readtable cannot handle double quotation marks very well

I have CSV files saved with LibreOffice with text flanked by double quotation marks (Format quoted field as text).

When I tried to read one of such CSV with two rows with readtable,

T0 =  readtable('file1.csv',...
    'Encoding','UTF-8','delimiter',',','ReadVariableNames',true);

readtable failed to read the first row,

Then I used this command and it can read both rows.

opts1 = delimitedTextImportOptions('Encoding','UTF-8','Delimiter',',','DataLines',[2 Inf],'VariableNamesLine',1);
T1 =  readtable('file1.csv',opts1);

However, the content of table wasn't great:

ans = 2×1 cell

'"optotagging"'

'"behaviour"'

The double quotation marks remained in some columns.

setvaropts' option 'QuoteRule','remove' appeared to be promissing, but I could not get it work.

setvaropts(opts1,'QuoteRule','remove')

How do I nicely remove double quotation marks in CSVs?

Answer 1

[Kouichi C. Nakamura]

I asked this to Mathworks and their answer was helpful:

opts = detectImportOptions('file1.csv','NumHeaderLines',0,'Delimiter',',') %will almost work for this case, but it detects the first line as a "meta-data" line because it is all string/blank
opts.DataLines = [2,inf] %will work around that issue
T2 = readtable('file1.csv',opts);

With this code, I can read both rows and remove double quotation marks nicely.

According to Mathworks:

> The solution shared, is very specific to your workflow and is an undocumented method which might change without notice.