How to generate CSV from array of hashes in Perl?

Generating CSV output from an array of hashes is a common task in Perl, especially when exporting data for spreadsheets or reports. Perl offers flexible ways to handle this, ranging from manual string manipulation to using built-in or core modules designed for CSV processing.

Understanding the Data Structure

An array of hashes is one of Perl's popular data structures for tabular data. Each hash represents a row, with keys as column names and values as the cell values. For example:

my @data = (
  { name => "Alice", age => 30, city => "London" },
  { name => "Bob",   age => 25, city => "Paris" },
  { name => "Carol", age => 27, city => "New York" },
);

Key Points When Generating CSV

• Consistent Columns: You usually want consistent columns for each row. This means choosing a fixed set of keys/fields.
• Quoting & escaping: CSV fields sometimes need quotes, especially if they contain commas, quotes, or newlines.
• Header Row: The CSV usually starts with a header line listing column names.
• Perl Context and Sigils: Use appropriate sigils to access scalars ($hash{key}) and arrays (@array).

Manual CSV Generation Example

This example demonstrates a simple way to generate CSV without external modules, handling minimal quoting:

use strict;
use warnings;

my @data = (
  { name => "Alice", age => 30, city => "London" },
  { name => "Bob",   age => 25, city => "Paris" },
  { name => "Carol", age => 27, city => "New York" },
);

# Define columns explicitly to ensure order & presence
my @columns = qw(name age city);

# Print CSV header
print join(",", @columns), "\n";

# Print each row
foreach my $row (@data) {
    my @fields = map {
        my $val = defined $row->{$_} ? $row->{$_} : "";
        # Simple quoting if needed
        if ($val =~ /[",\n]/) {
            $val =~ s/"/""/g;          # Escape double quotes by doubling
            $val = qq{"$val"};         # Surround by double quotes
        }
        $val
    } @columns;
    print join(",", @fields), "\n";
}

This code:

• Fixes the column order with @columns to avoid Perl's unordered hash keys.
• Escapes any embedded quotes by doubling them, per CSV spec.
• Wraps fields containing commas, quotes, or newlines inside quotes.
• Prints directly to STDOUT for easy redirection or capture.

Using Core Module Text::CSV (Recommended for Robustness)

Perl core includes Text::CSV from version 5.8 (bundled with Perl), which handles CSV specifics like Unicode, escaping, separators, and other edge cases reliably.

use strict;
use warnings;
use Text::CSV;

my @data = (
  { name => "Alice", age => 30, city => "London" },
  { name => "Bob, Jr.", age => 25, city => "Paris" },
  { name => "Carol \"CJ\"", age => 27, city => "New York" },
);

my @columns = qw(name age city);
my $csv = Text::CSV->new({ binary => 1, eol => "\n" });

# Print header
$csv->print(*STDOUT, \@columns);

# Print rows
for my $row (@data) {
    my @fields = map { $row->{$_} // "" } @columns;
    $csv->print(*STDOUT, \@fields);
}

Text::CSV gets rid of manual quoting headaches and handles tricky fields (quotes, commas, newlines) properly.

Common Gotchas

• Hash Key Ordering: Hash keys in Perl are unordered, so you should explicitly specify column order.
• Missing Data: Handle undefined or missing values, e.g. with the defined-or operator // to prevent warnings.
• Unicode & Binary: Use Text::CSV->new({ binary => 1 }) when dealing with UTF-8 or binary data.
• Never trust manual concatenation for CSV in complex cases; prefer Text::CSV.

Summary

You can generate CSV from an array of hashes in Perl either by:

• Writing your own printer with correct quoting and fixed columns, or
• Using the Text::CSV core module for robust and easy CSV handling.