textscan scuppered by '\n' mid entry

I have a csv file that annoyingly has some string entries that have newlines within the entry. Therefore when textscan comes across this entry it thinks it has reached the end of the row and tries to start a new one, where it halts scanning.

I tried to skip this field using %*s which doesn't work. Is there a way to ignore the '\n' signifier in between delimiters?

Answer 1

[Image Analyst]

You can pre-process the file. Pseudocode:

open file
while not EOF
  read a line with fgetl
  count commas
  if commas not the expected amount
     read next line
     stitch both lines together
     count commas again.
    if commas not the expected amount, it's an error
        Alert user via warndlg
    end
  end
  write line to output file
end

[harry]

Ok cool, that sounds good. So say instead of new lines (\n) commas were mid entry (as this damned file has), one wouldn't be able to stitch the next line as with the new line case, we would actually have to remove it. If this comma doesn't occur in a consistent position, how do you think we solve this?

Thanks!

[Image Analyst]

I didn't understand. Are you saying that you have two commas in a row or missing or extra commas? Please give an example of that type of line.