If something happens that the system does not expect, or if the system fails to read the CSV file, a message like this will pop up. This often points to one or more problems, which are shortly explained below.



Incorrect delimiters

The system expects commas as the delimiter of a comma separated file. Microsoft Excel often saves CSV files with a semicolon delimiter. You can select the delimiter used, the default setting is ‘automatic’. When the delimiter is set to automatic, the system is usually smart enough to detect this common error and compensate, but it may be better to ‘manually’ change the csv file to use commas. You can do this in many ways, but the easiest two ways are to open and save the file in a program that does use commas correctly (e.g. SPSS), or open the file in a text editor and use the find and replace function to replace all semicolons with commas. If, for some reason, your csv file uses even more exotic delimiters, you can use a text editor to replace them with commas.

Incorrect attribute names or invalid case

The system uses the first line of a CSV file to assign values to the correct token attributes. Mandatory attributes are first name, last name and email, and their names are “firstname”, “lastname” and “email”. Make sure your CSV file uses the correct names, without spaces, and case sensitive. Below you'll see the correct way (in green) to enter the field names in row 1 and invalid formats (in red):

Incorrect encoding

By default, the system expects UTF-8 encoding. Different spreadsheet programs may use different encoding sets. For example, Microsoft Excel uses Windows-1252 on computers with a US English operating system installed. This is not a problem for most characters, but accented characters such as é, õ, ç and ä may be stored differently depending on the character set. When the system comes upon such a character it does not recognize, it will throw an error and stop importing.

To fix this, you have two options. Either delete the offending characters, or force the file to be saved using the correct encoding. Replacing accented characters with their ‘regular’, unaccented counterparts is an easy fix in some cases, but may not be desirable.

The alternative is to force encoding. This option is absent in Microsoft Excel, but luckily most other software does offer this option. Google Docs (or Google Drive), OpenOffice and LibreOffice all have spreadsheet software that either saves in UTF-8 by default, or allows you to specify the encoding used when saving the file (you may have to use ‘save as’). Simply open the file or copy/paste the contents into one of the above spreadsheet programs, and save it as a new file. The system should now correctly recognize all characters.

Diagnosing other CSV problems

Sometimes CSV files can have other errors (like messed up columns), and sometimes these just are noticeable in Excel. You can use the free CSV viewer and editor CSVed, which you can download from http://csved.sjfrancke.nl/