2 * Copyright (C) 2000, HELM Team.
4 * This file is part of HELM, an Hypertextual, Electronic
5 * Library of Mathematics, developed at the Computer Science
6 * Department, University of Bologna, Italy.
8 * HELM is free software; you can redistribute it and/or
9 * modify it under the terms of the GNU General Public License
10 * as published by the Free Software Foundation; either version 2
11 * of the License, or (at your option) any later version.
13 * HELM is distributed in the hope that it will be useful,
14 * but WITHOUT ANY WARRANTY; without even the implied warranty of
15 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
16 * GNU General Public License for more details.
18 * You should have received a copy of the GNU General Public License
19 * along with HELM; if not, write to the Free Software
20 * Foundation, Inc., 59 Temple Place - Suite 330, Boston,
23 * For details, see the HELM World-Wide-Web page,
24 * http://cs.unibo.it/helm/.
27 open Http_getter_types;;
30 let string_of_ls_flag = function No -> "NO" | Yes -> "YES" | Ann -> "ANN"
31 let string_of_encoding = function
32 | Enc_normal -> "Normal"
33 | Enc_gzipped -> "GZipped"
35 let is_cic_uri uri = Pcre.pmatch ~pat:"^cic:" uri
36 let is_theory_uri uri = Pcre.pmatch ~pat:"^theory:" uri
37 let is_xml_uri uri = is_cic_uri uri || is_theory_uri uri
38 let is_rdf_uri uri = Pcre.pmatch ~pat:"^helm:rdf(.*):(.*)//(.*)" uri
39 let is_xsl_uri uri = Pcre.pmatch ~pat:"^\\w+\\.xsl" uri
41 let rec http_getter_uri_of_string = function
42 | uri when is_rdf_uri uri ->
43 (match Pcre.split ~pat:"//" uri with
46 match http_getter_uri_of_string uri with
47 | Xml_uri xmluri -> xmluri
48 | _ -> raise (Http_getter_invalid_URI uri)
50 Rdf_uri (prefix, rest)
51 | _ -> raise (Http_getter_invalid_URI uri))
52 | uri when is_cic_uri uri -> Xml_uri (Cic (Pcre.replace ~pat:"^cic:" uri))
53 | uri when is_theory_uri uri ->
54 Xml_uri (Theory (Pcre.replace ~pat:"^theory:" uri))
55 | uri -> raise (Http_getter_invalid_URI uri)
59 ~pat:(sprintf "DOCTYPE (.*) SYSTEM\\s+\"%s/" Http_getter_env.dtd_base_url)
61 sprintf "DOCTYPE $1 SYSTEM \"%s/getdtd?uri=" Http_getter_env.my_own_url)
64 let mk_patch_fun tag line =
66 ~pat:(sprintf "%s\\s+href=\"" tag)
68 sprintf "%s href=\"%s/getxslt?uri=" tag Http_getter_env.my_own_url)
71 let (patch_import, patch_include) =
72 (mk_patch_fun "xsl:import", mk_patch_fun "xsl:include")
74 fun line -> patch_include (patch_import line)
77 ~pat:"ENTITY (.*) SYSTEM\\s+\""
79 sprintf "ENTITY $1 SYSTEM \"%s/getdtd?uri=" Http_getter_env.my_own_url)
83 sprintf "<html><body><h1>Http Getter error: %s</h1></body></html>" s
84 let pp_internal_error s =
85 sprintf "<html><body><h1>Http Getter Internal error: %s</h1></body></html>" s
86 let pp_msg s = sprintf "<html><body><h1>%s</h1></body></html>" s
89 let mk_return_fun pp_fun contype msg outchan =
91 ~body:(pp_fun msg) ~headers:["Content-Type", contype] outchan
93 let return_html_error = mk_return_fun pp_error "text/html"
94 let return_html_internal_error = mk_return_fun pp_internal_error "text/html"
95 let return_html_msg = mk_return_fun pp_msg "text/html"
96 let return_xml_msg = mk_return_fun null_pp "text/xml"
97 let return_file ~fname ?contype ?contenc ?(patch_fun = fun x -> x) outchan =
99 match (contype, contenc) with
100 | (Some t, Some e) -> [ "Content-Type", t; "Content-Enconding", e ]
101 | (Some t, None) -> [ "Content-Type" , t ]
102 | (None, Some e) -> [ "Content-Enconding", e ]
105 Http_daemon.send_basic_headers ~code:200 outchan;
106 Http_daemon.send_headers headers outchan;
107 Http_daemon.send_CRLF outchan;
108 Http_getter_misc.iter_file
109 (fun line -> output_string outchan (patch_fun line ^ "\n"))
111 let return_400 body outchan = Http_daemon.respond_error ~code:400 ~body outchan
113 let wget ?output url =
115 (match output with Some file -> ["-O"; file] | None -> []) @ [url]
118 ~stdout:Shell.to_dev_null ~stderr:Shell.to_dev_null
119 [Shell.cmd "wget" flags]
121 (* TODO gzip and gunzip create executables file, but umask seems to be
122 correctly inherited from the shell .... boh *)
124 (* stderr shown as usual *)
125 let gzip ?(keep = false) fname =
126 if keep then (* keep original file *)
128 ~stdout:(Shell.to_file (fname ^ ".gz"))
129 [Shell.cmd "gzip" ["-f"; "-c"; fname]]
130 else (* don't keep original file *)
131 Shell.call [Shell.cmd "gzip" ["-f"; fname]]
133 (* stderr shown as usual *)
134 let gunzip ?(keep = false) fname =
135 if not (Pcre.pmatch ~pat:"\\.gz$" fname) then
136 failwith "gunzip: source file doesn't end with '.gz'";
137 let basename = Pcre.replace ~pat:"\\.gz$" fname in
138 if keep then (* keep original file *)
140 ~stdout:(Shell.to_file basename)
141 [Shell.cmd "gunzip" ["-f"; "-c"; fname]]
142 else (* don't keep original file *)
143 Shell.call [Shell.cmd "gunzip" ["-f"; fname]]
146 let buf = Buffer.create 28 in (* strlen("/tmp/fileSzb3Mw_http_getter") *)
148 ~stdout:(Shell.to_buffer buf)
149 [Shell.cmd "tempfile" ["--suffix=_http_getter"]];
150 Pcre.replace ~pat:"\n" (Buffer.contents buf)