X-Git-Url: http://matita.cs.unibo.it/gitweb/?a=blobdiff_plain;f=helm%2Fhttp_getter%2Fhttp_getter.pl.in;h=e73d9f9878689b039352916e90b29a0abcfcd303;hb=b564f0671c133928e372e352d972a7663b6a9136;hp=9c68ae71077c7e97ee7ffc87ffd719fd19cab60c;hpb=45eeb63cbcf9807064a4d67acf8396a3e42ad744;p=helm.git
diff --git a/helm/http_getter/http_getter.pl.in b/helm/http_getter/http_getter.pl.in
index 9c68ae710..e73d9f987 100755
--- a/helm/http_getter/http_getter.pl.in
+++ b/helm/http_getter/http_getter.pl.in
@@ -23,18 +23,34 @@
# For details, see the HELM World-Wide-Web page,
# http://cs.unibo.it/helm/.
+my $VERSION = "@VERSION@";
+
# First of all, let's load HELM configuration
use Env;
-my $HELM_LIBRARY_DIR = $ENV{"HELM_LIBRARY_DIR"};
+my $HELM_LIB_DIR = $ENV{"HELM_LIB_DIR"};
# this should be the only fixed constant
-my $DEFAULT_HELM_LIBRARY_DIR = "@DEFAULT_HELM_LIBRARY_DIR@";
-if (defined ($HELM_LIBRARY_DIR)) {
- $HELM_LIBRARY_PATH = $HELM_LIBRARY_DIR."/configuration.pl";
+my $DEFAULT_HELM_LIB_DIR = "@HELM_LIB_DIR@";
+if (defined ($HELM_LIB_DIR)) {
+ $HELM_LIB_PATH = $HELM_LIB_DIR."/configuration.pl";
} else {
- $HELM_LIBRARY_PATH = $DEFAULT_HELM_LIBRARY_DIR."/configuration.pl";
+ $HELM_LIB_PATH = $DEFAULT_HELM_LIB_DIR."/configuration.pl";
+}
+
+# : TODO temporary, move this setting to configuration file
+# 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";
}
+#
+
# next require defines: $helm_dir, $html_link, $dtd_dir, $uris_dbm
-require $HELM_LIBRARY_PATH;
+require $HELM_LIB_PATH;
+
+# Let's override the configuration file
+$style_dir = $ENV{"HELM_STYLE_DIR"} if (defined ($ENV{"HELM_STYLE_DIR"}));
+$dtd_dir = $ENV{"HELM_DTD_DIR"} if (defined ($ENV{"HELM_DTD_DIR"}));
use HTTP::Daemon;
use HTTP::Status;
@@ -42,17 +58,32 @@ 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, ">\n";
+print "Please contact me at: \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";
+
$SIG{CHLD} = "IGNORE"; # do not accumulate defunct processes
$SIG{USR1} = \&update; # sent by the child to make the parent update
while (my $c = $d->accept) {
@@ -60,11 +91,15 @@ while (my $c = $d->accept) {
while (my $r = $c->get_request) {
#CSC: mancano i controlli di sicurezza
- my $inputuri = $r->url;
+ 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");
+
+ print "\nUnescaped query: ".$http_query."\n";
if ($http_method eq 'GET' and $http_path eq "/getciconly") {
# finds the uri, url and filename
@@ -73,21 +108,19 @@ while (my $c = $d->accept) {
my $cicfilename = $cicuri;
$cicfilename =~ s/cic:(.*)/$1/;
$cicfilename =~ s/theory:(.*)/$1/;
-# $cicfilename = $helm_dir.$cicfilename.".xml";
-#
+
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};
-#
if (!defined($cicurl)) {
print "\nNOT FOUND!!!!!\n";
$c->send_error(RC_NOT_FOUND)
@@ -127,45 +160,41 @@ while (my $c = $d->accept) {
$cicfilename =~ s/theory:(.*)/$1/;
$cicfilename = $helm_dir.$cicfilename;
-# my $typesfilename = $cicfilename.".types.xml" if $typesuri;
-# my $annfilename = $cicfilename.$annsuffix.".xml" if $annuri;
-# $cicfilename .= ".xml";
-
-#
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 .= $cicext;
-#
-
if (!defined($cicurl) ||
(!defined($typesurl) && $typesuri) ||
@@ -208,14 +237,54 @@ EOT
print "DTD: $inputuri ==> ($filename)\n";
if (stat($filename)) {
print "Using local copy\n";
- open(FD, $filename);
+ open(FD, $filename) or die "Cannot open $filename\n";
$cont = "";
- while() { $cont .= $_; }
+ while() {
+ # Vary bad heuristic for substituion of absolute URLs
+ # for relative ones
+ s/ENTITY (.*) SYSTEM\s+"/ENTITY $1 SYSTEM "$myownurl\/getdtd?uri=/g;
+ $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() {
+ 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";
+ $cont = "";
+ while() {
+ # Vary bad heuristic for substituion of absolute URLs
+ # for relative ones
+ s/xsl:import\s+href="/xsl:import href="$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;
@@ -227,13 +296,52 @@ EOT
$cont = "$quoted_html_link";
answer($c,$cont);
} elsif ($http_method eq 'GET' and $http_path eq "/update") {
- print "Update requested...";
- update();
- kill(USR1,getppid());
+ # rebuild urls_of_uris.db
+ print "Update requested...\n";
+ mk_urls_of_uris();
+ kill(USR1,getppid()); # signal changes to parent
print " done\n";
answer($c,"
Update done
");
+ } 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,"
Invalid output type, may be ".
+ "\"txt\" or \"xml\"
");
+ } 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,"
Invalid uri , may begin with ".
+ "\"cic:\", \"theory:\" or \"*:\"