Changeset 35805 for lang/perl/HTTP-DetectUserAgent

Timestamp:

11/01/09 01:57:34 (4 years ago)

Author:

mizuno_takaaki

Message:

add iphone/some crawlers

Location:

lang/perl/HTTP-DetectUserAgent/trunk

Files:

• lib/HTTP/DetectUserAgent.pm (modified) (4 diffs)
• t/01.browser.t (modified) (2 diffs)
• t/03.crawler.t (modified) (2 diffs)

lang/perl/HTTP-DetectUserAgent/trunk/lib/HTTP/DetectUserAgent.pm

@@ -118 +118 @@
             $self->{name} = 'Yahoo! Japan Crawler';
             $self->{vendor} = 'Yahoo';
+        }elsif( index($ua, 'yahoofeedseeker') != -1){
+            $self->{name} = 'YahooFeedSeeker';
+            $self->{vendor} = 'Yahoo';
         }
     }elsif( index($ua, 'msnbot') != -1){
@@ -130 +133 @@
         # http://help.baidu.jp/system/05.html
         $self->{name} = 'Baiduspider';
+        $self->{vendor} = 'Baidu';
+    }elsif( index($ua, 'baidumobaider') != -1){
+        # http://help.baidu.jp/system/05.html
+        $self->{name} = 'BaiduMobaider';
         $self->{vendor} = 'Baidu';
     }elsif( index($ua, 'yeti') != -1 && index($ua, 'naver') != -1){
@@ -151 +158 @@
         $self->{name} = 'Internet Archive';
         $self->{vendor} = 'Internet Archive';
-
+    }elsif( index($ua, 'tagoobot') != -1){
+        # http://www.tagoo.ru
+        $self->{name} = 'Tagoobot';
+        $self->{vendor} = 'Tagoo';
+    }elsif( index($ua, 'sogou web spider') != -1){
+        #http://www.sogou.com/docs/help/webmasters.htm#07
+        $self->{name} = 'Sogou';
+        $self->{vendor} = 'Sogou';
+    }elsif( index($ua, 'daumoa') != -1){
+        #http://ws.daum.net/aboutWebSearch.html
+        $self->{name} = 'Daumoa';
+        $self->{vendor} = 'Daum';
     }elsif( index($ua, 'spider') != -1 || index($ua, 'crawler') != -1 ){
         $self->{name} = 'Unknown Crawler';
@@ -424 +442 @@
     my ( $self, $ua ) = @_;
     return unless $ua;
-    if( $ua =~ /win(?:9[58]|dows|nt)/ ){
+    if( $ua =~ /iphone/ ){
+        $self->{os} = 'iPhone OS';
+    }elsif( $ua =~ /win(?:9[58]|dows|nt)/ ){
         $self->{os} = 'Windows';
     }elsif( $ua =~ /mac(?:intosh|_(?:powerpc|68000))/ ){

lang/perl/HTTP-DetectUserAgent/trunk/t/01.browser.t

@@ -48 +48 @@
 name: "Internet Explorer"
 version: "6.0"
+vendor: "Microsoft"
+os: "Windows"
+
+=== Internet Explorer 3
+--- input
+Mozilla/4.0 (compatible; MSIE 8.0; Windows NT 5.1; Trident/4.0; GTB6; .NET CLR 1.1.4322) 
+--- expected
+name: "Internet Explorer"
+version: "8.0"
 vendor: "Microsoft"
 os: "Windows"
@@ -105 +114 @@
 os: "Windows"
 
+=== iPhone
+--- input
+Mozilla/5.0 (iPhone; U; CPU iPhone OS 3_0_1 like Mac OS X; ja-jp) AppleWebKit/528.18 (KHTML, like Gecko) Version/4.0 Mobile/7A400 Safari/528.16
+--- expected
+name: "Safari"
+version: "4.0"
+vendor: "Apple"
+os: "iPhone OS"
 

lang/perl/HTTP-DetectUserAgent/trunk/t/03.crawler.t

@@ -29 +29 @@
 vendor: "Google"
 
-=== googlebot mobile
+=== googlebot mobile 1
 --- input
 DoCoMo/1.0/N505i/c20/TB/W20H10 (compatible; Googlebot-Mobile/2.1; +http://www.google.com/bot.html)
+--- expected
+name: "Googlebot Mobile"
+vendor: "Google"
+
+=== googlebot mobile 2
+--- input
+DoCoMo/2.0 N905i(c100;TB;W24H16) (compatible; Googlebot-Mobile/2.1; +http://www.google.com/bot.html) 
 --- expected
 name: "Googlebot Mobile"
@@ -63 +70 @@
 name: "Twiceler"
 vendor: "Cuil"
+
+=== BaiduMobaider 1
+--- input
+DoCoMo/2.0 P05A(c100;TB;W24H15) (compatible; BaiduMobaider/1.0; +http://www.baidu.jp/spider/)
+--- expected
+name: "BaiduMobaider"
+vendor: "Baidu"
+
+=== BaiduMobaider 2
+--- input
+DoCoMo/1.0/D506i/c20/TB/W20H10 (compatible; BaiduMobaider/1.0; +http://www.baidu.jp/spider/)
+--- expected
+name: "BaiduMobaider"
+vendor: "Baidu"
+
+=== BaiduMobaider 3
+--- input
+KDDI-CA3A UP.Browser/6.2.0.13.2 (GUI) MMP/2.0 (compatible; BaiduMobaider/1.0;+http://www.baidu.jp/spider/)
+--- expected
+name: "BaiduMobaider"
+vendor: "Baidu"
+
+=== BaiduMobaider 4
+--- input
+SoftBank/1.0/912SH/SHJ002/SN001111111111000 Browser/NetFront/3.4 Profile/MIDP-2.0 (compatible; BaiduMobaider/1.0;+http://www.baidu.jp/spider/)
+--- expected
+name: "BaiduMobaider"
+vendor: "Baidu"
+
+=== Tagoobot
+--- input
+Mozilla/5.0 (compatible; Tagoobot/3.0; +http://www.tagoo.ru)
+--- expected
+name: "Tagoobot"
+vendor: "Tagoo"
+
+=== Sogou web spider
+--- input
+Sogou web spider/4.0(+http://www.sogou.com/docs/help/webmasters.htm#07)
+--- expected
+name: "Sogou"
+vendor: "Sogou"
+
+=== Daumoa
+--- input
+Mozilla/5.0 (compatible; MSIE or Firefox mutant; not on Windows server; +http://ws.daum.net/aboutWebSearch.html) Daumoa/2.0
+--- expected
+name: "Daumoa"
+vendor: "Daum"
+
+=== YahooFeedSeeker
+--- input
+YahooFeedSeeker/1.0 (compatible; Mozilla 4.0; MSIE 5.5; http://my.yahoo.com/s/publishers.html)
+--- expected
+name: "YahooFeedSeeker"
+vendor: "Yahoo"