Question

I have a method that checks for broken links:

def self.check_prod_links
  require 'net/http'
  results = []
  Product.find_each(:conditions =>{:published => 1}) do |product|
    url = product.url 
    id = product.id
    uri = URI(url)
    begin
      response = Net::HTTP.get_response(uri)
    rescue
      http = Net::HTTP.new(uri.host, uri.port)
      http.use_ssl = true
      http.verify_mode = OpenSSL::SSL::VERIFY_NONE
      request = Net::HTTP::Get.new(uri.request_uri)
      response = http.request(request)
    rescue
      response = Net::HTTP.get_response("http://" + uri)  
    rescue => e
      p "Problem getting url: #{url} Error Message: #{e.message}"
    end
    p "Checking URL = #{url}. ID = #{id}. Response Code = #{response.code}" 
    unless response.code.to_i == 200
      product.update_attribute(:published, 0) 
      results << product
    end
  end
  return results
end

My understanding was that rescue => e should log all exceptions not caught by the previous rescue statements, and the method should continue running, however for some reason when certain URL’s are checked, the script quits with the following exception:

SSL_connect returned=1 errno=0 state=SSLv2/v3 read server hello A:
unknown protocol

How can I set this up so that if an exception is caught, it will be printed, and the task will continue running?

Also, how can I call the results array to render in the mailer view, is there a better way to do this so that I can add all products that were unpublished to an already existing mailer?

Thanks!

Answer 1

I’m pretty sure your first rescue will be the only one that works. Specifying => e just tells Ruby to store the exception in a variable called e. I’m thinking the code in your first rescue block is raising the error you are seeing and there is nothing else rescuing it. To be honest, this is kind of a mess of code and you might be better off refactoring it into smaller methods.