CS 40 - Lecture 19

Cover page image

Cay S. Horstmann

Lecture 18 Recap

Plain text: letters, numbers, symbols, no fonts, no colors, no bold, no italic, one size
Common use: Scrape text from web, spreadsheet, process contents
Iterate through words: Extend Text class
```
for (String word : getWords()) { ... }
```
Counting, Finding, Collecting: Return int, String, WordList

The Internet

Network of interconnected computers
Each computer has an address (e.g. 130.65.86.60)
Each computer has a host name (e.g. oslo.cs.sjsu.edu)
Developed by DARPA for sharing computer resources
First killer app: email
Second killer app: the web

The World-Wide Web

Collection of documents, stored on computers (“web servers”) connected to the internet
“Hyperlinks” let you jump from one document to another
Need a “browser” to display documents
Documents coded in HTML

HTML

Cay rowing

Hypertext markup language
Plain text, with tags <h1>Welcome to my page!<h1> to denote styles, fonts, colors, etc.
Tags for links <a href="http://horstmann.com/index.html">Click me!</a>
...and images <img src="http://horstmann.com/cay_rowing.gif">

Lab 1. View Tags

In your web browser, load the page http://horstmann.com/index.html. Then select View -> Page Source.
What examples of a heading and of an image tag do you see?ActiveLecture.org

Lab 2. Read Tags

Start Greenfoot and open TextWorld
Right-click on Text class, select new Text(), drag to world
Right-click on text icon, select void load(String source)
In dialog, enter "http://horstmann.com/index.html". Be sure to include the ""
Select explore
What happens? ActiveLecture.org

Lab 2 Continued

Now the web page is in a Text object. We can analyze it in the usual way.
Complete this class WebText1 that collects all hyperlinks.
Hyperlinks start with href
Try the program like this:
- Right-click on WebText1, select new WebText1(), drag to world
- Right-click on icon, select act
- The act method contains the following code, which runs your method and puts the result into the world
```
 public void act() 
   {
      load("http://horstmann.com/index.html")
      WordList words = getLinks();
      words.addToWorld(this);
   }
}
```
- Now right-click on the list icon and select explore

Bored?

There is a problem with the output of the previous program
We just want the link, not the surrounding href="..."
Use the substring method to clean up the string.
```
str = str.substring(6);
```
Gets the substring starting at position 6 (i.e. the first position after the ")
Put that statement before you add str to the word list. Now what happens?

Bored? (Cont'd)

You still have one more " to worry about
Not so easy--it isn't at a fixed position
First get length of string
```
int n = str.length();
```
Then call
```
str = str.substring(6, n - 1);
```
Try it out. You should get just the links.
You can become Google!

Google

Search Engine (duh)
How does it know all those pages?
It “crawls”, loading a page and reading its links
...and then loading those pages...
...and following the links in them.

Google

How does it know which pages are relevant?
Page rank algorithm
A page with many links to it is likely to be authoritative
Links from that page are also likely to be authoritative

Lab 3

To determine the "Google number" of a person, launch Google, type in the name, hit Search, and keep looking at results until you have information about the given person (and not just some other person with the same name). The page number is the Google number.
Type Cay Horstmann into Google. Whoa! Google number 1
Try Jeff Smith (another professor in the CS department). What is his Google number?
Why doesn't he have Google number 1?

HTTP

Browser uses very simple protocol (HTTP = Hypertext transfer protocol)
Connects to web server (e.g. horstmann.com)
Sends request: GET /index.html
Gets HTML page
Disconnects from web server
Stateless

Shopping Carts

Stateless protocol makes it hard to implement shopping cart
You tell browser to visit store page
Browser connects, gets HTML, disconnects, shows page
You click on "add to cart" link
Browser connects again. Server says “Hello, stranger!”
How can server remember you?

Cookies

Solution 1: Put funny numbers in links
Solution 2: Use cookies
You tell browser to visit store page
Server says “Hello stranger, next time we meet tell me that the secret handshake is w00z1e”
Browser gets HTML, disconnects, shows page
You click on "add to cart" link
Browser connects again and says “Hello server, the secret handshake is w00z1e. Give me this page.”

Lab 4: Cookies

In Mozilla, select Edit -> Preferences -> Privacy -> Show cookies
Find some ad related cookie. (If you don't have one, visit a site with lots of ads, then try again.) Expand the tree display. Click on a "Cookie name" and read the "Content" below. What do you get? ActiveLecture.org
That content is set back whenever you visit another site with ads from the same server.

Lab 4: Cookies

Uncheck "Accept Cookies from Sites"
Visit http://nytimes.com, click on some articles until you get a login prompt.
You do have an account with the New York Times, right? (If not, make one now.)
Log in.
What happens? Why? ActiveLecture.org

Reminders

No class on May 6, May 8
May 13: Last day of classes, project presentations
May 1: Present project progress in class (3 min./team)
Homework 7 now published
Want missing homework credit?
- Do any or all of homework #1 - #6
- Your solution must differ from the published solution in some way
- Bring your solution to the class on May 1 and explain it to me
- You will get 50% of the original score