HTTP Overview

1
HTTP Overview

• Vijayan Sugumaran
• School of Business Administration
• Oakland University

2
What is HTTP?

• Hypertext Transfer Protocol
• Language used to describe how documents are sent
  over the internet
• Rules by which browsers make requests and servers
  supply responses
• Includes ways to
• Ask for a document by name
• Agree on data format
• Determine who the user is
• Decide how to handle outdated resources
• Indicate the results of the request
• When specifying a URL or clicking on a hyperlink,
  the browser translates into HTTP commands
• The server finds the resource and sends it back

3
HTTP Request Model

• HTTP is a stateless request/response protocol
• Client opens a socket to the Web servers HTTP
  port (80 by default)
• Client sends a ASCII text request line, followed
  by zero or more HTTP headers, an empty line, and
  any data that accompanies the request
• The web server parses the request and locates the
  specified resource
• Server writes a copy of the resource to the
  socket, where it is read by the client
• The server closes connection
• Once the connection is closed, the server doesnt
  remember anything about this request

4
Connecting to the Web Server

• Web server operates by listening for requests on
  a particular port (default port 80)
• If the server is running on any other port, it
  has to be part of the URL
• E.g. http//www.mycompany.com4311/myfile.htm
• Client initiates an HTTP request by opening a
  TCP/IP socket to the Web server port, and then
  opening input and output stream over the socket
• Parameters required to open the socket are the
  Web server host name, and the port number
• Server host name is extracted from the URL
• Output stream used to send commands to server
• Input stream used to read the server response

5
Sending the HTTP Request

• Once the socket connection is made, Web browser
  writes an HTTP command to request the document
• A request has up to four parts
• First part is request line
• Second part one or more request headers
• Third part Empty line consisting of carriage
  return or line feed
• Fourth part User data

6
Part I - Request Line

• Mandatory part of HTTP Request
• Consists of three tokens, separated by spaces
• First Token Request method
• Second Token request URI (derived from the URL
  stuff that comes after the host name)
• http//www.mycompany.com/myfile.html
• Third Token HTTP version the client understands
  (allowable values are HTTP/1.0 and HTTP/1.1)
• Example of Request line
• GET /myfile.html HTTP/1.0
• There are eight possible methods defined in HTTP
  (GET and POST are the most frequently used)

7
Request Methods

• GET A simple request to retrieve the resource
  identified in the URI
• HEAD Same as GET, except the server doesnt
  return the requested document. The server only
  returns the status line and headers
• POST A request for the server to accept data
  that will be written to the clients output
  stream
• PUT A request for the server to store the data
  in the request as the new contents of the
  specified URI
• DELETE A request for the server to delete the
  resource named in the URI
• OPTIONS A request for information about what
  request methods the server supports
• TRACE A request for the Web server to echo the
  HTTP request and its headers
• CONNECT Currently unimplemented method reserved
  for use with a tunneling proxy

8
Part II - Request Headers

• Second part one or more request headers
• Request headers inform the server of clients
  identity and capabilities
• Key/value pairs (key and value separated by
  colon)
• E.g Accept text/html, image/bmp,
  text/x-wap.wml
• one pair per line
• Typical Request Headers
• User-Agent (The vendor and the version of the
  client)
• Accept (A list of content types that
  client recognizes)
• Content-Length (number of bytes of data appended
  to the request)

9
Part III Empty Line

• This informs the server that no more headers
  follow
• Even if no headers exist, this empty line has to
  be sent so that the server doesnt look for any
  more headers

10
HTTP Response from Server

• Response also has up to four parts
• A status line
• Zero or more response headers
• An empty line signaling the end of the response
  headers, and
• The requested data

11
HTTP Response - Status line

• The status line consists of three tokens
• HTTP version
• Just as the client indicates the highest version
  it can understand, the server indicates its
  capabilities
• Response code
• Three digit numeric code that indicates whether
  the request succeeded or failed, if it failed,
  the reason why.
• An optional response description, which is a
  human readable explanation of the response code
• Status line example
• HTTP/1.0 200 OK

12
Client (Browser) Asks

• User types the following URL in the browser
• http//www.lyricnote.com/simple.htm
• Browser opens a socket connection to
  www.lyricnote.com host on port 80 and then writes
  the following line and an empty line after that.
• GET /simple.htm HTTP/1.0

13
Web Server Returns
HTTP/1.1 200 OK Date Tue, 12 Feb 2002 035543
GMT Server Apache/1.3.12 (Win32) Content-Length
241 Content-Type text/html ltHTMLgt ltBODYgt ltH3gtWel
comelt/H3gt to ltbgtThe Lyric Notelt/bgt, the best
Internet source for ltULgt ltLIgtsheet
music ltLIgtmusical instruments ltLIgtbooks on
musical topics ltLIgtmusic software,
and ltLIgtmusical gift items lt/ULgt lt/BODYgt lt/HTMLgt
14
Browser Renders

• Browser parses the status line and checks the
  status code request was successful
• The parses response headers
• Finds out that 241 bytes of html follows
• Reads html data and renders the document
• If the html document contains references to other
  resources, the browser requests the server for
  those resources
• E.g if the document contains an ltimggt tag, then
  the browser requests that image.
• The server doesnt send the image automatically
  along with the original request

15
Complex Request

• http//www.lyricnote.com/compound.htm
• HTTP Request
• GET /compound.htm HTTP/1.0
• Then the server sends the following.

16
Server Sends
HTTP/1.1 200 OK Date Tue, 12 Feb 2002 235543
GMT Server Apache/1.3.12 (Win32) Content-Length
380 Content-Type text/html ltHTMLgt ltHEADgt ltLINK
RELstylesheet HREFlyricnote.cssgt lt/HEADgt ltBO
DYgt ltIMG SRCimages/logo.jpggt ltHR
COLOR005A9C ALIGNLEFT WIDTH500gt ltH3gtWelc
omelt/H3gt to ltbgtThe Lyric Notelt/bgt, the best
Internet source for ltULgt ltLIgtsheet
music ltLIgtmusical instruments ltLIgtbooks on
musical topics ltLIgtmusic software,
and ltLIgtmusical gift items lt/ULgt lt/BODYgt lt/HTMLgt
17
Another Request

• Browser encounters style sheet request
• Makes a second HTTP request
• GET /lyricnote.css HTTP/1.0
• Web server retrieves the style sheet and sends it
  to the client

HTTP/1.1 200 OK Date Tue, 12 Feb 2002 235553
GMT Server Apache/1.3.12 (Win32) Connection
Keep-alive, close Content-Length
73 Content-Type text/plain H3 font-size
20px font-weight bold color 005A9C
18
Yet Another Request

• Browser interprets the style sheet and applies it
  to the document
• Encounters image tag and makes a request for the
  logo
• GET /images/logo.jpg HTTP/1.0
• Server responds with image data stream

HTTP/1.1 200 OK Date Tue, 12 Feb 2002 235559
GMT Server Apache/1.3.12 (Win32) Connection
Keep-alive, close Content-Length
1280 Content-Type text/plain (binary image data
follows)
19
GET Versus POST

• GET
• Form data appended to the URI
• Simplifies the work the server needs to do
• Limit to the length of the URI (4KB)
• Some servers wont even accept 2KB
• If you submit form data using GET and book mark
  the resulting page, then, you dont have to type
  the data
• However, the data is visible defeats having a
  password
• POST
• Write the data to the clients output stream
• Server takes time in processing this data

20
More Info on HTTP

• W3C World Wide Web Consortium
• http//www.w3.org/Protocols/
• Freesoft.org
• http//www.freesoft.org/CIE/RFC/1945/index.htm
• HTTP IETF (Internet Engineering Task Force)
• http//www-old.ics.uci.edu/pub/ietf/http/