When you first need to get some data into a spreadsheet, it's tempting to treat a CSV as just a text file where things are separated by commas. I've seen a lot of developers—myself included, early in my career—take the "quick and dirty" route by just interpolating strings. It feels efficient because you aren't loading a library or learning a new API; you're just manipulating strings, which you already know how to do.

The temptation of string interpolation

Imagine we're building a simple report that exports a list of users and their subscription plans. A naive implementation usually looks something like this:

users = [
  { name: "Alice Smith", plan: "Premium", email: "alice@example.com" },
  { name: "Bob Jones", plan: "Basic", email: "bob@example.com" }
]

csv_string = "Name,Plan,Email\n"
users.each do |user|
  csv_string += "#{user[:name]},#{user[:plan]},#{user[:email]}\n"
end

File.write("report.csv", csv_string)

On the surface, this works. If your data is clean and predictable, you've got a CSV file in five lines of code. But in a production environment, "clean and predictable" data is a myth. This approach assumes that none of your data fields will ever contain a comma, a newline, or a double quote.

Where the string approach falls apart

Here is where things get messy. What happens when a user signs up as "Company, Inc."? Or what if a field contains a quoted string? If you use the string interpolation method above, a name like "Company, Inc." will be written into the file as Company, Inc.,Premium,email@example.com. When you open that in Excel or Google Sheets, the "Inc." gets pushed into the "Plan" column, and everything shifts one cell to the right. Your report is now corrupted.

You could try to fix this by writing your own regex to find commas and wrap them in quotes, but that's a rabbit hole you don't want to go down. You'd essentially be rewriting a CSV parser from scratch, and you'll almost certainly miss an edge case that will crash a client's import script three months from now.

Leveraging the CSV standard library

The professional way to handle this in Ruby is to use the built-in CSV library. It doesn't just join strings; it handles the RFC 4180 standard, which means it knows exactly how to escape quotes and handle commas within values. I prefer using CSV.open because it streams the data to the file rather than building one massive string in memory, which is critical if you're ever exporting thousands of rows.

require 'csv'

users = [
  { name: "Company, Inc.", plan: "Premium", email: "contact@corp.com" },
  { name: "Bob \"The Builder\" Jones", plan: "Basic", email: "bob@example.com" }
]

CSV.open("report.csv", "wb") do |csv|
  # Write the header row
  csv << ["Name", "Plan", "Email"]

  # Write the data rows
  users.each do |user|
    csv << [user[:name], user[:plan], user[:email]]
  end
end

Notice the difference in the logic. Instead of worrying about delimiters and newlines, you're simply passing an array of values to the csv object. The library handles the heavy lifting: "Company, Inc." will be automatically wrapped in double quotes in the final file, and the double quotes in Bob's nickname will be properly escaped. It's cleaner, it's safer, and it's the way you'll see it done in any serious Ruby codebase.

📋 Practical Task

Exercise: Build an Inventory Audit Export Tool

Your task is to create a script that generates an inventory report for a warehouse. You must use the CSV library to ensure the data is formatted correctly, as some of the product names contain characters that would break a naive string-based implementation.

Requirements:

• Create an array of hashes representing products. Each product should have a name, sku, and description.
• Include at least one product name that contains a comma (e.g., "Widget, Deluxe Edition").
• Include at least one description that contains double quotes (e.g., "Known as the \"Gold Standard\" of widgets").
• Write the data to a file named inventory_audit.csv.
• The CSV must have a header row: Product Name, SKU, Description.

Once you've run your script, open the resulting inventory_audit.csv in a text editor to verify that the library automatically added quotes around the fields containing commas and handled the internal quotes correctly.