while (my $r = $c->get_request) {
#CSC: mancano i controlli di sicurezza
- my $inputuri = uri_unescape($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");
+ my $inputuri = $cgi->param('uri');
+ print "\nRequest: ".$r->url."\n\n";
print "\nUnescaped query: ".$http_query."\n";
if ($http_method eq 'GET' and $http_path eq "/getciconly") {
# finds the uri, url and filename
my $cicuri = $inputuri;
+ my $answerformat = $cgi->param('format');
+ $answerformat = "" if (not defined($answerformat));
+
+ if (($answerformat ne "gz") and ($answerformat ne "normal")
+ and ($answerformat ne "")) {
+ die "Wrong output format: $answerformat, must be 'normal' ".
+ "or 'gz'\n";
+ }
my $cicfilename = $cicuri;
$cicfilename =~ s/cic:(.*)/$1/;
print_request("cic",$cicuri,$cicurl,$cicfilename);
# Retrieves the file
- my $ciccontent = download(0,"cic",$cicurl,$cicfilename);
+ my $ciccontent = download(0,"cic",$cicurl,$cicfilename,$answerformat);
# Answering the client
answer($c,$ciccontent);
}
- } elsif ($http_method eq 'GET' and $http_path eq "/get") {
- # finds the uris, urls and filenames
- my $cicuri = $inputuri,
- $typesuri = $inputuri,
- $annuri = $inputuri;
- my $annsuffix;
- if ($inputuri =~ /\.types$/) {
- $cicuri =~ s/(.*)\.types$/$1/;
- undef($annuri);
- } elsif ($inputuri =~ /\.types\.ann$/) {
- $cicuri =~ s/(.*)\.types\.ann$/$1/;
- $typesuri =~ s/(.*)\.ann$/$1/;
- $annsuffix = ".types.ann";
- } elsif ($inputuri =~ /\.ann$/) {
- $cicuri =~ s/(.*)\.ann$/$1/;
- undef($typesuri);
- $annsuffix = ".ann";
- } else {
- undef($typesuri);
- undef($annuri);
- }
-
- my $cicfilename = $cicuri;
- $cicfilename =~ s/cic:(.*)/$1/;
- $cicfilename =~ s/theory:(.*)/$1/;
- $cicfilename = $helm_dir.$cicfilename;
-
- 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 $typesfilename = $cicfilename.$typesext if $typesuri;
- my $annfilename = $cicfilename.$annsuffix.$annext if $annuri;
- $cicfilename .= $cicext;
-
- if (!defined($cicurl) ||
- (!defined($typesurl) && $typesuri) ||
- (!defined($annuri) && $annuri))
- {
- print "\nNOT FOUND!!!!!\n";
- $c->send_error(RC_NOT_FOUND)
- } else {
- print_request("cic",$cicuri,$cicurl,$cicfilename);
- print_request("types",$typesuri,$typesurl,$typesfilename)
- if ($typesuri);
- print_request("ann",$annuri,$annurl,$annfilename)
- if ($annuri);
-
- # Retrieves the files
-
- my $ciccontent = download(1,"cic",$cicurl,$cicfilename);
- my $typescontent =
- download(1,"types",$typesurl,$typesfilename) if ($typesuri);
- my $anncontent =
- download(1,"ann",$annurl,$annfilename) if ($annuri);
-
- # Merging the files together
-
- my $merged = <<EOT;
-<?xml version="1.0" encoding="UTF-8"?>
-<cicxml uri="$cicuri">
-$ciccontent
-$typescontent
-$anncontent
-</cicxml>
-EOT
-
- # Answering the client
- answer($c,$merged);
- }
- } elsif ($http_method eq 'GET' and $http_path eq "/getdtd") {
+ } elsif ($http_method eq 'GET' and $http_path eq "/getdtd") {
my $filename = $inputuri;
$filename = $dtd_dir."/".$filename;
print "DTD: $inputuri ==> ($filename)\n";
sub download
{
- my ($remove_headers,$str,$url,$filename) = @_;
+ my ($remove_headers,$str,$url,$filename,$format) = @_;
my ($gz, $buffer);
#my $debug=1; # for DEBUG only
."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);
-
- # 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
+ 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);
+
+ # 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";
+ }
+ # $cont now contained uncompressed data
}
if ($remove_headers) {
$cont =~ s/<\?xml [^?]*\?>//sg;
} else {
$cont =~ s/DOCTYPE (.*) SYSTEM\s+"http:\/\/www.cs.unibo.it\/helm\/dtd\//DOCTYPE $1 SYSTEM "$myownurl\/getdtd?uri=/g;
}
+ if ($format eq "gz") {
+ gzip($cont,"$basefname.tmp");
+ open (TMP, "< $basefname.tmp")
+ or die "Can't open tempfile: $filename.tmp, exiting!\n";
+ $cont = "";
+ while(<TMP>) {
+ $cont .= $_;
+ }
+ close TMP;
+ unlink ($basefname . ".tmp") or
+ die "cannot unlink temporary file: $basefname.tmp\n";
+ }
+
return $cont;
}