab/octave

Time:
11:00
Day:
2019-11-05
Company:
Hyperdrones
Drones:
Jupiter, alvalade, 20, 2000, 500.0, 20.0, 2019-11-05, 10:15
Terra, ameixoeira, 15, 1500, 400.0, 20.0, 2019-11-05, 10:20
V125, ameixoeira, 20, 2000, 350.0, 20.0, 2019-11-05, 10:20
Saturno, lumiar, 10, 1000, 600.0, 20.0, 2019-11-05, 10:30
Neptuno, lumiar, 15, 1500, 600.0, 15.0, 2019-11-05, 10:30
Mercurio, alvalade, 25, 2500, 200.0, 20.0, 2019-11-05, 10:40
Marte, campogrande, 10, 1500, 100.0, 10.0, 2019-11-05, 10:50

Answer 1

Assuming you want to read the matrix portion:

C = textscan(fileID, '%s %s %d %d %f %f %{yyyy-dd-MM}D %{HH:mm}D','HeaderLines',7,'Delimiter,',')

The 7 assumes there aren't really blank lines in your file. If you have blank lines, adjust that to 14 or whatever is appropriate).

Answer 2

Regarding the Octave version specifically , I would recommend something like this:

pkg load io                          % required for `csv2cell` function
tmp = fileread('data');              % read file as string
tmp = strsplit( tmp, '\n' );         % split on line endings to create rows
tmp = tmp(1:6);                      % keep only rows 1 to 6
Headers = struct( tmp{:} );          % create struct from comma-separated-list
Headers.Drones = csv2cell('data', 7) % use csv2cell to read csv starting from row 7

Result:

Headers =
  scalar structure containing the fields:
    Time:: 1x5 sq_string
    Day:: 1x10 sq_string
    Company:: 1x11 sq_string
    Drones: 8x8 cell

Matlab does not come with an equivalent csv2cell function, but there are similar ones on matlab fileexchange; here's one that seems to have similar syntax as the octave version .

In general I'd use csv2cell for non-numeric csv data of this nature; it's much easier to work with than textscan. I'd only use textscan as a last resort for non-csv files with lines that are unusual but otherwise consistent in some way...

_{PS. If your csv file ends with an empty line, csv2cell might result in an extra 'empty' cell row. Remove this manually if you don't want it.}