Character encoding issue exporting rails data to CSV

Question

I'm exporting data to a CSV file in rails and in some of my fields, I'm getting character encoding issues like this when I open in Excel:

didn’t

I borrowed this code from an example and I'm assuming the encoding is off. Any idea what it should be?

send_data csv_data,
      :type => 'text/csv; charset=iso-8859-1; header=present',
      :disposition => "attachment; filename=#{filename}.csv"

Answer 1

When Excel opens the CSV file it just assumes an "iso-8859-1" character encoding. I guess it doesn't even know about the encoding information you send along within your HTTP reply. That's why setting this to UTF-8 doesn't work.

So in order to export your CSV file for Excel in Rails you could do this:

send_data Iconv.conv('iso-8859-1//IGNORE', 'utf-8', csv_data),
  :type => 'text/csv; charset=iso-8859-1; header=present',
  :disposition => "attachment; filename=#{filename}.csv"

This re-encodes your UTF-8 data string (that's the Rails default) to ISO-8859 and sends it. Along goes the information that this reply is actually ISO-8859-1 encoded (which won't make a difference for Excel but is technically correct if you should open it in a browser etc.).

Answer 2

This worked for me, with Chinese characters!excel csv fromat (BOM + UTF8)

def export_csv_excel
  ....

  # Add BOM to make excel using utf8 to open csv file
  head = 'EF BB BF'.split(' ').map{|a|a.hex.chr}.join()

  csv_str = CSV.generate(csv = head) do |csv|
    csv << [ , , , ...]
    @invoices.each do |invoice|
      csv << [ , , , ...]
    end
  end

  send_data csv_str, filename: "Invoices-#{Time.now.strftime("%y%m%d%H%M%S")}.csv", type: "text/csv"
end

source(Chinese): http://blog.inheart.tw/2013/09/rubyraisl-csv-excel.html

Answer 3

The answers above did not work for me on Mac Excel: Using iso-8859-1 would require I replace/remove weird characters, which is not a good enough solution for me, and using BOM with UTF8 worked under Windows but not under Mac Excel.

What worked for me is the WINDOWS-1252 encoding as suggested by https://stackoverflow.com/a/20194266/226255

def self.to_csv(options = {})
  (CSV.generate(options) do |csv|
    csv << self.headers

    all.each do |e|
      csv << e.values
    end
   end).encode('WINDOWS-1252', :undef => :replace, :replace => '')
end

Answer 4

module DownloadService
  def student_list
   File.open("#{file_name}", "w+:UTF-16LE:UTF-8") do |f|
      file = CSV.generate({:col_sep => "\t"}) do |c| 
        c << ['Canción ', 'años', 'etc']
      end
      f.write "\xEF\xBB\xBF"
      f.write(file)
   end
  end
end