How do I ignore delimiter inside quotation marks using regexp?

Hi everyone,

I'm writing a script to import data. The size (coloumns) of the data can vary. The main delimiter is a semicolon.

My data looks like the following:

Header ; Value ; Data

Example1; 12 ; "air"

Example2; 24 ; "water"

For this usecase I'm using

filebyfield = regexp(filebyline, ';', 'split'); %filebyline- textfile including my data

Unfortunately the data I try to import isn't uniform and could look like the following:

Header ; Value ; Data

Example1; 12 ; "air;fire"

Example2; 24 ; "water"

Is there any way to ignore the delimiter (";") if its inside of quotation marks? I thought about textscan, but I don't have the same number of coloumns every time and therefore cant specify the data format.

Any help would be appreciated! Best regards

Answer 1

[Walter Roberson]

sometimes the easiest way is to detect the quoted delimiter and replace them with some special character such as ¡ and then do the split and then fix-up the results.

[Kilian Brandt]

hi walter, ok thanks for the hint, I tried it but it turned out that my whole approach was too complicated for the task.

best regards!

[Ameer Hamza]

I think Walter's approch is quite simple. For example

filebyfield = {'Header ; Value ; Data';
     'Example1; 12 ; "Air;fire"';
     'Example2; 24 ; "water"'};
 
filebyfield = regexprep(filebyfield, '("\w*);(\w*")', '$1@$2');
filebyfield = regexp(filebyfield, ';', 'split');
filebyfield = cellfun(@(x) {strrep(x, '@', ';')}, filebyfield)

[Kilian Brandt]

Thanks alot! It works perfectly.

I have an additional question about the first line:

filebyfield = regexprep(filebyfield, '("\w*);(\w*")', '$1@$2');

Could you please tell me if my assumptions about the terms

'("\w*);(\w*")' % finds the quoted ";" | "/w*" means characters with undefined length 

and

'$1@$2' % "$1" stands for the first "("\w*)" and $2 for the second "("\w*)"

are correct?

[Ameer Hamza]

Yes, all the points are correct.