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/
29 open Http_getter_debugger;;
32 let fold_file f init fname =
33 let inchan = open_in fname in
34 let rec fold_lines' value =
36 let line = input_line inchan in
37 fold_lines' (f value line)
38 with End_of_file -> value
40 let res = (try fold_lines' init with e -> (close_in inchan; raise e)) in
43 let iter_file f = fold_file (fun _ line -> f line) ()
45 let hashtbl_sorted_fold f tbl init =
47 List.sort compare (Hashtbl.fold (fun key _ keys -> key::keys) tbl [])
49 List.fold_left (fun acc k -> f k (Hashtbl.find tbl k) acc) init sorted_keys
53 ~stdout:Shell.to_dev_null ~stderr:Shell.to_dev_null
54 [Shell.cmd "cp" [src; dst]]
56 let file_scheme_RE = Pcre.regexp "^file://"
57 let wget ?output url =
60 (match output with Some file -> ["-O"; file] | None -> []) @ [url]
62 debug_print ("wget " ^ String.concat " " flags);
64 ~stdout:Shell.to_dev_null ~stderr:Shell.to_dev_null
65 [Shell.cmd "wget" flags]
67 if Pcre.pmatch ~rex:file_scheme_RE url then begin (* file:// URL *)
68 let src_fname = Pcre.replace ~rex:file_scheme_RE url in
70 | Some dst_fname -> cp src_fname dst_fname
72 let dst_fname = Filename.basename src_fname in
73 if src_fname <> dst_fname then
74 cp src_fname dst_fname
75 else (* src and dst are the same: do nothing *)
77 end else (* other URL, pass it to wget *)
80 (* stderr shown as usual *)
81 let gzip ?(keep = false) fname =
82 if keep then begin (* keep original file *)
83 debug_print ("gzip -f -c " ^ fname);
85 ~stdout:(Shell.to_file (fname ^ ".gz"))
86 [Shell.cmd "gzip" ["-f"; "-c"; fname]]
87 end else begin (* don't keep original file *)
88 debug_print ("gzip -f " ^ fname);
89 Shell.call [Shell.cmd "gzip" ["-f"; fname]]
92 (* stderr shown as usual *)
93 let gunzip ?(keep = false) fname =
94 if not (Pcre.pmatch ~pat:"\\.gz$" fname) then
95 failwith "gunzip: source file doesn't end with '.gz'";
96 let basename = Pcre.replace ~pat:"\\.gz$" fname in
97 if keep then begin (* keep original file *)
98 debug_print ("gunzip -f -c " ^ fname);
100 ~stdout:(Shell.to_file basename)
101 [Shell.cmd "gunzip" ["-f"; "-c"; fname]]
102 end else begin (* don't keep original file *)
103 debug_print ("gunzip -f " ^ fname);
104 Shell.call [Shell.cmd "gunzip" ["-f"; fname]]
108 let buf = Buffer.create 28 in (* strlen("/tmp/fileSzb3Mw_http_getter") *)
110 ~stdout:(Shell.to_buffer buf)
111 [Shell.cmd "tempfile" ["--suffix=_http_getter"]];
112 Pcre.replace ~pat:"\n" (Buffer.contents buf)
114 let mkdir ?(parents = false) dirname =
115 if not (Sys.file_exists dirname) then begin
116 let flags = if parents then ["-p"; dirname] else [dirname] in
117 debug_print ("mkdir " ^ String.concat " " flags);
118 Shell.call [Shell.cmd "mkdir" flags]
121 let string_of_proc_status = function
122 | Unix.WEXITED code -> sprintf "[Exited: %d]" code
123 | Unix.WSIGNALED sg -> sprintf "[Killed: %d]" sg
124 | Unix.WSTOPPED sg -> sprintf "[Stopped: %d]" sg
127 if Pcre.pmatch ~rex:file_scheme_RE url then begin
128 (* file:// URL. Read data from file system *)
129 let fname = Pcre.replace ~rex:file_scheme_RE url in
131 let size = (Unix.stat fname).Unix.st_size in
132 let buf = String.create size in
133 let ic = open_in fname in
134 really_input ic buf 0 size;
137 with Unix.Unix_error (Unix.ENOENT, "stat", _) -> None
138 end else (* other URL, pass it to netclient *)
140 Some (Http_client.Convenience.http_get url)
141 with Http_client.Http_error (code, _) -> None