2 * Copyright (C) 2003-2004:
3 * Stefano Zacchiroli <zack@cs.unibo.it>
4 * for the HELM Team http://helm.cs.unibo.it/
6 * This file is part of HELM, an Hypertextual, Electronic
7 * Library of Mathematics, developed at the Computer Science
8 * Department, University of Bologna, Italy.
10 * HELM is free software; you can redistribute it and/or
11 * modify it under the terms of the GNU General Public License
12 * as published by the Free Software Foundation; either version 2
13 * of the License, or (at your option) any later version.
15 * HELM is distributed in the hope that it will be useful,
16 * but WITHOUT ANY WARRANTY; without even the implied warranty of
17 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
18 * GNU General Public License for more details.
20 * You should have received a copy of the GNU General Public License
21 * along with HELM; if not, write to the Free Software
22 * Foundation, Inc., 59 Temple Place - Suite 330, Boston,
25 * For details, see the HELM World-Wide-Web page,
26 * http://helm.cs.unibo.it/
31 open Http_getter_common
33 open Http_getter_types
34 open Http_getter_debugger
38 let configuration_file = "/projects/helm/etc/http_getter.conf.xml"
40 let common_headers = [
41 "Cache-Control", "no-cache";
46 (* HTTP queries argument parsing *)
48 (* parse encoding ("format" parameter), default is Enc_normal *)
49 let parse_enc (req: Http_types.request) =
51 (match req#param "format" with
52 | "normal" -> Enc_normal
54 | s -> raise (Bad_request ("Invalid format: " ^ s)))
55 with Http_types.Param_not_found _ -> Enc_normal
57 (* parse "patch_dtd" parameter, default is true *)
58 let parse_patch (req: Http_types.request) =
60 (match req#param "patch_dtd" with
61 | s when String.lowercase s = "yes" -> true
62 | s when String.lowercase s = "no" -> false
63 | s -> raise (Bad_request ("Invalid patch_dtd value: " ^ s)))
64 with Http_types.Param_not_found _ -> true
66 (* parse output format ("format" parameter), no default value *)
67 let parse_output_format meth (req: Http_types.request) =
68 match req#param "format" with
69 | s when String.lowercase s = "txt" -> Fmt_text
70 | s when String.lowercase s = "xml" -> Fmt_xml
71 | s -> raise (Bad_request ("Invalid /" ^ meth ^ " format: " ^ s))
73 (* parse "baseuri" format for /ls method, no default value *)
75 let parse_ls_RE = Pcre.regexp "^(\\w+):(.*)$" in
76 let trailing_slash_RE = Pcre.regexp "/+$" in
78 raise (Bad_request ("Invalid /ls baseuri: " ^ uri))
80 fun (req: Http_types.request) ->
81 let baseuri = req#param "baseuri" in
84 Pcre.extract ~rex:parse_ls_RE
85 (Pcre.replace ~rex:trailing_slash_RE baseuri)
87 (match (subs.(1), subs.(2)) with
88 | "cic", uri -> Cic uri
89 | "theory", uri -> Theory uri
90 | _ -> wrong_uri baseuri)
91 with Not_found -> wrong_uri baseuri
93 (* parse "position" argument, default is 0 *)
94 let parse_position (req: Http_types.request) =
96 let res = int_of_string (req#param "position") in
98 raise (Failure "int_of_string");
101 | Http_types.Param_not_found _ -> 0
102 | Failure "int_of_string" ->
104 (sprintf "position must be a non negative integer (%s given)"
105 (req#param "position")))
107 let parse_rdf_class (req: Http_types.request) =
108 match req#param "class" with
109 | "forward" -> `Forward
110 | "backward" -> `Backward
111 | c -> raise (Bad_request ("Invalid RDF class: " ^ c))
113 let mk_return_fun pp_fun contype msg outchan =
115 ~body:(pp_fun msg) ~headers:["Content-Type", contype] outchan
117 sprintf "<html><body>Http Getter error: %s</body></html>" s
118 let pp_internal_error s =
119 sprintf "<html><body>Http Getter Internal error: %s</body></html>" s
120 let pp_msg s = sprintf "<html><body>%s</body></html>" s
122 let return_html_error = mk_return_fun pp_error "text/html"
123 let return_html_internal_error = mk_return_fun pp_internal_error "text/html"
124 let return_html_msg = mk_return_fun pp_msg "text/html"
125 let return_html_raw = mk_return_fun null_pp "text/html"
126 let return_xml_raw = mk_return_fun null_pp "text/xml"
127 let return_400 body outchan = Http_daemon.respond_error ~code:400 ~body outchan
129 let return_all_foo_uris doctype uris outchan =
130 Http_daemon.send_basic_headers ~code:200 outchan;
131 Http_daemon.send_header "Content-Type" "text/xml" outchan;
132 Http_daemon.send_headers common_headers outchan;
133 Http_daemon.send_CRLF outchan;
137 "<?xml version=\"1.0\" encoding=\"ISO-8859-1\"?>
138 <!DOCTYPE %s SYSTEM \"%s/getdtd?uri=%s.dtd\">
143 (Lazy.force Http_getter_env.my_own_url)
147 (fun uri -> output_string outchan (sprintf "\t<uri value=\"%s\" />\n" uri))
149 output_string outchan (sprintf "</%s>\n" doctype)
151 let return_all_xml_uris fmt outchan =
152 let uris = Http_getter.getalluris () in
155 let buf = Buffer.create 10240 in
156 List.iter (bprintf buf "%s\n") uris ;
157 let body = Buffer.contents buf in
159 ~headers:(("Content-Type", "text/plain") :: common_headers)
162 return_all_foo_uris "alluris" uris outchan
164 let return_all_rdf_uris classs outchan =
165 return_all_foo_uris "allrdfuris" (Http_getter.getallrdfuris classs) outchan
167 let return_ls xmluri fmt outchan =
168 let ls_items = Http_getter.ls xmluri in
169 let buf = Buffer.create 10240 in
174 | Ls_section dir -> bprintf buf "dir, %s\n" dir
176 bprintf buf "object, %s, <%s,%s,%s,%s>\n"
177 obj.uri (if obj.ann then "YES" else "NO")
178 (string_of_ls_flag obj.types)
179 (string_of_ls_flag obj.body)
180 (string_of_ls_flag obj.proof_tree))
183 Buffer.add_string buf "<?xml version=\"1.0\" encoding=\"ISO-8859-1\"?>\n";
184 bprintf buf "<!DOCTYPE ls SYSTEM \"%s/getdtd?uri=ls.dtd\">\n"
185 (Lazy.force Http_getter_env.my_own_url);
186 Buffer.add_string buf "<ls>\n";
189 | Ls_section dir -> bprintf buf "<section>%s</section>\n" dir
192 "<object name=\"%s\">
193 \t<ann value=\"%s\" />
194 \t<types value=\"%s\" />
195 \t<body value=\"%s\" />
196 \t<proof_tree value=\"%s\" />
199 obj.uri (if obj.ann then "YES" else "NO")
200 (string_of_ls_flag obj.types)
201 (string_of_ls_flag obj.body)
202 (string_of_ls_flag obj.proof_tree))
204 Buffer.add_string buf "</ls>\n");
205 let body = Buffer.contents buf in
207 ~headers:(("Content-Type", "text/plain") :: common_headers)
210 let return_help outchan = return_html_raw (Http_getter.help ()) outchan
212 let return_resolve uri outchan =
215 (sprintf "<url value=\"%s\" />\n" (Http_getter.resolve uri))
217 with Unresolvable_URI uri ->
218 return_xml_raw "<unresolved />\n" outchan
220 let return_list_servers outchan =
222 (sprintf "<html><body><table>\n%s\n</table></body></html>"
225 (fun (pos, server) ->
226 sprintf "<tr><td>%d</td><td>%s</td></tr>" pos server)
227 (Http_getter.list_servers ()))))
230 let log_failure msg = debug_print ("Request not fulfilled: " ^ msg)
232 (** given an action (i.e. a function which expects a logger and do something
233 * using it as a logger), perform it sending its output incrementally to the
234 * given output channel. Response is sent embedded in an HTML document.
235 * Channel is closed afterwards. *)
236 let send_log_to ?prepend action outchan =
237 Http_daemon.send_basic_headers ~code:200 outchan;
238 Http_daemon.send_header "Content-Type" "text/html" outchan;
239 Http_daemon.send_CRLF outchan;
240 output_string outchan "<html><body>\n"; flush outchan;
243 | Some text -> output_string outchan text; flush outchan);
245 output_string outchan (HelmLogger.html_of_html_tag tag);
249 output_string outchan "\n</body></html>";
254 let callback (req: Http_types.request) outchan =
256 debug_print ("Connection from " ^ req#clientAddr);
257 debug_print ("Received request: " ^ req#path);
259 | "/help" -> return_help outchan
261 let uri = req#param "uri" in
262 Http_getter_cache.respond_xml ~url:(Http_getter.resolve uri) ~uri
263 ~enc:(parse_enc req) ~patch:(parse_patch req) outchan
265 Http_getter_cache.respond_xsl
266 ~url:(Http_getter.resolve (req#param "uri"))
267 ~patch:(parse_patch req) outchan
269 Http_getter_cache.respond_dtd ~patch:(parse_patch req)
270 ~url:(sprintf "%s/%s"
271 (Helm_registry.get "getter.dtd_dir") (req#param "uri"))
273 | "/resolve" -> return_resolve (req#param "uri") outchan
275 Http_getter.register ~uri:(req#param "uri") ~url:(req#param "url");
276 return_html_msg "Register done" outchan
278 Http_getter.clean_cache ();
279 return_html_msg "Done." outchan
281 Http_getter_env.reload (); (* reload servers list from servers file *)
282 send_log_to (fun logger -> Http_getter.update ~logger ()) outchan
283 | "/list_servers" -> return_list_servers outchan
285 let name = req#param "url" in
286 let position = parse_position req in
288 sprintf "Added server %s in position %d)<br />\n" name position
291 (fun logger -> Http_getter.add_server ~logger ~position name) outchan
292 | "/remove_server" ->
293 let position = parse_position req in
294 if not (Http_getter.has_server position) then
295 raise (Bad_request (sprintf "no server with position %d" position))
298 sprintf "Removed server at position %d<br />\n" position
301 (fun logger -> Http_getter.remove_server ~logger position) outchan
303 return_all_xml_uris (parse_output_format "getalluris" req) outchan
304 | "/getallrdfuris" -> return_all_rdf_uris (parse_rdf_class req) outchan
306 return_ls (parse_ls_uri req) (parse_output_format "ls" req) outchan
308 Http_daemon.respond ~body:Http_getter_const.empty_xml outchan
310 Http_daemon.respond_error ~status:(`Client_error `Bad_request) outchan);
311 debug_print "Done!\n"
313 | Http_types.Param_not_found attr_name ->
314 let msg = sprintf "Parameter '%s' is missing" attr_name in
316 return_400 msg outchan
319 return_html_error msg outchan
320 | Internal_error msg ->
322 return_html_internal_error msg outchan
323 | Shell.Subprocess_error l ->
327 sprintf "Command '%s' returned %s" cmd (string_of_proc_status code))
330 log_failure (String.concat ", " msgs);
331 return_html_internal_error (String.concat "<br />\n" msgs) outchan
333 let msg = "Uncaught exception: " ^ (Printexc.to_string exc) in
335 return_html_error msg outchan
340 Helm_registry.load_from configuration_file;
341 Http_getter_env.reload ();
342 print_string (Http_getter_env.env_to_string ());
344 at_exit Http_getter.close_maps;
345 Sys.catch_break true;
347 Http_daemon.start' ~mode:`Thread
348 ~timeout:(Some 600) ~port:(Helm_registry.get_int "getter.port")
350 with Sys.Break -> () (* 'close_maps' already registered with 'at_exit' *)