An instrument I use outputs a CSV file of data, but I cannot control the column names. I would like to change the header row before importing the data (without editing the raw CSV file) so that I can use the following code to import into my database:
def self.import(file)
CSV.foreach(file.path, headers: true) do |row|
Foo.create! row.to_hash
end
end
How do I completely replace the header row with one of my own?
You can use the :header_converters option with a lambda that uses a lookup table.
convert = {"from" => "to", "bad" => "good"}
CSV.foreach(file.path, headers: true, header_converters: lambda { |name| convert[name] }) do |row|
...
If the transformations you want to make to the names are simpler, you can just apply changes to the name in the lambda.
You can use the method in Ivan's answer if you want to completely rename the headers.
convert will becoming empty
convert[name] || "Placeholder" in the lambda. This will default the label to "Placeholder" if it's not in the lookup hash.new that might be creating this issue, as I only want to change the given one and all other should remain same CSV.new(body, headers: true, header_converters: lambda { |name| convert[name].to_sym }, converters: :all)lambda { |name| (convert[name] || name).to_sym } so the final version is: csv = CSV.new(body, headers: true, header_converters: lambda { |name| (convert[name] || name).to_sym }, converters: :all)You can pass array of custom column names instead of a simple true value to headers option:
CSV.foreach(file.path, headers: ['name', 'birthdate', 'gender'])
source: http://ruby-doc.org/stdlib-2.2.0/libdoc/csv/rdoc/CSV.html#method-c-new
