2 * Copyright (C) 2003, 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_debugger;;
30 let fold_file f init fname =
31 let inchan = open_in fname in
32 let rec fold_lines' value =
34 let line = input_line inchan in
35 fold_lines' (f value line)
36 with End_of_file -> value
38 let res = (try fold_lines' init with e -> (close_in inchan; raise e)) in
41 let iter_file f = fold_file (fun _ line -> f line) ()
43 let hashtbl_sorted_fold f tbl init =
45 List.sort compare (Hashtbl.fold (fun key _ keys -> key::keys) tbl [])
47 List.fold_left (fun acc k -> f k (Hashtbl.find tbl k) acc) init sorted_keys
51 ~stdout:Shell.to_dev_null ~stderr:Shell.to_dev_null
52 [Shell.cmd "cp" [src; dst]]
54 let file_scheme_RE = Pcre.regexp "^file://"
55 let wget ?output url =
58 (match output with Some file -> ["-O"; file] | None -> []) @ [url]
60 debug_print ("wget " ^ String.concat " " flags);
62 ~stdout:Shell.to_dev_null ~stderr:Shell.to_dev_null
63 [Shell.cmd "wget" flags]
65 if Pcre.pmatch ~rex:file_scheme_RE url then begin (* file:// URL *)
66 let src_fname = Pcre.replace ~rex:file_scheme_RE url in
68 | Some dst_fname -> cp src_fname dst_fname
70 let dst_fname = Filename.basename src_fname in
71 if src_fname <> dst_fname then
72 cp src_fname dst_fname
73 else (* src and dst are the same: do nothing *)
75 end else (* other URL, pass it to wget *)
78 (* stderr shown as usual *)
79 let gzip ?(keep = false) fname =
80 if keep then begin (* keep original file *)
81 debug_print ("gzip -f -c " ^ fname);
83 ~stdout:(Shell.to_file (fname ^ ".gz"))
84 [Shell.cmd "gzip" ["-f"; "-c"; fname]]
85 end else begin (* don't keep original file *)
86 debug_print ("gzip -f " ^ fname);
87 Shell.call [Shell.cmd "gzip" ["-f"; fname]]
90 (* stderr shown as usual *)
91 let gunzip ?(keep = false) fname =
92 if not (Pcre.pmatch ~pat:"\\.gz$" fname) then
93 failwith "gunzip: source file doesn't end with '.gz'";
94 let basename = Pcre.replace ~pat:"\\.gz$" fname in
95 if keep then begin (* keep original file *)
96 debug_print ("gunzip -f -c " ^ fname);
98 ~stdout:(Shell.to_file basename)
99 [Shell.cmd "gunzip" ["-f"; "-c"; fname]]
100 end else begin (* don't keep original file *)
101 debug_print ("gunzip -f " ^ fname);
102 Shell.call [Shell.cmd "gunzip" ["-f"; fname]]
106 let buf = Buffer.create 28 in (* strlen("/tmp/fileSzb3Mw_http_getter") *)
108 ~stdout:(Shell.to_buffer buf)
109 [Shell.cmd "tempfile" ["--suffix=_http_getter"]];
110 Pcre.replace ~pat:"\n" (Buffer.contents buf)
112 let mkdir ?(parents = false) dirname =
113 if not (Sys.file_exists dirname) then begin
114 let flags = if parents then ["-p"; dirname] else [dirname] in
115 debug_print ("mkdir " ^ String.concat " " flags);
116 Shell.call [Shell.cmd "mkdir" flags]
119 let string_of_proc_status = function
120 | Unix.WEXITED code -> sprintf "[Exited: %d]" code
121 | Unix.WSIGNALED sg -> sprintf "[Killed: %d]" sg
122 | Unix.WSTOPPED sg -> sprintf "[Stopped: %d]" sg
125 if Pcre.pmatch ~rex:file_scheme_RE url then begin
126 (* file:// URL. Read data from file system *)
127 let fname = Pcre.replace ~rex:file_scheme_RE url in
129 let size = (Unix.stat fname).Unix.st_size in
130 let buf = String.create size in
131 let ic = open_in fname in
132 really_input ic buf 0 size;
135 with Unix.Unix_error (Unix.ENOENT, "stat", _) -> None
136 end else (* other URL, pass it to netclient *)
138 Some (Http_client.Convenience.http_get url)
139 with Http_client.Http_error (code, _) -> None