From: "Jesús Gabriel y Galán" <jgabrielygalan@...>
Date: 2012-10-10T05:27:40+09:00
Subject: Re: SocketError when letting ruby open url from txt-file

On Tue, Oct 9, 2012 at 7:58 PM, Sybren Kooistra <lists@ruby-forum.com> wrote:
> It's half a million files.. So I don't know if printing the file first
> would help (or what exactly do you mean)?

Yes, I meant that there's probably some bad (invalid) uris. Or maybe
garbage in the file.

> Is there a way to workaround possible bad-urls

Catch the exception and log the line, for example:


File.foreach("websites.txt") do |line| #loop on basis urls textfile
  begin
    searchablefile = Nokogiri::HTML(open(line)) #open each url
   # the rest of the logic
  rescue
    puts "there was an error opening #{line}"
  end
end

Jesus.