Commit | Line | Data |
fe3b3201 |
1 | # $Id: RSS.pm 1934 2006-04-22 05:13:55Z btrott $ |
0d5e38d1 |
2 | |
3 | package XML::Feed::RSS; |
4 | use strict; |
5 | |
6 | use base qw( XML::Feed ); |
0d5e38d1 |
7 | use DateTime::Format::Mail; |
8 | use DateTime::Format::W3CDTF; |
9 | |
973e1f9e |
10 | our $PREFERRED_PARSER = "XML::RSS"; |
11 | |
12 | sub init_empty { |
4e9c4625 |
13 | my ($feed, %args) = @_; |
14 | $args{'version'} ||= '2.0'; |
973e1f9e |
15 | eval "use $PREFERRED_PARSER"; die $@ if $@; |
4e9c4625 |
16 | $feed->{rss} = $PREFERRED_PARSER->new(%args); |
fe3b3201 |
17 | $feed->{rss}->add_module(prefix => "content", uri => 'http://purl.org/rss/1.0/modules/content/'); |
813f78d8 |
18 | $feed->{rss}->add_module(prefix => "dcterms", uri => 'http://purl.org/dc/terms/'); |
9a36f82c |
19 | $feed->{rss}->add_module(prefix => "atom", uri => 'http://www.w3.org/2005/Atom'); |
20 | $feed->{rss}->add_module(prefix => "geo", uri => 'http://www.w3.org/2003/01/geo/wgs84_pos#'); |
973e1f9e |
21 | $feed; |
22 | } |
23 | |
0d5e38d1 |
24 | sub init_string { |
25 | my $feed = shift; |
26 | my($str) = @_; |
973e1f9e |
27 | $feed->init_empty; |
0d5e38d1 |
28 | if ($str) { |
973e1f9e |
29 | $feed->{rss}->parse($$str); |
0d5e38d1 |
30 | } |
31 | $feed; |
32 | } |
33 | |
34 | sub format { 'RSS ' . $_[0]->{rss}->{'version'} } |
35 | |
36 | ## The following elements are the same in all versions of RSS. |
973e1f9e |
37 | sub title { shift->{rss}->channel('title', @_) } |
38 | sub link { shift->{rss}->channel('link', @_) } |
39 | sub description { shift->{rss}->channel('description', @_) } |
0d5e38d1 |
40 | |
813f78d8 |
41 | # This doesn't exist in RSS |
42 | sub id { } |
43 | |
0d5e38d1 |
44 | ## This is RSS 2.0 only--what's the equivalent in RSS 1.0? |
973e1f9e |
45 | sub copyright { shift->{rss}->channel('copyright', @_) } |
0d5e38d1 |
46 | |
47 | ## The following all work transparently in any RSS version. |
48 | sub language { |
973e1f9e |
49 | my $feed = shift; |
50 | if (@_) { |
51 | $feed->{rss}->channel('language', $_[0]); |
52 | $feed->{rss}->channel->{dc}{language} = $_[0]; |
53 | } else { |
54 | $feed->{rss}->channel('language') || |
55 | $feed->{rss}->channel->{dc}{language}; |
56 | } |
0d5e38d1 |
57 | } |
58 | |
9a36f82c |
59 | sub self_link { |
60 | my $feed = shift; |
61 | |
62 | if (@_) { |
63 | my $uri = shift; |
64 | |
65 | $feed->{rss}->channel->{'atom'}{'link'} = |
66 | { |
67 | rel => "self", |
68 | href => $uri, |
69 | type => "application/rss+xml", |
70 | }; |
71 | } |
72 | |
73 | return $feed->{rss}->channel->{'atom'}{'link'}; |
74 | } |
75 | |
76 | |
0d5e38d1 |
77 | sub generator { |
973e1f9e |
78 | my $feed = shift; |
79 | if (@_) { |
80 | $feed->{rss}->channel('generator', $_[0]); |
81 | $feed->{rss}->channel->{'http://webns.net/mvcb/'}{generatorAgent} = |
82 | $_[0]; |
83 | } else { |
84 | $feed->{rss}->channel('generator') || |
85 | $feed->{rss}->channel->{'http://webns.net/mvcb/'}{generatorAgent}; |
86 | } |
0d5e38d1 |
87 | } |
88 | |
89 | sub author { |
973e1f9e |
90 | my $feed = shift; |
91 | if (@_) { |
92 | $feed->{rss}->channel('webMaster', $_[0]); |
93 | $feed->{rss}->channel->{dc}{creator} = $_[0]; |
94 | } else { |
95 | $feed->{rss}->channel('webMaster') || |
96 | $feed->{rss}->channel->{dc}{creator}; |
97 | } |
0d5e38d1 |
98 | } |
99 | |
100 | sub modified { |
973e1f9e |
101 | my $rss = shift->{rss}; |
102 | if (@_) { |
103 | $rss->channel('pubDate', |
104 | DateTime::Format::Mail->format_datetime($_[0])); |
105 | ## XML::RSS is so weird... if I set this, it will try to use |
106 | ## the value for the lastBuildDate, which I don't want--because |
107 | ## this date is formatted for an RSS 1.0 feed. So it's commented out. |
108 | #$rss->channel->{dc}{date} = |
109 | # DateTime::Format::W3CDTF->format_datetime($_[0]); |
fe3b3201 |
110 | } else { |
111 | my $date; |
112 | eval { |
113 | if (my $ts = $rss->channel('pubDate')) { |
114 | $date = DateTime::Format::Mail->parse_datetime($ts); |
115 | } elsif ($ts = $rss->channel->{dc}{date}) { |
116 | $date = DateTime::Format::W3CDTF->parse_datetime($ts); |
117 | } |
118 | }; |
119 | return $date; |
0d5e38d1 |
120 | } |
121 | } |
122 | |
123 | sub entries { |
124 | my $rss = $_[0]->{rss}; |
125 | my @entries; |
126 | for my $item (@{ $rss->{items} }) { |
973e1f9e |
127 | push @entries, XML::Feed::Entry::RSS->wrap($item); |
0d5e38d1 |
128 | } |
129 | @entries; |
130 | } |
131 | |
973e1f9e |
132 | sub add_entry { |
33d4cb3f |
133 | my $feed = shift; |
134 | my $entry = shift || return; |
135 | $entry = $feed->_convert_entry($entry); |
973e1f9e |
136 | $feed->{rss}->add_item(%{ $entry->unwrap }); |
137 | } |
138 | |
139 | sub as_xml { $_[0]->{rss}->as_string } |
140 | |
141 | package XML::Feed::Entry::RSS; |
0d5e38d1 |
142 | use strict; |
143 | |
a749d9b9 |
144 | use XML::Feed::Content; |
145 | |
0d5e38d1 |
146 | use base qw( XML::Feed::Entry ); |
147 | |
973e1f9e |
148 | sub init_empty { $_[0]->{entry} = { } } |
149 | |
150 | sub title { |
151 | my $entry = shift; |
152 | @_ ? $entry->{entry}{title} = $_[0] : $entry->{entry}{title}; |
153 | } |
154 | |
155 | sub link { |
156 | my $entry = shift; |
157 | if (@_) { |
158 | $entry->{entry}{link} = $_[0]; |
159 | ## For RSS 2.0 output from XML::RSS. Sigh. |
160 | $entry->{entry}{permaLink} = $_[0]; |
161 | } else { |
162 | $entry->{entry}{link} || $entry->{entry}{guid}; |
163 | } |
164 | } |
a749d9b9 |
165 | |
166 | sub summary { |
973e1f9e |
167 | my $item = shift->{entry}; |
168 | if (@_) { |
169 | $item->{description} = ref($_[0]) eq 'XML::Feed::Content' ? |
170 | $_[0]->body : $_[0]; |
171 | ## Because of the logic below, we need to add some dummy content, |
172 | ## so that we'll properly recognize the description we enter as |
173 | ## the summary. |
fe3b3201 |
174 | if (!$item->{content}{encoded} && |
973e1f9e |
175 | !$item->{'http://www.w3.org/1999/xhtml'}{body}) { |
fe3b3201 |
176 | $item->{content}{encoded} = ' '; |
973e1f9e |
177 | } |
178 | } else { |
179 | ## Some RSS feeds use <description> for a summary, and some use it |
180 | ## for the full content. Pretty gross. We don't want to return the |
181 | ## full content if the caller expects a summary, so the heuristic is: |
182 | ## if the <entry> contains both a <description> and one of the elements |
183 | ## typically used for the full content, use <description> as summary. |
184 | my $txt; |
185 | if ($item->{description} && |
fe3b3201 |
186 | ($item->{content}{encoded} || |
973e1f9e |
187 | $item->{'http://www.w3.org/1999/xhtml'}{body})) { |
188 | $txt = $item->{description}; |
189 | } |
190 | XML::Feed::Content->wrap({ type => 'text/plain', body => $txt }); |
a749d9b9 |
191 | } |
a749d9b9 |
192 | } |
0d5e38d1 |
193 | |
194 | sub content { |
973e1f9e |
195 | my $item = shift->{entry}; |
196 | if (@_) { |
197 | my $c = ref($_[0]) eq 'XML::Feed::Content' ? $_[0]->body : $_[0]; |
fe3b3201 |
198 | $item->{content}{encoded} = $c; |
973e1f9e |
199 | } else { |
200 | my $body = |
fe3b3201 |
201 | $item->{content}{encoded} || |
973e1f9e |
202 | $item->{'http://www.w3.org/1999/xhtml'}{body} || |
203 | $item->{description}; |
204 | XML::Feed::Content->wrap({ type => 'text/html', body => $body }); |
205 | } |
0d5e38d1 |
206 | } |
207 | |
208 | sub category { |
973e1f9e |
209 | my $item = shift->{entry}; |
210 | if (@_) { |
211 | $item->{category} = $item->{dc}{subject} = $_[0]; |
212 | } else { |
213 | $item->{category} || $item->{dc}{subject}; |
214 | } |
0d5e38d1 |
215 | } |
216 | |
217 | sub author { |
973e1f9e |
218 | my $item = shift->{entry}; |
219 | if (@_) { |
220 | $item->{author} = $item->{dc}{creator} = $_[0]; |
221 | } else { |
222 | $item->{author} || $item->{dc}{creator}; |
223 | } |
0d5e38d1 |
224 | } |
225 | |
226 | ## XML::RSS doesn't give us access to the rdf:about for the <item>, |
227 | ## so we have to fall back to the <link> element in RSS 1.0 feeds. |
228 | sub id { |
973e1f9e |
229 | my $item = shift->{entry}; |
230 | if (@_) { |
231 | $item->{guid} = $_[0]; |
232 | } else { |
233 | $item->{guid} || $item->{link}; |
234 | } |
0d5e38d1 |
235 | } |
236 | |
237 | sub issued { |
973e1f9e |
238 | my $item = shift->{entry}; |
239 | if (@_) { |
240 | $item->{dc}{date} = DateTime::Format::W3CDTF->format_datetime($_[0]); |
241 | $item->{pubDate} = DateTime::Format::Mail->format_datetime($_[0]); |
242 | } else { |
ecac864a |
243 | ## Either of these could die if the format is invalid. |
244 | my $date; |
245 | eval { |
246 | if (my $ts = $item->{pubDate}) { |
247 | my $parser = DateTime::Format::Mail->new; |
248 | $parser->loose; |
249 | $date = $parser->parse_datetime($ts); |
250 | } elsif ($ts = $item->{dc}{date}) { |
251 | $date = DateTime::Format::W3CDTF->parse_datetime($ts); |
252 | } |
253 | }; |
254 | return $date; |
0d5e38d1 |
255 | } |
256 | } |
257 | |
258 | sub modified { |
973e1f9e |
259 | my $item = shift->{entry}; |
260 | if (@_) { |
fe3b3201 |
261 | $item->{dcterms}{modified} = |
973e1f9e |
262 | DateTime::Format::W3CDTF->format_datetime($_[0]); |
263 | } else { |
fe3b3201 |
264 | if (my $ts = $item->{dcterms}{modified}) { |
ecac864a |
265 | return eval { DateTime::Format::W3CDTF->parse_datetime($ts) }; |
973e1f9e |
266 | } |
0d5e38d1 |
267 | } |
268 | } |
269 | |
9a36f82c |
270 | sub lat { |
271 | my $item = shift->{entry}; |
272 | if (@_) { |
273 | $item->{geo}{lat} = $_[0]; |
274 | } else { |
275 | return $item->{geo}{lat}; |
276 | } |
277 | } |
278 | |
279 | sub long { |
280 | my $item = shift->{entry}; |
281 | if (@_) { |
282 | $item->{geo}{long} = $_[0]; |
283 | } else { |
284 | return $item->{geo}{long}; |
285 | } |
286 | } |
287 | |
288 | |
0d5e38d1 |
289 | 1; |