Created
May 14, 2010 03:27
-
-
Save kga/400775 to your computer and use it in GitHub Desktop.
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
<ul> | |
<li><img src="http://img.imatsubu.jp/img/upload/account/20100513/17758_20100513110735906.jpg"><a href="http://imatsubu.jp/info/17758/">mai03</a> <p>うめこと茎わかめなう</p></li> | |
<li><img src="http://img.imatsubu.jp/img/upload/account/20100513/17754_20100513110715252.jpg"><a href="http://imatsubu.jp/info/17754/">erica06</a> <p>ホントにすぐ更新されるーおもしろい</p></li> | |
<li><img src="http://img.imatsubu.jp/img/upload/account/20100513/17750_20100513110645554.jpg"><a href="http://imatsubu.jp/info/17750/">erika07</a> <p> バリバリデコメ使えてるよぉ( 」´0`)」 ERIKA </p></li> | |
<li><img src="http://img.imatsubu.jp/img/upload/account/20100513/17746_20100513110558541.jpg"><a href="http://imatsubu.jp/info/17746/">fonchi08</a> <p>電車にガタンゴトンと揺られながらお仕事に向かっています。皆様今日も一日頑張りまっしょいp(^^)q</p></li> | |
<li><img src="http://img.imatsubu.jp/img/upload/account/20100513/17742_20100513110532782.jpg"><a href="http://imatsubu.jp/info/17742/">rurika09</a> <p>いまつぶ初投稿☆デコメがよくわからないので色合いは淡白になりますが汗よろしくお願いします☆</p></li> | |
<li><img src="http://img.imatsubu.jp/img/upload/account/20100513/17738_20100513110509519.jpg"><a href="http://imatsubu.jp/info/17738/">suzuka11</a> <p>外の空気が心地好い〜♪</p></li> | |
<li><img src="http://img.imatsubu.jp/img/upload/account/20100513/17734_20100513110447260.jpg"><a href="http://imatsubu.jp/info/17734/">yui12</a> <p>まいぷるさんと一緒になうったなう。</p></li> | |
<li><img src="http://img.imatsubu.jp/img/upload/account/20100513/17730_20100513110425997.jpg"><a href="http://imatsubu.jp/info/17730/">serina13</a> <p>アイドリング!!!の曲の歌詞を出すので、何の曲か当てて下さい○Level1「よりどりみどり」わっかるかなー</p></li> | |
<li><img src="http://img.imatsubu.jp/img/upload/account/20100513/17726_20100513110401287.jpg"><a href="http://imatsubu.jp/info/17726/">hitomi14</a> <p>酒井瞳です。こんにちは〜♪</p></li> | |
<li><img src="http://img.imatsubu.jp/img/upload/account/20100513/17722_20100513110338223.jpg"><a href="http://imatsubu.jp/info/17722/">nao15</a> <p>POPPOP(笑) 楽しも〜う</p></li> | |
<li><img src="http://img.imatsubu.jp/img/upload/account/20100513/17718_20100513110311694.jpg"><a href="http://imatsubu.jp/info/17718/">ami16</a> <p>ないすファン様!(^ω^)タイトルつけたほーが見やすいのね ないすないす了解まかせて!(゜▽゜)</p></li> | |
<li><img src="http://img.imatsubu.jp/img/upload/account/20100513/17714_20100513110210314.jpg"><a href="http://imatsubu.jp/info/17714/">hitomi17</a> <p>改札口無事発見なうHitomi</p></li> | |
<li><img src="http://img.imatsubu.jp/img/upload/account/20100513/17710_20100513110149735.jpg"><a href="http://imatsubu.jp/info/17710/">yurika19</a> <p>ターピータピオカなう♪今日もお天気や!</p></li> | |
<li><img src="http://img.imatsubu.jp/img/upload/account/20100513/17706_20100513110124467.jpg"><a href="http://imatsubu.jp/info/17706/">ai20</a> <p>はい。ゆっぺとお仕事前にタピオカなう。</p></li> | |
<li><img src="http://img.imatsubu.jp/img/upload/account/20100513/17698_20100513110058764.jpg"><a href="http://imatsubu.jp/info/17698/">kaede21</a> <p>電車 1駅だけど寝過ごしてしまったー!!でも全然 間に合うから大丈夫なう</p></li> | |
<li><img src="http://img.imatsubu.jp/img/upload/account/20100513/17690_20100513110031923.jpg"><a href="http://imatsubu.jp/info/17690/">ruka22</a> <p></p></li> | |
<li><img src="http://img.imatsubu.jp/img/upload/account/20100513/17686_20100513110011290.jpg"><a href="http://imatsubu.jp/info/17686/">yuna23</a> <p>今日あったかい〜幸せ!</p></li> | |
<li><img src="http://img.imatsubu.jp/img/upload/account/20100513/17682_20100513105829738.jpg"><a href="http://imatsubu.jp/info/17682/">manami24</a> <p>今日も一日頑張りましょう</p></li> | |
<li><img src="http://img.imatsubu.jp/img/upload/account/20100512/17166_20100512211937670.jpg"><a href="http://imatsubu.jp/info/17166/">kaoru25</a> <p></p></li> | |
<li><img src="http://img.imatsubu.jp/img/upload/account/20100512/17154_20100512211959566.jpg"><a href="http://imatsubu.jp/info/17154/">chika26</a> <p>勉強中</p></li> | |
</ul> |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
#!/usr/bin/env perl | |
use strict; | |
use warnings; | |
use Config::Pit; | |
use Encode qw(encode); | |
use URI; | |
use WWW::Mechanize; | |
use Web::Scraper; | |
my $url = URI->new('http://imatsubu.jp'); | |
my $conf = Config::Pit::get('imatsubu.jp', require => +{ | |
id => 'id for imatsubu', | |
pw => 'pw for imatsubu', | |
}); | |
my $mech = WWW::Mechanize->new; | |
$mech->get($url); | |
$mech->submit_form( | |
form_id => 'loginForm', | |
fields => +{ | |
'frontAccountBean.loginId' => $conf->{id}, | |
'frontAccountBean.loginPassword' => $conf->{pw}, | |
}, | |
); | |
$url->path('/list/19'); | |
$mech->get($url); | |
my $list19 = $mech->content; | |
#use File::Slurp qw(slurp); | |
#my $list19 = slurp('list19.html'); | |
$list19 =~ s/[\n\r]+/ /g; | |
$list19 =~ s/ / /g; | |
my $tsubu = scraper { | |
process 'div.timeline div.mutters', 'tsubu[]' => scraper { | |
process 'div.thn', 'info' => scraper { | |
process 'img', 'avatar' => '@src'; | |
process 'img', 'id' => '@alt'; | |
process 'a', 'link' => '@href'; | |
}; | |
process 'div.spouting div.container', 'body' => scraper { | |
process 'div.content', 'content' => 'TEXT'; | |
process 'p.tim', 'date' => 'TEXT'; | |
} | |
}; | |
}; | |
my $res = $tsubu->scrape($list19, $url); | |
print "<ul>\n"; | |
for my $tsubu (@{ $res->{tsubu} }) { | |
printf qq!<li><img src="%s"><a href="%s">%s</a><span>%s</span><p>%s</p></li>\n!, | |
$tsubu->{info}{avatar}, | |
$tsubu->{info}{link}, | |
$tsubu->{info}{id}, | |
$tsubu->{body}{date}, | |
encode('utf8', $tsubu->{body}{content}); | |
} | |
print "</ul>\n"; |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment