What changes have been made to the CSV import process recent

What changes have been made to the import process recently? I ran an import update with a csv file using a unique field. I noticed that the number of updates did not match the number of rows in the file, and did not identify the rows that failed ["Cannot find a matching value...]. It appears that if there is a comma at the end of a line, it reads the next line as a continuation, not a new row. I checked a similar file from last month for the same zone and this did not happen, so I can only assume that this a change in the code.