# For details, see the HELM World-Wide-Web page,
# http://cs.unibo.it/helm/.
-#use strict;
-
-my $VERSION = "@VERSION@";
-
-# various variables
-my ($HELM_LIB_PATH);
-my (%map);
-
# First of all, let's load HELM configuration
use Env;
my $HELM_LIB_DIR = $ENV{"HELM_LIB_DIR"};
# set the cache mode, may be "gzipped" or "normal"
my $cachemode = $ENV{'HTTP_GETTER_CACHE_MODE'} || 'gzipped';
if (($cachemode ne 'gzipped') and ($cachemode ne 'normal')) {
- die "Invalid HTTP_GETTER_CACHE_MODE environment variable, must be".
- "'normal' or 'gzipped'\n";
+ die "Invalid HTTP_GETTER_CACHE_MODE environment variable, must be 'normal' or 'gzipped'\n";
}
# </ZACK>
# next require defines: $helm_dir, $html_link, $dtd_dir, $uris_dbm
require $HELM_LIB_PATH;
-# Let's override the configuration file
-my $style_dir = $ENV{"HELM_STYLE_DIR"} if (defined ($ENV{"HELM_STYLE_DIR"}));
-my $dtd_dir = $ENV{"HELM_DTD_DIR"} if (defined ($ENV{"HELM_DTD_DIR"}));
-
use HTTP::Daemon;
use HTTP::Status;
use HTTP::Request;
use LWP::UserAgent;
use DB_File;
use Compress::Zlib;
-use CGI;
-use URI::Escape;
#CSC: mancano i controlli sulle condizioni di errore di molte funzioni
#CSC: ==> non e' robusto
#CSC: altra roba da sistemare segnata con CSC
my $d = new HTTP::Daemon LocalPort => 8081;
-my $myownurl = $d->url;
-
-# Let's patch the returned URL
-$myownurl =~ s/\/$//; # chop the final slash
-my $myownport = $myownurl;
-$myownport =~ s/http:\/\/(.*):(.*)/$2/;
-$myownurl =~ s/http:\/\/(.*):(.*)/$1/;
-($myownurl) = gethostbyname($myownurl);
-$myownurl = "http://".$myownurl.":".$myownport;
-
tie(%map, 'DB_File', $uris_dbm.".db", O_RDONLY, 0664);
-print "Please contact me at: <URL:", $myownurl."/", ">\n";
+print "Please contact me at: <URL:", $d->url, ">\n";
print "helm_dir: $helm_dir\n";
-print "style_dir: $style_dir\n";
print "dtd_dir: $dtd_dir\n";
print "urls_of_uris.db: $uris_dbm.db\n";
print "cache mode: $cachemode\n";
while (my $r = $c->get_request) {
#CSC: mancano i controlli di sicurezza
- my $inputuri = uri_unescape($r->url);
+ my $inputuri = $r->url;
$inputuri =~ s/^[^?]*\?uri=(.*)/$1/;
print "\nRequest: ".$r->url."\n\n";
my $http_method = $r->method;
my $http_path = $r->url->path;
- my $http_query = uri_unescape($r->url->query);
- my $cgi = new CGI("$http_query");
-
- print "\nUnescaped query: ".$http_query."\n";
if ($http_method eq 'GET' and $http_path eq "/getciconly") {
# finds the uri, url and filename
my $cicfilename = $cicuri;
$cicfilename =~ s/cic:(.*)/$1/;
$cicfilename =~ s/theory:(.*)/$1/;
-
+# $cicfilename = $helm_dir.$cicfilename.".xml";
+# <gzip>
my $cicurl = $map{$cicuri};
- my $extension;
- if ($cicurl =~ /\.xml$/) { # non gzipped file
- $extension = ".xml";
- } elsif ($cicurl =~ /\.xml\.gz$/) { # gzipped file
- $extension = ".xml.gz";
- } else { # error: unknown extension
- die "unexpected extension in url: $cicurl, might be '.xml'".
- "or '.xml.gz'";
- }
+ my $extension;
+ if ($cicurl =~ /\.xml$/) { # non gzipped file
+ $extension = ".xml";
+ } elsif ($cicurl =~ /\.xml\.gz$/) { # gzipped file
+ $extension = ".xml.gz";
+ } else { # error: unknown extension
+ die "unexpected extension in url: $cicurl, might be '.xml' or '.xml.gz'";
+ }
$cicfilename = $helm_dir.$cicfilename.$extension;
+ #my $cicurl = $map{$cicuri};
+# </gzip>
if (!defined($cicurl)) {
print "\nNOT FOUND!!!!!\n";
$c->send_error(RC_NOT_FOUND)
$cicfilename =~ s/theory:(.*)/$1/;
$cicfilename = $helm_dir.$cicfilename;
+# my $typesfilename = $cicfilename.".types.xml" if $typesuri;
+# my $annfilename = $cicfilename.$annsuffix.".xml" if $annuri;
+# $cicfilename .= ".xml";
+
+# <gzip>
my $cicurl = $map{$cicuri};
my $typesurl = $map{$typesuri} if (defined($typesuri));
my $annurl = $map{$annuri} if (defined($annuri));
- my ($cicext, $typesext, $annext);
- if ($cicurl =~ /\.xml$/) { # normal file
- $cicext = ".xml";
- } elsif ($cicurl =~ /\.xml\.gz$/) { # gzipped file
- $cicext = ".xml.gz";
- } else {
- die "unexpected extension in url: $cicurl;".
- "might be '.xml' or '.xml.gz'";
- }
- if (defined($typesuri)) { # extension selection for types file
- if ($typesurl =~ /\.xml$/) { # normal file
- $typesext = ".types.xml";
- } elsif ($typesurl =~ /\.xml\.gz$/) { # gzipped file
- $typesext = ".types.xml.gz";
- } else {
- die "unexpected extension in url: $typesurl;".
- "might be '.xml' or '.xml.gz'";
- }
- }
- if (defined($annuri)) { # extension selection for annotation file
- if ($annurl =~ /\.xml$/) { # normal file
- $annext = ".xml";
- } elsif ($annurl =~ /\.xml\.gz$/) { # gzipped file
- $annext = ".xml.gz";
- } else {
- die "unexpected extension in url: $annurl".
- "might be '.xml' or '.xml.gz'";
- }
- }
+ my ($cicext, $typesext, $annext);
+ if ($cicurl =~ /\.xml$/) { # normal file
+ $cicext = ".xml";
+ } elsif ($cicurl =~ /\.xml\.gz$/) { # gzipped file
+ $cicext = ".xml.gz";
+ } else {
+ die "unexpected extension in url: $cicurl; might be '.xml' or '.xml.gz'";
+ }
+ if (defined($typesuri)) { # extension selection for types file
+ if ($typesurl =~ /\.xml$/) { # normal file
+ $typesext = ".types.xml";
+ } elsif ($typesurl =~ /\.xml\.gz$/) { # gzipped file
+ $typesext = ".types.xml.gz";
+ } else {
+ die "unexpected extension in url: $typesurl; might be '.xml' or '.xml.gz'";
+ }
+ }
+ if (defined($annuri)) { # extension selection for annotation file
+ if ($annurl =~ /\.xml$/) { # normal file
+ $annext = ".xml";
+ } elsif ($annurl =~ /\.xml\.gz$/) { # gzipped file
+ $annext = ".xml.gz";
+ } else {
+ die "unexpected extension in url: $annurl might be '.xml' or '.xml.gz'";
+ }
+ }
my $typesfilename = $cicfilename.$typesext if $typesuri;
my $annfilename = $cicfilename.$annsuffix.$annext if $annuri;
$cicfilename .= $cicext;
+# </gzip>
+
if (!defined($cicurl) ||
(!defined($typesurl) && $typesuri) ||
print "Using local copy\n";
open(FD, $filename) or die "Cannot open $filename\n";
$cont = "";
- while(<FD>) {
- # Vary bad heuristic for substituion of absolute URLs
- # for relative ones
- s/ENTITY (.*) SYSTEM\s+"/ENTITY $1 SYSTEM "$myownurl\/getdtd?uri=/g;
- $cont .= $_;
- }
+ while(<FD>) { $cont .= $_; }
close(FD);
answer($c,$cont);
} else {
die "Could not find DTD!";
}
- } elsif ($http_method eq 'GET' and $http_path eq "/getstyleconf") {
- my $filename = $inputuri;
- $filename = $style_dir."/config/".$filename;
- if (stat($filename)) {
- open(FD, $filename) or die "Cannot open $filename\n";
- $cont = "";
- while(<FD>) {
- s/DOCTYPE (.*) SYSTEM\s+"/DOCTYPE $1 SYSTEM "$myownurl\/getstyleconf?uri=/g;
- $cont .= $_;
- }
- close(FD);
- answer($c,$cont);
- } else {
- die "Could not find Style Configuration File!";
- }
- } elsif ($http_method eq 'GET' and $http_path eq "/getxslt") {
- my $filename = $inputuri;
- $filename = $style_dir."/".$filename;
- print "XSLT: $inputuri ==> ($filename)\n";
- if (stat($filename)) {
- print "Using local copy\n";
- open(FD, $filename) or die "Cannot open $filename\n";
- $cont = "";
- while(<FD>) {
- # Vary bad heuristic for substituion of absolute URLs
- # for relative ones
- s/xsl:import\s+href="/xsl:import href="$myownurl\/getxslt?uri=/g ;
- s/xsl:include\s+href="/xsl:include href="$myownurl\/getxslt?uri=/g ;
- $cont .= $_;
- }
- close(FD);
- answer($c,$cont);
- } else {
- die "Could not find XSLT!";
- }
} elsif ($http_method eq 'GET' and $http_path eq "/conf") {
my $quoted_html_link = $html_link;
$quoted_html_link =~ s/&/&/g;
$cont = "<?xml version=\"1.0\"?><html_link>$quoted_html_link</html_link>";
answer($c,$cont);
} elsif ($http_method eq 'GET' and $http_path eq "/update") {
- # rebuild urls_of_uris.db
- print "Update requested...\n";
- mk_urls_of_uris();
- kill(USR1,getppid()); # signal changes to parent
+ print "Update requested...";
+ update();
+ kill(USR1,getppid());
print " done\n";
answer($c,"<html><body><h1>Update done</h1></body></html>");
- } elsif ($http_method eq 'GET' and $http_path eq "/ls") {
- # send back keys that begin with a given uri
- my ($uritype,$uripattern,$uriprefix);
- my $baseuri = $cgi->param('baseuri');
- chop $baseuri if ($baseuri =~ /.*\/$/); # remove trailing "/"
- my $outype = $cgi->param('format'); # output type, might be 'txt' or 'xml'
- $uripattern = $baseuri;
- $uripattern =~ s/^.*:(.*)/$1/;
- if ($baseuri =~ /^cic:/) {
- $uritype = "cic";
- } elsif ($baseuri =~ /^theory:/) {
- $uritype = "theory";
- } else {
- $uritype = "invalid";
- }
- if ($uritype ne "invalid") { # uri is valid
- if (($outype ne 'txt') and ($outype ne 'xml')) { # invalid out type
- print "Invalid output type specified: $outype\n";
- answer($c,"<html><body><h1>Invalid output type, may be ".
- "\"txt\" or \"xml\"</h1></body></html>");
- } else { # valid output
- print "BASEURI $baseuri, FORMAT $outype\n";
- $cont = finduris($uritype,$uripattern,$outype);
- answer($c,$cont);
- }
- } else { # invalid uri
- print "Invalid uri: $baseuri, may begin with 'cic:', ".
- "'theory:' or '*:'\n";
- answer($c,"<html><body><h1>Invalid uri , may begin with ".
- "\"cic:\", \"theory:\" or \"*:\"</h1></body></html>");
- }
- } elsif ($http_method eq 'GET' and $http_path eq "/help") {
- print "Help requested!";
- answer($c,"<html><body><h1>HTTP Getter Version ".
- $VERSION."</h1></body></html>");
} else {
- print "\n";
- print "INVALID REQUEST!!!!!\n";
- print "(PATH: ",$http_path,", ";
- print "QUERY: ",$http_query,")\n";
+ print "\nINVALID REQUEST!!!!!\n";
$c->send_error(RC_FORBIDDEN)
}
print "\nRequest solved: ".$r->url."\n\n";
#================================
-sub finduris { # find uris for cic and theory trees generation
- my ($uritype,$uripattern,$format) = @_;
- my $content = "";
- my ($uri,$localpart,$basepart,$dirname,$suffix,$flags,$key);
- my (@itemz,@already_pushed_dir);
- my (%objects,%dirs); # map uris to suffixes' flags
-
- print "FINDURIS, uritype: $uritype, uripattern: $uripattern, ".
- "format: $format\n\n";
-
- if (($uritype eq "cic") or ($uritype eq "theory")) {
- # get info only of one type: cic or theory
- foreach (keys(%map)) { # select matching uris
- $uri = $_;
- if ($uri =~ /^$uritype:$uripattern\//) {
- $localpart = $uri;
- $localpart =~ s/^$uritype:$uripattern\/(.*)/$1/;
-
- if ($localpart =~ /^[^\/]*$/) { # no slash, an OBJECT
- $basepart = $localpart;
- $basepart =~ s/^([^.]*\.[^.]*)(\.types)?(\.ann)?/$1/; # remove exts .types or
- # .types.ann
- $flags = $objects{$basepart}; # get old flags
- if ($localpart =~ /\.ann$/) {
- $flags = add_flag("ann","YES",$flags);
- } else {
- $flags = add_flag("ann","NO",$flags);
- }
- if ($localpart =~ /\.types$/) {
- $flags = add_flag("types","YES",$flags);
- } elsif ($localpart =~ /\.types\.ann$/) {
- $flags = add_flag("types","ANN",$flags);
- } else {
- $flags = add_flag("types","NO",$flags);
- }
- $objects{$basepart} = $flags; # save new flags
- } else { # exists at least one slash, a DIR
- ($dirname) = split (/\//, $localpart);
- $dirs{$dirname} = ""; # no flags requirement for dir
- }
- }
- }
- } else {
- die "getter internal error: unsupported uritype: \"$uritype\"";
- }
- # now generate OUTPUT:
- # output will be saved in $content
- if ($format eq "txt") { # TXT output
- foreach $key (sort(keys %dirs)) {
- $content .= "dir, " . $key . "\n";
- }
- foreach $key (sort(keys %objects)) {
- $content .= "object, $key, " . $objects{$key} . "\n";
- }
- } elsif ($format eq "xml") { # XML output
- $content .= '<?xml version="1.0" encoding="ISO-8859-1"?>' . "\n";
- $content .= '<!DOCTYPE ls SYSTEM ';
- $content .= '"http://localhost:8081/getdtd?uri=ls.dtd">' . "\n\n";
- $content .= "<ls>\n";
- foreach $key (sort(keys %dirs)) {
- $content .= "\t<section>$key</section>\n";
- }
- foreach $key (sort(keys %objects)) {
- $content .= "\t<object name=\"$key\">\n";
- $flags = $objects{$key};
- $flags =~ s/^<(.*)>$/$1/;
- my ($annflag,$typesflag) = split /,/,$flags;
- $content .= "\t\t<ann value=\"$annflag\" />\n";
- $content .= "\t\t<types value=\"$typesflag\" />\n";
- $content .= "\t</object>\n";
- }
- $content .= "</ls>\n";
- } else { # may not enter this branch
- die "Getter internal error: invalid format received by finduris sub";
- }
- return $content;
-}
-
-sub add_flag {
-# manage string like: "<ann_flag,type_flag>"
-# "ann_flag" may be one of "ann_YES", "ann_NO"
-# "type_flag" may be one of "types_NO", "types_YES", "types_ANN"
-# when adding a flag the max between the current flag and the new flag
-# is taken, the orders are ann_NO < ann_YES and types_NO < types_YES <
-# types_ANN
- my ($flagtype,$newflag,$str) = @_;
- $str = "<,>" if ($str eq "");
- ($str =~ s/^<(.*,.*)>$/$1/) or die "Internal error: ".
- "wrong string format for flag adding in $str";
- my ($annflag,$typeflag) = split /,/,$str;
- if ($flagtype eq "ann") { # ANN flag handling
- if ($newflag eq "YES") {
- $annflag = "YES";
- } elsif ($newflag eq "NO") {
- $annflag = "NO" unless ($annflag eq "YES");
- } else {
- die "Internal error: annflag must be \"YES\" or \"NO\"";
- }
- } elsif ($flagtype eq "types") { # TYPES flag handling
- if ($newflag eq "ANN") {
- $typeflag = "ANN";
- } elsif ($newflag eq "YES") {
- $typeflag = "YES" unless ($typeflag eq "ANN");
- } elsif ($newflag eq "NO") {
- $typeflag = "NO"
- unless (($typeflag eq "ANN") or ($typeflag eq "YES"));
- } else {
- die "Internal error: typeflag must be \"YES\", \"NO\" or \"ANN\"";
- }
- } else {
- die "Internal error: unsupported flagtype \"$flagtype\"";
- }
- $str = "<$annflag,$typeflag>";
-}
#CSC: Too much powerful: creates even /home, /home/users/, ...
#CSC: Does not raise errors if could not create dirs/files
my ($gz, $buffer, $cont);
print "deflating $filename ...\n";
- $gz = gzopen($filename, "r")
- or die "Cannot open gzip'ed file $filename: $gzerrno";
+ $gz = gzopen($filename, "r") or die "Cannot open gzip'ed file $filename: $gzerrno";
$cont = "";
while ( $gz->gzread($buffer) > 0 ) {
$cont .= $buffer;
sub gzip { # gzip the content argument and save it to filename argument
my ($cont, $filename) = @_;
- #my $debug=1; # debug only
-
- print "gzopening $filename ...\n" if (defined($debug));
- my $gz = gzopen($filename, "w")
- or die "Cannot gzopen for writing file $filename: $gzerrno";
- print "gzwriting ...\n" if (defined($debug));
- $gz->gzwrite($cont) or die "error writing: $gzerrno, exiting!\n";
- print "gzclosing ...\n" if (defined($debug));
+
+ my ($gz, $cont);
+
+ $gz = gzopen($filename, "w") or die "Cannot gzopen for writing file $filename: $gzerrno";
+ $gz->gzwrite($cont) or die "error writing: $gzerrno\n" ;
$gz->gzclose();
}
sub download
{
my ($remove_headers,$str,$url,$filename) = @_;
+# <gzip>
my ($gz, $buffer);
- #my $debug=1; # for DEBUG only
-
my $resourcetype; # retrieve mode: "normal" (.xml) or "gzipped" (.xml.gz)
if ($filename =~ /\.xml$/) { # set retrieve mode
$resourcetype = "normal";
die "Unsupported download extension, might be '.gz' or '.xml'\n";
}
my $basefname = $filename;
- $basefname =~ s/\.gz$//; # get base resource name removing trailing .gz
+ $basefname =~ s/\.gz$//; # get base resource name removing trailing .gz
+# </gzip>
$cont = ""; # modified by side-effect by the callback function
my $localfname="";
} elsif (stat($basefname.".gz")) {
$localfname=$basefname.".gz";
}
- if ($localfname ne "") { # we already have local copy of requested file
- # check both possible cache entry: gzipped or normal
- print "Using local copy for the $str file\n";
- if ($localfname =~ /\.xml\.gz$/) { # deflating cached file and return it
- $cont = gunzip($localfname);
- } elsif ($localfname =~ /\.xml$/) { # just return cached file
- open(FD, $localfname) or die "Cannot open $localfname";
- while(<FD>) { $cont .= $_; }
- close(FD);
- } else { # error
- die "Internal error: unexpected file name $localfname,"
- ."must end with '.gz' or '.xml.gz'\n";
- }
- } else { # download file from net
- print "Downloading the $str file\n"; # download file
- my $ua = LWP::UserAgent->new;
- my $request = HTTP::Request->new(GET => "$url");
- my $response = $ua->request($request, \&callback);
+ if ($localfname ne "") { # we already have local copy of requested file
+ # check both possible cache entry: gzipped or normal
+ print "Using local copy for the $str file\n";
+# <gzip>
+ if ($localfname =~ /\.xml\.gz$/) { # deflating cached file and return it
+ $cont = gunzip($localfname);
+ } elsif ($localfname =~ /\.xml$/) { # just return cached file
+ open(FD, $localfname) or die "Cannot open $localfname";
+ while(<FD>) { $cont .= $_; }
+ close(FD);
+ } else { # error
+ die "Internal error: unexpected file name $localfname, must end with '.gz' or '.xml.gz'\n";
+ }
+# </gzip>
+ } else { # download file from net
+ print "Downloading the $str file\n"; # download file
+ $ua = LWP::UserAgent->new;
+ $request = HTTP::Request->new(GET => "$url");
+ $response = $ua->request($request, \&callback);
- # cache retrieved file to disk
-# <ZACK/> TODO: inefficent, I haven't yet undestood how to deflate
-# in memory gzipped file, without call "gzopen"
- print "Storing the $str file\n";
- print "Making dirs ...\n" if (defined($debug));
- mkdirs($filename);
- print "Opening tmp file for writing ...\n" if (defined($debug));
- open(FD, ">".$filename.".tmp") or die "Cannot open $filename.tmp\n";
- print "Writing on tmp file ...\n" if (defined($debug));
- print FD $cont;
- print "Closing tmp file ...\n" if (defined($debug));
- close(FD);
-
- # handle cache conversion normal->gzipped or gzipped->normal as user choice
- print "cachemode:$cachemode, resourcetype:$resourcetype\n" if (defined($debug));
- if (($cachemode eq 'normal') and ($resourcetype eq 'normal')) {
- # cache the file as is
- rename "$filename.tmp", $filename;
- } elsif (($cachemode eq 'gzipped') and ($resourcetype eq 'gzipped')) {
- # cache the file as is
- # and update the $cont variabile with deflated content
- rename "$filename.tmp", $filename;
- $cont = gunzip($filename);
- } elsif (($cachemode eq 'normal') and ($resourcetype eq 'gzipped')) {
- # deflate cache entry
- # and update $cont
- open(FD, "> $basefname") or die "cannot open $basefname\n";
- $cont = gunzip($filename.".tmp");
- print FD $cont;
- close(FD);
- unlink "$filename.tmp"; # delete old gzipped file
- } elsif (($cachemode eq 'gzipped') and ($resourcetype eq 'normal')) {
- # compress cache entry
- print "gzipping ...\n" if (defined($debug));
- gzip($cont, $basefname.".gz");
- unlink "$filename.tmp"; # delete old uncompressed file
- } else {
- die "Internal error, unsopported cachemode, resourcetype couple\n";
- }
- # $cont now contained uncompressed data
+ # cache retrieved file to disk
+# <ZACK/> TODO: inefficent, I haven't yet undestood how to deflate in memory gzipped file,
+# without call "gzopen"
+# <gzip>
+ print "Storing the $str file\n";
+ mkdirs($filename);
+ open(FD, ">".$filename.".tmp") or die "Cannot open $filename.tmp\n";
+ print FD $cont;
+ close(FD);
+
+ # handle cache conversion normal->gzipped or gzipped->normal as user choice
+ if (($cachemode eq 'normal') and ($resourcetype eq 'normal')) { # cache the file as is
+ rename "$filename.tmp", $filename;
+ } elsif (($cachemode eq 'gzipped') and ($resourcetype eq 'gzipped')) { # cache the file as is
+ # and update the $cont variabile with deflated content
+ rename "$filename.tmp", $filename;
+ $cont = gunzip($filename);
+ } elsif (($cachemode eq 'normal') and ($resourcetype eq 'gzipped')) { # deflate cache entry
+ # and update $cont
+ open(FD, "> $basefname") or die "cannot open $basefname\n";
+ $cont = gunzip($filename.".tmp");
+ print FD $cont;
+ close(FD);
+ unlink "$filename.tmp"; # delete old gzipped file
+ } elsif (($cachemode eq 'gzipped') and ($resourcetype eq 'normal')) { # compress cache entry
+ gzip($cont, $basefname.".gz");
+ unlink "$filename.tmp"; # delete old uncompressed file
+ } else {
+ die "Internal error, unsopported cachemode, resourcetype couple\n";
+ }
+ # $cont now contained uncompressed data
+# </gzip>
}
if ($remove_headers) {
$cont =~ s/<\?xml [^?]*\?>//sg;
$cont =~ s/<!DOCTYPE [^>]*>//sg;
- } else {
- $cont =~ s/DOCTYPE (.*) SYSTEM\s+"http:\/\/www.cs.unibo.it\/helm\/dtd\//DOCTYPE $1 SYSTEM "$myownurl\/getdtd?uri=/g;
}
return $cont;
}
$c->send_response($res);
}
-sub helm_wget {
-#retrieve a file from an url and write it to a temp dir
-#used for retrieve resource index from servers
- $cont = "";
- my ($prefix, $URL) = @_;
- my $ua = LWP::UserAgent->new;
- my $request = HTTP::Request->new(GET => "$URL");
- my $response = $ua->request($request, \&callback);
- my ($filename) = reverse (split "/", $URL); # get filename part of the URL
- open (TEMP, "> $prefix/$filename")
- || die "Cannot open temporary file: $prefix/$filename\n";
- print TEMP $cont;
- close TEMP;
-}
-
sub update {
untie %map;
tie(%map, 'DB_File', $uris_dbm.".db", O_RDONLY, 0664);
}
-
-sub mk_urls_of_uris {
-#rebuild $uris_dbm.db fetching resource indexes from servers
- my (
- $server, $idxfile, $uri, $url, $comp, $line,
- @servers,
- %urls_of_uris
- );
-
- untie %map;
- if (stat $uris_dbm.".db") { # remove old db file
- unlink($uris_dbm.".db") or
- die "cannot unlink old db file: $uris_dbm.db\n";
- }
- tie(%urls_of_uris, 'DB_File', $uris_dbm.".db", O_RDWR|O_CREAT, 0664);
-
- open (SRVS, "< $servers_file") or
- die "cannot open servers file: $servers_file\n";
- @servers = <SRVS>;
- close (SRVS);
- while ($server = pop @servers) { #cicle on servers in reverse order
- print "processing server: $server ...\n";
- chomp $server;
- helm_wget($tmp_dir, $server."/".$indexname); #get index
- $idxfile = $tmp_dir."/".$indexname;
- open (INDEX, "< $idxfile") or
- die "cannot open temporary index file: $idxfile\n";
- while ($line = <INDEX>) { #parse index and add entry to urls_of_uris
- chomp $line;
- ($uri,$comp) = split /[ \t]+/, $line;
- # build url:
- if ($comp =~ /gz/) {
- $url = $uri . ".xml" . ".gz";
- } else {
- $url = $uri . ".xml";
- }
- $url =~ s/cic:/$server/;
- $url =~ s/theory:/$server/;
- $urls_of_uris{$uri} = $url;
- }
- close INDEX;
- die "cannot unlink temporary file: $idxfile\n"
- if (unlink $idxfile) != 1;
- }
-
- untie(%urls_of_uris);
- tie(%map, 'DB_File', $uris_dbm.".db", O_RDONLY, 0664);
-}
-