NAME XML::OverHTTP - A base class for XML over HTTP-styled web service interface DESCRIPTION This module is not used directly from end-users. As a child class of this, module authors can easily write own interface module for XML over HTTP-styled web service. METHODS PROVIDED This module provides some methods and requires other methods overridden by child classes. The following methods are to be called in your module or by its users. new This constructor method returns a new object for your users. It accepts query parameters by hash. my $api = MyAPI->new( %param ); MyAPI.pm inherits this XML::OverHTTP modules. add_param This method adds query parameters for the request. $api->add_param( %param ); It does not validate key names. get_param This method returns a current query parameter. $api->get_param( 'key' ); treepp This method returns an XML::TreePP object to make the request. $api->treepp->get( 'key' ); And you can set its object as well. my $mytpp = XML::TreePP->new; $api->treepp( $mytpp ); total_entries, entries_per_page and current_page parameters in $mytpp are updated. request This method makes the request for the web service and returns its response tree. my $tree = $api->request; After calling this method, the following methods are available. tree This method returns the whole of the response parsed by XML::TreePP parser. my $tree = $api->tree; Every element is blessed when "elem_class" is defined. root This method returns the root element in the response. my $root = $api->root; xml This method returns the response context itself. print $api->xml, "\n"; code This method returns the response status code. my $code = $api->code; # usually "200" when succeeded page This method returns a Data::Page object to create page navigation. my $pager = $api->page; print "Last page: ", $pager->last_page, "\n"; And you can set its object as well. my $pager = Data::Page->new; $api->page( $pager ); pageset This method returns a Data::Pageset object to create page navigation. The paging mode is "fixed" as default. my $pager = $api->pageset; $pager->pages_per_set( 10 ); print "First page of next page set: ", $page_info->next_set, "\n"; Or set it to "slide" mode if you want. my $pager = $api->pageset( 'slide' ); page_param This method returns pair(s) of query key and value to set the page number for the next request. my $hash = $api->page_param( $page ); The optional second argument specifies the number of entries per page. my $hash = $api->page_param( $page, $size ); The optional third argument incluedes some other query parameters. my $newhash = $api->page_param( $page, $size, $oldhash ); page_query This method returns a processed query string which is joined by '&' delimiter. my $query = $api->page_query(); # current page my $query = $api->page_query( $page, $size, $hash ); # specified page METHOD YOU MUST OVERRIDE You MUST override at least one method below: url This is a method to specify the url for the request to the web service. E.g., sub url { 'http://www.example.com/api/V1/' } METHODS YOU SHOULD OVERRIDE The methods that you SHOULD override in your module are below: root_elem This is a method to specify a root element name in the response. E.g., sub root_elem { 'rdf:RDF' } is_error This is a method to return "true" value when the response seems to have error. This returns "undef" when it succeeds. E.g., sub is_error { $_[0]->root->{status} != 'OK' } total_entries This is a method to return the number of total entries for "Data::Page". E.g., sub total_entries { $_[0]->root->{hits} } entries_per_page This is a method to return the number of entries per page for "Data::Page". E.g., sub entries_per_page { $_[0]->root->{-count} } current_page This is a method to return the current page number for "Data::Page". E.g., sub current_page { $_[0]->root->{-page} } page_param This is a method to return paging parameters for the next request. E.g., sub page_param { my $self = shift; my $page = shift || $self->current_page(); my $size = shift || $self->entries_per_page(); my $hash = shift || {}; $hash->{page} = $page if defined $page; $hash->{count} = $size if defined $size; $hash; } When your API uses SQL-like query parameters, offset and limit: sub page_param { my $self = shift; my $page = shift || $self->current_page() or return; my $size = shift || $self->entries_per_page() or return; my $hash = shift || {}; $hash->{offset} = ($page-1) * $size; $hash->{limit} = $size; $hash; } METHODS YOU CAN OVERRIDE You CAN override the following methods as well. http_method This is a method to specify the HTTP method, 'GET' or 'POST', for the request. This returns 'GET' as default. E.g., sub http_method { 'GET' } default_param This is a method to specify pairs of default query parameter and its value for the request. E.g., sub default_param { { method => 'search', lang => 'perl' } } notnull_param This is a method to specify a list of query parameters which are required by the web service. E.g., sub notnull_param { [qw( api_key secret query )] } These keys are checked before makeing a request for safe. query_class This is a method to specify a class name for query parameters. E.g., sub elem_class { 'MyAPI::Query' } The default value is "undef", it means a normal hash is used instead. attr_prefix This is a method to specify a prefix for each attribute in the response tree. XML::TreePP uses it. E.g., sub attr_prefix { '' } The default prefix is zero-length string "" which is recommended. text_node_key This is a method to specify a hash key for text nodes in the response tree. XML::TreePP uses it. E.g., sub text_node_key { '_text' } The default key is "#text". elem_class This is a method to specify a base class name for each element in the response tree. XML::TreePP uses it. E.g., sub elem_class { 'MyAPI::Element' } The default value is "undef", it means each elements is a just hashref and not bless-ed. force_array This is a method to specify a list of element names which should always be forced into an array representation in the response tree. XML::TreePP uses it. E.g., sub force_array { [qw( rdf:li item xmlns )] } force_hash This is a method to specify a list of element names which should always be forced into an hash representation in the response tree. XML::TreePP uses it. E.g., sub force_hash { [qw( item image )] } SEE ALSO XML::TreePP AUTHOR Yusuke Kawasaki COPYRIGHT AND LICENSE Copyright (c) 2007 Yusuke Kawasaki. All rights reserved. This program is free software; you can redistribute it and/or modify it under the same terms as Perl itself.