A Ruby script to find broken links

find_broken_links.rb

#!/usr/bin/env ruby
require 'net/http'
require 'timeout'

# A Ruby script to find broken links.
#
# Not perfect, but it gets the job done.
# The output from STDOUT can be used as a CSV:
#
#   ruby find_broken_links.rb > broken_links.csv

DIR = './_posts/'
SEARCH = 'https?:\/\/[a-zA-Z0-9!#$%&_+=,.?\/\-]+'

results = `grep -E -o -r '#{SEARCH}' '#{DIR}'`

results.split("\n").map(&:strip).each do |line|
    filename, url = line.split(':', 2)

    $stderr.puts "Checking: #{url}"
    begin
        Timeout.timeout(2) do
            res = Net::HTTP.get_response(URI(url))
        
            unless res.is_a?(Net::HTTPSuccess)
                puts "#{res.code},\"#{filename}\",\"#{url}\""
                $stderr.puts "BROKEN: #{res.code} #{url} #{filename}"
            end
        end
    rescue => e
        $stderr.puts "Skipping due to error!"
        $stderr.puts e
    end
end