A Brief Summary of an HTTP Client/Server Negotiation

It is beyond the scope of this book to explore all the information exchanged between server and client when a request is made, not least because PHP handles most of these details for you. You should gain a basic understanding of this process, however, especially if you intend to write scripts that fetch Web pages or check the status of Web addresses.

HTTP stands for Hypertext Transfer Protocol. It is essentially a set of rules that defines the process by which a client sends a request and a server returns a response. Both client and server provide information about themselves and the data to be transferred. Much of this information becomes available to you in superglobal arrays.

The Request

A client requests data from the server according to a strict set of rules. The request consists of up to three components:

• A request line

• A header section

• An entity body

The request line is mandatory. It consists of a request method, typically GET, HEAD, or POST; the address of the required document; and the HTTP version to be used (HTTP/1.0 or HTTP/1.1). A typical request for a document called mydoc.html might look like this:

GET /mydoc.html HTTP/1.0

The client is making a GET request. In other words, it is requesting an entire document but sending no data itself (in fact, you can send small amounts of data as part of a GET request by adding a query string to the URL). The HEAD method would be used if you wanted only information about a document. The POST method is used to transfer data from a client to the server, usually from an HTML form.

The request line is enough in itself to make a valid GET request. To inform the server that a request is complete, an empty line must be sent.

Most clients follow the request line with a header section in which name/value pairs can be sent to the server. Some of these become available to you as environment variables. Each client header consists of a key and value on one line separated by a colon. Table 14.2 lists a few of these.

Table 14.2. Some Client Headers

Name	Description
Accept	The media types with which the client can work.
Accept-Encoding	The types of data compression the client can handle.
Accept-Charset	The character sets the client prefers.
Accept-Language	The language the client prefers (en for English).
Host	The host to which a request is being made. Some servers that maintain multiple virtual hosts rely heavily on this header.
Referer	The document from which a request is being made.
User-Agent	Information about the client type and version.

For GET and HEAD methods, the header section ends the request and an empty line is sent to the server. For requests made using the POST method, an empty line is followed by the entity body. An entity body consists of any data to be sent to the server; this is usually a set of URL-encoded name/value pairs similar to those found in a query string.

Listing 14.2 shows a request sent to a server by Mozilla 5.0.

Listing 14.2 Typical Client Headers Sent by a Mozilla Browser

1: GET /index.html HTTP/1.1
2: Host: resources.corrosive.co.uk:9090
3: User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.2.1) Gecko/20030225
4: Accept: text/xml,application/xml,application/xhtml+xml,text/html;
5: Accept-Encoding: gzip, deflate, compress;q=0.9
6: Accept-Charset: ISO-8859-1, utf-8;q=0.66, *;q=0.66
7: Keep-Alive: 300
8: Connection: keep-alive

The Response

After a server has received a client's request, it sends a response to the client. The response usually consists of three parts:

• A status line

• A header section

• An entity body

As you can see, there's a lot of symmetry between a request and a response. In fact, certain headers can be sent by either client or server, especially those that provide information about an entity body.

The status line consists of the HTTP version the server is using (HTTP/1.0 or HTTP/1.1), a response code, and a text message that clarifies the meaning of the response code.

Many response codes are available that a server can send to a browser. Each code provides some information about the success or otherwise of the request. Table 14.3 lists some of the more common response codes.

Table 14.3. Some Response Codes

Code	Text	Description
200	OK	The request was successful, and the requested data will follow.
301	Moved Permanently	The requested data no longer exists on the server. A location header will contain a new address.
302	Moved Temporarily	The requested data has been moved. A location header will contain a new address.
404	Not Found	The data could not be found at the supplied address.
500	Internal Server Error	The server or a CGI script has encountered a severe problem in attempting to serve the data.

A typical response line, therefore, might look something like the following:

HTTP/1.1 200 OK

The header section includes a series of response headers, formatted in the same way as request headers. Table 14.4 lists some headers commonly sent by servers.

Table 14.4. Some Common Server Headers

Name	Description
Date	The current date
Server	The server name and version
Content-Type	The MIME type of content in the entity body
Content-Length	The size of the entity in bytes
Location	The full address of an alternative document

Listing 14.3 shows a typical server response. After the headers have been sent (lines 2–6), the server sends an empty line to the client (line 7) followed by the entity body (the document originally requested).

Listing 14.3 A Server Response

 1: HTTP/1.1 200 OK
 2: Date: Mon, 08 Sep 2003 19:24:35 GMT
 3: Server: Apache/2.0.47 (Unix) PHP/5.0.0b1
 4: X-Powered-By: PHP/5.0.0b1
 5: Connection: close
 6: Content-Type: text/html; charset=ISO-8859-1
 7:
 8: <!DOCTYPE html PUBLIC
 9:   "-//W3C//DTD XHTML 1.0 Strict//EN"
10:  "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
11: <html>
12: <head>
13: <title>Listing 14.3 A server response</title>
14: </head>
15: <body>
16: <div>
17: Hello
18: </div>
19: </body>
20: </html>