--- trunk/perl/scripts/est-spider	2005/09/10 09:31:53	18
+++ trunk/perl/scripts/est-spider	2005/09/18 18:21:06	30
@@ -5,6 +5,8 @@
 use File::Which;
 use HyperEstraier;
 use Text::Iconv;
+#use File::MMagic;
+use File::MMagic::XS qw/:compat/;
 
 # do we use Node API?
 my $node_url;
@@ -15,14 +17,17 @@
 my $exclude;
 
 #$verbose = 1;
+my $debug = 0;
+my $force = 0;
 
 my $result = GetOptions(
 	"collection=s" => \$collection,
 	"path=s" => \$path_add,
 	"verbose!" => \$verbose,
-	"debug!" => \$verbose,
+	"debug!" => \$debug,
 	"exclude=s" => \$exclude,
 	"node=s" => \$node_url,
+	"force!" => \$force,
 );
 
 my $dir = shift @ARGV || die "usage: $0 [dir]";
@@ -33,6 +38,9 @@
 
 my $pdftotext = which('pdftotext');
 
+#my $mm = new File::MMagic('/usr/share/misc/file/magic');
+my $mm = new File::MMagic::XS();
+
 my $iconv = new Text::Iconv('iso-8859-2', 'utf-8');
 
 select(STDERR); $|=1;
@@ -62,7 +70,8 @@
 
 find({ wanted => \&file, 
 	follow => 1,
-	no_chdir => 1
+	follow_skip => 2,
+	no_chdir => 1,
 }, $dir);
 
 unless ($node_url) {
@@ -92,11 +101,16 @@
 	# create a document object 
 	my $doc = HyperEstraier::Document->new;
 
-	my $title = $1 if ($contents =~ m#<title>(.+)</title>#is);
+	my $title = $1 if ($contents =~ m#<title>(.+?)</title>#is);
+
+	# chop long titles to 100 chars
+	$title = substr($title, 0, 100) . '...' if ($title && length($title) > 100);
+	# use path if no title is found
+	$title ||= $path;
 
 	# add attributes to the document object 
 	$doc->add_attr('@uri', "file:///$path");
-	$doc->add_attr('@title', $title || $path);
+	$doc->add_attr('@title', $iconv->convert($title));
 	$doc->add_attr('@size', $size);
 	$doc->add_attr('@mtime', $mtime);
 
@@ -122,9 +136,27 @@
 	my $path = $_;
 	my $contents;
 
-	return if (-l $path);
+	return if (! $force && -l $path || $path =~ m#/.svn# || $path =~ m/(~|.bak)$/);
+
+	my $mtime = (stat($path))[9];
+	my $mtime_db = $db->get_doc_attr_by_uri("file:///$path", '@mtime') || -2;
+
+	if ($mtime == $mtime_db) {
+		print STDERR "# same: $path $mtime\n" if ($verbose);
+		return unless($force);
+	} else {
+		print STDERR "# changed: $path $mtime != $mtime_db\n" if ($debug);
+	}
+
+	# skip files on which File::MMagic::XS croaks
+	return if ($path =~ m#\.au$#);
+
+	my $type = $mm->checktype_filename($path);
+	$type =~ s/\s+/ /gs;
+
+	print STDERR "# $path $type\n" if ($debug);
 
-	if ($pdftotext && -f $path && $path =~ m/\.pdf$/i) {
+	if ($pdftotext && -f $path && $type =~ m/pdf/i) {
 
 		print STDERR "$path {converting}" if ($verbose);
 
@@ -159,19 +191,22 @@
 			print STDERR " $page_nr" if ($verbose);
 			my $pre_tmp = $pre_html;
 			$pre_tmp =~ s/##page_nr##/$page_nr<\/title>/s;
-			dump_contents($db, $pre_tmp . $page . $post_html,time(), $path) if ($page !~ m/^\s*$/s);
+			dump_contents($db, $pre_tmp . $page . $post_html, $mtime, "$path#$page_nr") if ($page !~ m/^\s*$/s);
 			$page_nr++;
 		}
 
 	} else {
 
-		return if (! -f $path || ! m/\.(html*|php|pl|txt|info|log|text)$/i);
+#		return if (! -f $path || ! m/\.(html*|php|pl|txt|info|log|text)$/i);
+		return unless (-f $path && $type =~ m/html/ ||
+			($type =~ m#text# && $path =~ m/\.(php|pl|txt|info|log|text)$/io)
+		);
 
 		# skip index files
 		return if (m/index_[a-z]\.html*/i || m/index_symbol\.html*/i);
 
 		open(F,"$path") || die "can't open file: $path";
-		print STDERR "$path" if ($verbose);
+		print STDERR "$path ($type)" if ($verbose);
 		while(<F>) {
 			$contents .= "$_";
 		}
@@ -182,7 +217,7 @@
 		# add optional components to path
 		$path .= " $path_add" if ($path_add);
 
-		dump_contents($db, $contents,time(), $path);
+		dump_contents($db, $contents, $mtime, $path);
 	}
 
 	print STDERR "\n" if ($verbose);