| 1 | # PRWire.pm | ||
| 2 | # by Jim Smyser | ||
| 3 | # Copyright (C) 2000 by Jim Smyser | ||
| 4 | # $Id: PRWire.pm,v 1.00 2000/04/07 02:33:19 jims Exp $ | ||
| 5 | |||
| 6 | |||
| 7 | package WWW::Search::PRWire; | ||
| 8 | |||
| 9 | =head1 NAME | ||
| 10 | |||
| 11 | WWW::Search::PRWire - class for viewing latest Press Releases | ||
| 12 | |||
| 13 | =head1 SYNOPSIS | ||
| 14 | |||
| 15 | use WWW::Search; | ||
| 16 | my $search = new WWW::Search('PRWire'); | ||
| 17 | $search->native_query(WWW::Search::escape_query('NULL')); | ||
| 18 | while (my $result = $search->next_result()) | ||
| 19 | { | ||
| 20 | print $result->url, "\n"; | ||
| 21 | } | ||
| 22 | |||
| 23 | =head1 DESCRIPTION | ||
| 24 | |||
| 25 | Class for WWW::Search for fetching and parsing latest PRWire news | ||
| 26 | headlines. F<http://www.prnewswire.com>. This code should | ||
| 27 | serve as an example of using WWW::Search to parse useful data | ||
| 28 | from pages other than its usual searching methods. Yes, WWW::Search | ||
| 29 | is useful for retrieving data when there is "next" pages to get and | ||
| 30 | yet, no search interface to interact with. | ||
| 31 | See USAGE. | ||
| 32 | |||
| 33 | This class exports no public interface; all interaction should | ||
| 34 | be done through WWW::Search objects. | ||
| 35 | |||
| 36 | =head1 USAGE | ||
| 37 | |||
| 38 | PRWire.pm does not deal with options $native_query (Query) or any | ||
| 39 | others except $maximum_to_retrieve. In a sense, this is not a | ||
| 40 | "search" backend. It simply parses all the latest headlines and | ||
| 41 | retrieves more as defined by $maximum_to_retrieve. | ||
| 42 | |||
| 43 | If you use with WebSearch or AutoSearch you will need to to send a | ||
| 44 | bogus query to prevent complaining of NO query. Search for NULL or | ||
| 45 | something. On a web page you could just have a button with a caption | ||
| 46 | "View Latest Press Releases" and optionally perhaps a option for how | ||
| 47 | many to return. | ||
| 48 | |||
| 49 | $result->title returns just the date and time of the article, so you | ||
| 50 | will also want to print $result->description after $result->title | ||
| 51 | so users will have descriptive text identifying the article. | ||
| 52 | |||
| 53 | =head1 AUTHOR | ||
| 54 | |||
| 55 | C<WWW::Search::PRWire> is written and maintained by Jim Smyser | ||
| 56 | <jsmyser@bigfoot.com>. | ||
| 57 | |||
| 58 | =head1 COPYRIGHT | ||
| 59 | |||
| 60 | (c) PR Newswire Redistribution, retransmission, republication or | ||
| 61 | commercial exploitation of the contents of PR Newswire are expressly | ||
| 62 | prohibited without the written consent of PR Newswire. | ||
| 63 | |||
| 64 | WWW::Search Copyright (c) 1996-1998 University of Southern California. | ||
| 65 | All rights reserved. | ||
| 66 | |||
| 67 | THIS SOFTWARE IS PROVIDED "AS IS" AND WITHOUT ANY EXPRESS OR IMPLIED | ||
| 68 | WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED WARRANTIES OF | ||
| 69 | MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. | ||
| 70 | |||
| 71 | =cut | ||
| 72 | #' | ||
| 73 | |||
| 74 | ##################################################################### | ||
| 75 | require Exporter; | ||
| 76 | @EXPORT = qw(); | ||
| 77 | @EXPORT_OK = qw(); | ||
| 78 | @ISA = qw(WWW::Search Exporter); | ||
| 79 | $VERSION = '1.0'; | ||
| 80 | |||
| 81 | 1 | ▶ | use Carp (); |
|
Subroutine
BEGIN: called CC=1 SCAR=0.0
| |||
| 82 | 0 | ▶ | use WWW::Search(qw(generic_option strip_tags)); |
|
Subroutine
BEGIN: called CC=1 SCAR=2.6
| |||
| 83 | |||
| 84 | require WWW::SearchResult; | ||
| 85 | |||
| 86 | sub native_setup_search { | ||
| 87 | |||
| 88 | my($self, $native_query, $native_options_ref) = @_; | ||
| 89 | $self->{_debug} = $native_options_ref->{'search_debug'}; | ||
| 90 | $self->{_debug} = 2 if ($native_options_ref->{'search_parse_debug'}); | ||
| 91 | $self->{_debug} = 0 if (!defined($self->{_debug})); | ||
| 92 | $self->{agent_e_mail} = 'jsmyser@bigfoot.com'; | ||
| 93 | $self->user_agent('user'); | ||
| 94 | $self->{_next_to_retrieve} = 1; | ||
| 95 | $self->{'_num_hits'} = 0; | ||
| 96 | if (!defined($self->{_options})) { | ||
| 97 | $self->{_options} = { | ||
| 98 | 'search_url' => 'http://www.prnewswire.com/tnw/tnw.shtml', | ||
| 99 | }; | ||
| 100 | } | ||
| 101 | |||
| 102 | my $options_ref = $self->{_options}; | ||
| 103 | if (defined($native_options_ref)) | ||
| 104 | { | ||
| 105 | # Copy in new options. | ||
| 106 | foreach (keys %$native_options_ref) | ||
| 107 | { | ||
| 108 | $options_ref->{$_} = $native_options_ref->{$_}; | ||
| 109 | } | ||
| 110 | } | ||
| 111 | # Process the options. | ||
| 112 | my($options) = ''; | ||
| 113 | foreach (sort keys %$options_ref) | ||
| 114 | { | ||
| 115 | next if (generic_option($_)); | ||
| 116 | $options .= $_ . '=' . $options_ref->{$_} . '&'; | ||
| 117 | } | ||
| 118 | chop $options; | ||
| 119 | $self->{_next_url} = $self->{_options}{'search_url'}; | ||
| 120 | } | ||
| 121 | # private | ||
| 122 | sub native_retrieve_some { | ||
| 123 | |||
| 124 | my ($self) = @_; | ||
| 125 | print STDERR "**PRWire Get Latest**\n" if $self->{_debug}; | ||
| 126 | |||
| 127 | # Fast exit if already done: | ||
| 128 | return undef if (!defined($self->{_next_url})); | ||
| 129 | |||
| 130 | # If this is not the first page of results, sleep so as to not | ||
| 131 | # overload the server: | ||
| 132 | $self->user_agent_delay if 1 < $self->{'_next_to_retrieve'}; | ||
| 133 | |||
| 134 | # Get some if were not already scoring somewhere else: | ||
| 135 | print STDERR "*Sending request (",$self->{_next_url},")\n" if $self->{_debug}; | ||
| 136 | my($response) = $self->http_request('GET', $self->{_next_url}); | ||
| 137 | |||
| 138 | $self->{response} = $response; | ||
| 139 | if (!$response->is_success) | ||
| 140 | { | ||
| 141 | return undef; | ||
| 142 | } | ||
| 143 | $self->{'_next_url'} = undef; | ||
| 144 | print STDERR "**Response\n" if $self->{_debug}; | ||
| 145 | # parse the output | ||
| 146 | my ($HEADER, $HITS, $TITLE, $DESC) = qw(HE HI TI DE); | ||
| 147 | my $hits_found = 0; | ||
| 148 | my $state = $HEADER; | ||
| 149 | my $hit = (); | ||
| 150 | foreach ($self->split_lines($response->content())) | ||
| 151 | { | ||
| 152 | next if m@^$@; # short circuit for blank lines | ||
| 153 | print STDERR " $state ===$_=== " if 2 <= $self->{'_debug'}; | ||
| 154 | |||
| 155 | if (m|<TITLE>.*?</TITLE>|i) | ||
| 156 | { | ||
| 157 | $state = $HITS; | ||
| 158 | } | ||
| 159 | elsif ($state eq $HITS && m@^<A HREF="(.*)">@i) | ||
| 160 | { | ||
| 161 | print "**Found Hit URL**\n" if 2 <= $self->{_debug}; | ||
| 162 | my ($url) = ($1); | ||
| 163 | if (defined($hit)) | ||
| 164 | { | ||
| 165 | push(@{$self->{cache}}, $hit); | ||
| 166 | }; | ||
| 167 | $hit = new WWW::SearchResult; | ||
| 168 | $hits_found++; | ||
| 169 | $url = "http://www.prnewswire.com" . $url; | ||
| 170 | $hit->add_url($url); | ||
| 171 | $state = $TITLE; | ||
| 172 | } | ||
| 173 | elsif ($state eq $TITLE && m|^(.+)</A>|i) | ||
| 174 | { | ||
| 175 | my $sTitle = $1; | ||
| 176 | $hit->title($sTitle); | ||
| 177 | $state = $DESC; | ||
| 178 | } | ||
| 179 | elsif ($state eq $DESC && m|^<DD>(.+)|i) | ||
| 180 | { | ||
| 181 | $hit->description($1); | ||
| 182 | $state = $HITS; | ||
| 183 | } | ||
| 184 | elsif ($state eq $HITS && m|Click\s<A HREF="(.*?)">.*?<IMG SRC.*?>|i) | ||
| 185 | { | ||
| 186 | $sURL = $1; | ||
| 187 | $self->{'_next_url'} = 'http://www.prnewswire.com' . $sURL; | ||
| 188 | print STDERR " **Next Tag is: ", $self->{'_next_url'}, "\n" if 2 <= $self->{_debug}; | ||
| 189 | } | ||
| 190 | else | ||
| 191 | { | ||
| 192 | print STDERR "**Nothing matched.**\n" if 2 <= $self->{_debug}; | ||
| 193 | } | ||
| 194 | } | ||
| 195 | if (defined($hit)) | ||
| 196 | { | ||
| 197 | push(@{$self->{cache}}, $hit); | ||
| 198 | } | ||
| 199 | return $hits_found; | ||
| 200 | } | ||
| 201 | 1; | ||
| 202 | |||
| 203 | |||