perl/modules/W3C/LinkChecker/bin checklink,4.115,4.116

• From: Ville Skytta via cvs-syncmail <cvsmail@w3.org>
• Date: Mon, 22 Sep 2008 19:33:33 +0000
• To: www-validator-cvs@w3.org
• Message-Id: <E1KhrA5-00004s-CJ@lionel-hutz.w3.org>

Update of /sources/public/perl/modules/W3C/LinkChecker/bin
In directory hutz:/tmp/cvs-serv32752

Modified Files:
	checklink 
Log Message:
Add non-robot developer mode, thanks to Michael Ernst.

Index: checklink
===================================================================
RCS file: /sources/public/perl/modules/W3C/LinkChecker/bin/checklink,v
retrieving revision 4.115
retrieving revision 4.116
diff -u -d -r4.115 -r4.116
--- checklink	11 Sep 2008 16:51:41 -0000	4.115
+++ checklink	22 Sep 2008 19:33:31 -0000	4.116
@@ -33,9 +33,17 @@
 
 package W3C::UserAgent;
 
-use LWP::RobotUA 1.19 qw();
+use LWP::RobotUA   1.19 qw();
+use LWP::UserAgent      qw();
 
-@W3C::UserAgent::ISA = qw(LWP::RobotUA);
+# if 0, ignore robots exclusion (useful for testing)
+use constant USE_ROBOT_UA => 1;
+
+if (USE_ROBOT_UA) {
+  @W3C::UserAgent::ISA = qw(LWP::RobotUA);
+} else {
+  @W3C::UserAgent::ISA = qw(LWP::UserAgent);
+}
 
 sub new
 {
@@ -50,7 +58,16 @@
   # WWW::RobotRules <= 5.78 have bugs which cause suboptimal results with
   # User-Agent substring matching against robots.txt files; "User-Agent: *"
   # should work ok with all though, and "User-Agent: W3C-checklink" for >= 5.77
-  my $self = $class->SUPER::new($name, $from, $rules);
+  my $self;
+  if (USE_ROBOT_UA) {
+    $self = $class->SUPER::new($name, $from, $rules);
+  } else {
+    my %cnf;
+    @cnf{qw(agent from)} = ($name, $from);
+    $self = LWP::UserAgent->new(%cnf);
+    $self = bless $self, $class;
+  }
+
   $self->from(undef) unless $from_ok;
 
   $self->env_proxy();
@@ -244,7 +261,10 @@
   $REVISION    = sprintf('version %s (c) 1999-2008 W3C', $VERSION);
   my ($cvsver) = q$Revision$ =~ /(\d+[\d\.]*\.\d+)/;
   $AGENT       = sprintf('%s/%s [%s] %s',
-                         $PROGRAM, $VERSION, $cvsver, LWP::RobotUA->_agent());
+                         $PROGRAM, $VERSION, $cvsver,
+                         (W3C::UserAgent::USE_ROBOT_UA
+                           ? LWP::RobotUA->_agent()
+                           : LWP::UserAgent->_agent()));
 
   # Pull in mod_perl modules if applicable.
   eval {
@@ -393,7 +413,9 @@
 my $ua = W3C::UserAgent->new($AGENT); # @@@ TODO: admin address
 # @@@ make number of keep-alive connections customizable
 $ua->conn_cache({ total_capacity => 1}); # 1 keep-alive connection
-$ua->delay($Opts{Sleep_Time}/60);
+if ($ua->can('delay')) {
+  $ua->delay($Opts{Sleep_Time}/60);
+}
 $ua->timeout($Opts{Timeout});
 eval {
   $ua->allow_private_ips($Opts{Allow_Private_IPs});

Received on Monday, 22 September 2008 19:34:08 UTC