If your webbots are going to access sensitive information or handle money, they’ll need to authenticate, or sign in as registered users of websites. This chapter teaches you how to write webbots that access password-protected websites. As in previous chapters, you can practice what you learn with example scripts and special test pages on the book’s website.
Authentication is the processes of proving that you are who you say you are. You authenticate yourself by presenting something that only you can produce. Table 20-1 describes the three categories of things used to prove a person’s identity.
Table 20-1. Things That Prove a Person’s Identity
You Authenticate Yourself With . . . | Examples |
|---|---|
Something you know | Usernames and passwords; Social Security numbers |
Something you are (biometrics) | DNA samples; thumbprints; retina, voice, and facial scans |
Something you have | House keys, digital certificates, encoded magnetic cards, wireless key fobs, implanted canine microchips |
Most websites that require authentication ask for usernames and passwords (something you know). The username and password—also known as login criteria—are compared to records in a database. The user is allowed access to the website if the login criteria match the records in the database. Based on the login criteria, the website may optionally restrict the user to specific parts of the website or grant specific functionality.
Usernames and passwords are the most convenient way to authenticate people online because they can be authenticated with a browser and without the need for additional hardware or software.
Websites also authenticate through the use of digital certificates (something you have), which must be exchanged between client and server and validated before access to a website or service is granted. The intricacies of digital certificates are described in Chapter 19. If you skipped this chapter, this is a good time to read it. Otherwise, all you need to know is that digital certificates are files that reside on servers, or less frequently, on the hard drives of client computers. The contents of these certificate files are automatically exchanged to authenticate the computer that holds the certificate. You’re most apt to encounter digital certificates when using the HTTPS protocol (also know as SSL) to access secure websites. Here, the certificate authenticates the website and facilitates the use of an encrypted data channel. Less frequently, a certificate is required on the client computer as well, to access virtual private networks (VPNs), which allow remote users to access private corporate networks. PHP/CURL manages certificates automatically if you specify the https: protocol in the URL. PHP/CURL also facilitates the use of local certificates; in the odd circumstance that you require a client-side certificate, PHP/CURL and client-side certificates are covered in Appendix A.
Biometrics (something you are) are generally not used in online authentication and are beyond the scope of this chapter. Personally, I have only seen biometrics used to authenticate users to online services when biometric information is readily available, as in telemedicine.
Your webbots may encounter websites that use multiple forms of authentication, since authentication is strengthened when two or more techniques are combined. For example, ATMs require both an ATM card (something you have) and a personal identification number (PIN) (something you know). Similarly, the retailer Target experimented with an ATM-style authentication scheme when it introduced USB credit card readers that worked in conjunction with Target.com.
You may very well encounter certificates—and even biometrics—as a webbot developer, so the more familiar you are with the various forms of authentication, the more potential targets your webbots will have. You’ll find, however, that most webbots authenticate with simple usernames and passwords. The following sections describe the most common techniques for using usernames and passwords.