The Perl Toolchain Summit needs more sponsors. If your company depends on Perl, please support this very important event.

NAME

Netscape::Server::Request - Perl interface to Netscape server Request

SYNOPSIS

 package Netscape::Server::Something;
 use Netscape::Server qw/:all/;

 sub handler {
     my($pb, $sn, $rq) = @_;
     ...
     $auth_type = $rq->auth_type;
     $path_info = $rq->path_info;
     $query_string = $rq->query_string;
     $remote_user = $rq->remote_user;
     $request_method = $rq->request_method;
     $server_protocol = $rq->server_protocol;
     $user_agent = $rq->user_agent;
     $vars = $rq->vars;
     $reqpb = $rq->reqpb;
     $headers = $rq->headers;
     $srvhdrs = $rq->srvhdrs;
     $rq->protocol_status($sn, $status, $reason);
     $proceed = $rq->protocol_start_response($sn, $rq);
     ...
 }

DESCRIPTION

The Netscape::Server::Request class provides a Perl-object interface to the Netscape Server API Request structure. Instances of the Netscape::Server::Request structure are passed as arguments to all Perl subroutines that are executed by a Netscape server that has been integrated with Perl using nsapi_perl.

For an overview of integrating Perl and NSAPI, see nsapi_perl. Suffice it to say here that nsapi_perl provides a mechanism by which a Perl interpreter is embedded into a Netscape server. The NSAPI can then be programmed to in Perl rather than in C. This is acheived by placing the appropriate hooks in the server configuration files; nsapi_perl will then call whatever Perl subroutines you wish at various stages of processing a request from a client.

When a Perl subroutine is called by nsapi_perl, it is passed three arguments:

 my($pb, $sn, $rq) = @_;

$pb is a reference to hash to the key=value pairs passed to the argument from the server configuration files; see nsapi_perl for more details. $sn is an instance of Netscape::Server::Session which has its own man page. $rq is an instance of Netscape::Server::Request and is the subject of the rest of this document.

OBJECT ABSTRACTION

A request in the world of the NSAPI is an entity that encapsulates the header data sent from the client and header data sent back by the server during an http transaction. In addition, information internal to the server generated in during a transaction is considered part of a request. In the NSAPI all this information is maintained in a C structure called a Request.

Netscape::Server::Request provides a Perl interface to the Request structure. It also provides methods that cause the server to perform an action based on the current state of the request.

Some Netscape::Server::Request methods require that an instance of Netcape::Server::Session be passed as an argument. You will find that those methods requiring this will have a synonym method defined in Netscape::Server::Session. This is so you don't have to remember whether such a method is to be written as

 $rq->method($sn);

or

 $sn->method($rq);

where $rq is an instance of Netscape::Server::Request and $sn is an instance of Netscape::Server::Session. Either method call will do the same thing.

INSTANCE METHODS

Netscape::Server::Request methods can be divided into those that return request attributes, those that access server variables, and those that perform actions.

Request Attributes

These methods return or set attributes of the request.

auth_type
 $auth_type = $rq->auth_type($type);

With no arguments returns the authorization type of the request. This will either be 'basic' if authorization has been provided by the client or undef if the client provided no authorization. The $type argument sets the authorization to the value of $type.

path_info
 $path_info = $rq->path_info($path);

With no arguments returns the extra path information appended to the URI of the request. Note that this may not be defined during the initial stages of processing a request; usually it is defined by NameTrans directives. The optional argument allows you to set the extra path information.

query_string
 $query_string = $rq->query_string($path);

With no arguments returns the query string information appended to the URI of the request. Note that this may not be defined during the initial stages of processing a request; usually it is defined by NameTrans directives. The optional argument allows you to set the query string information.

remote_user
 $remote_user = $rq->remote_user($user);

With no arguments returns the remote user as determined from http authentication. If no authentication has been provided with the request this will be undef. The optional argument allows you to set the remote user.

request_method
 $request_method = $rq->request_method($method);

With no arguments returns the request method of the request. This will probably be one of GET, POST, or HEAD. The optional argument allows you to set the request method.

server_protocol
 $server_protocol = $rq->server_protocol($protocol);

With no arguments returns something like "HTTP/1.0", or whatever protocol the server is using to process the request. The optional argument allows you to set the server protocol, although I'm not sure how this affects server internals.

user_agent
 $user_agent = $rq->user_agent($agent);

With no arguments returns the value sent by the browser in the "User-Agent" field of the http request header. The optional argument allows you to set the user agent.

Server Variable Methods

These methods allow you to work with the variables the server uses when constructing a response to a request. Each such variable is basically a key=value pair. The variables are grouped into 4 different categories. Each category of variable is accessed using a method specific to that category.

vars

This method accesses server working variables.

 $vars = $rq->vars;
 $vars = $rq->vars($key);
 $vars = $rq->vars($key, $value);

With no arguments returns a reference to hash containing the names and values of server working variables for this request. Server working variables are internal variables set up by the server during the various stages of processing a request. These variables describe things like the document root directory for the server, the full path to the requested file, and so on.

With one argument returns the value of the variable named by $key, or undef if the variable doesn't exist.

With two arguments sets the variable named by $key to $value. This is the "official" way for one nsapi_perl subroutine to pass data to a second subroutine that will run later on during the request.

I do not know of a definitive list of server working variable names, but the following seem to be in common usage in NSAPI documentation and examples.

auth-type: This variable is defined by subroutines running in the AuthTrans stage of processing a request. It should be to the literal basic if http authentication is in effect for this request.

auth-user: This variable is defined by subroutines running in the AuthTrans stage of processing a request. It should be set to the username as determined from the client's authentication information if authentication was successful.

ppath: This variable initially contains the partial path as determined from the clients request header. For instance if a client requests http://server/file.html, ppath would initially contain the string /file.html.

name: This variable can be created by a NameTrans subroutine to indicate that the object named by name is to be added to the set of active objects processing this request. See nsapi_perl and your Netscape server documentation for what an object is in this context.

path: This variable contains the full path to the requested file after all NameTrans subroutines have run. If don't plan on having the built-in NameTrans function run in addition to your subroutines, then you are responsible for setting this variable to the correct value.

You can create any number of your own variables to pass information to subroutines that will run later on in processing of the request.

reqpb

This method accesses client request line variables.

 $reqpb = $rq->reqpb;
 $reqpb = $rq->reqpb($key);
 $reqpb = $rq->reqpb($key, $value);

With no arguments returns a reference to a hash containing these entries:

method: The method of the request: either GET, HEAD, or POST.

uri: The URI the client asked for in the request.

protocol: The protocol of the request, as in HTTP/1.0 or something like that.

clf-request: The full test of the first line of the client's request.

With one argument this method returns the value of the field named by $key.

With two arguments this method sets the field named by $key to $value.

headers

This method accesses client request header variables.

 $headers = $rq->headers;
 $headers = $rq->headers($key);
 $headers = $rq->headers($key, $value);

With no arguments returns a reference to hash containing the client's request headers. The keys of this hash are the names of the various header lines conveted to all lower case, such as user-agent and cookie. The values are whatever text appeared in that field of the header.

With one argument this method returns the value of the request header field named by $key. Be sure to use the all lower-case form of the field name.

With two arguments sets the value of the request header field named by $key (again all lower-case) to $value. Setting one of these values is not an ethical thing.

srvhdrs

This method accesses server response header variables.

 $srvhdrs = $rq->srvhdrs;
 $srvhdrs = $rq->srvhdrs($key);
 $srvhdrs = $rq->srvhdrs($key, $value);

With no arguments returns a reference to a hash containing the server's response headers. The keys of this hash are the names of the various header lines converted to all lower case, such as content-type and set-cookie. The values are whatever text appears in that field of the header.

With one argument this method returns the value of the request header field name dby $key. Be sure to user the all lower-case form of the field name.

With two arguments sets the value of the request header field named by $key (again all lower-case) to $value. This is the "official" way to set the server's response headers. Make sure you do this before using protocol_start_response() (a method defined below).

Action Methods

These methods cause the server to take an action based on the current state of the request. They are listed in the approximate sequence in which they should be used by the nsapi_perl subroutine. Some of these methods indicate success or failure by returning one of the constants defined in the Netscape::Server module.

protocol_status
 $rq->protocol_status($sn, $status, $reason);

Sets the HTTP status of the session. $sn is an instance of Netscape::Server::Session. $status is one of the protocol-status constants, like PROTOCOL_OK, that can imported from Netscape::Server. $reason is an optional string sent to the client in the status line. If $reason is omitted the server will pick one based on $status defaulting to "unknown reason" in degenerate cases. This method returns nothing.

protocol_start_response
 $proceed = $rq->protocol_start_response($sn, $status, $reason);

Initiates an http response to the client by sending an http header based on the current state of $sn and $rq. $sn is an instance of Netscape::Server::Session. Returns either REQ_PROCEED, REQ_NOACTION or REQ_ABORTED. If REQ_PROCEED is returned the subroutine can continue as normal. If REQ_NOACTION is returned, the method succeeded, but the client needs no actual data (perhaps because the client has the data in its cache.) If REQ_ABORTED is returned, the method did not succeed.

To send data to the client see net_write() in Netscape::Server::Session.

AUTHOR

Benjamin Sugars <bsugars@canoe.ca>

SEE ALSO

perl(1), nsapi_perl, Netscape::Server, Netscape::Server::Session