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_debugger
33 let trailing_dot_gz_RE = Pcre.regexp "\\.gz$" (* for g{,un}zip *)
34 let url_RE = Pcre.regexp "^([\\w.-]+)(:(\\d+))?(/.*)?$"
35 let http_scheme_RE = Pcre.regexp ~flags:[`CASELESS] "^http://"
36 let file_scheme_RE = Pcre.regexp ~flags:[`CASELESS] "^file://"
37 let dir_sep_RE = Pcre.regexp "/"
38 let heading_slash_RE = Pcre.regexp "^/"
40 let bufsiz = 16384 (* for file system I/O *)
41 let tcp_bufsiz = 4096 (* for TCP I/O *)
43 let fold_file f init fname =
44 let ic = open_in fname in
46 let line = try Some (input_line ic) with End_of_file -> None in
49 | Some line -> aux (f line acc)
51 let res = try aux init with e -> close_in ic; raise e in
55 let iter_file f = fold_file (fun line _ -> f line) ()
57 let hashtbl_sorted_fold f tbl init =
59 List.sort compare (Hashtbl.fold (fun key _ keys -> key::keys) tbl [])
61 List.fold_left (fun acc k -> f k (Hashtbl.find tbl k) acc) init sorted_keys
63 let hashtbl_sorted_iter f tbl =
65 List.sort compare (Hashtbl.fold (fun key _ keys -> key::keys) tbl [])
67 List.iter (fun k -> f k (Hashtbl.find tbl k)) sorted_keys
70 let (ic, oc) = (open_in src, open_out dst) in
71 let buf = String.create bufsiz in
74 let bytes = input ic buf 0 bufsiz in
75 if bytes = 0 then raise End_of_file else output oc buf 0 bytes
77 with End_of_file -> ());
78 close_in ic; close_out oc
80 let wget ?output url =
82 (sprintf "wgetting %s (output: %s)" url
83 (match output with None -> "default" | Some f -> f));
85 | url when Pcre.pmatch ~rex:file_scheme_RE url -> (* file:// *)
86 (let src_fname = Pcre.replace ~rex:file_scheme_RE url in
88 | Some dst_fname -> cp src_fname dst_fname
90 let dst_fname = Filename.basename src_fname in
91 if src_fname <> dst_fname then
92 cp src_fname dst_fname
93 else (* src and dst are the same: do nothing *)
95 | url when Pcre.pmatch ~rex:http_scheme_RE url -> (* http:// *)
97 open_out (match output with Some f -> f | None -> Filename.basename url)
99 Http_client.http_get_iter (fun data -> output_string oc data) url;
101 | scheme -> (* unsupported scheme *)
102 failwith ("Http_getter_misc.wget: unsupported scheme: " ^ scheme)
104 let gzip ?(keep = false) ?output fname =
105 let output = match output with None -> fname ^ ".gz" | Some fname -> fname in
106 debug_print (sprintf "gzipping %s (keep: %b, output: %s)" fname keep output);
107 let (ic, oc) = (open_in fname, Gzip.open_out output) in
108 let buf = String.create bufsiz in
111 let bytes = input ic buf 0 bufsiz in
112 if bytes = 0 then raise End_of_file else Gzip.output oc buf 0 bytes
114 with End_of_file -> ());
115 close_in ic; Gzip.close_out oc;
116 if not keep then Sys.remove fname
119 let gunzip ?(keep = false) ?output fname =
120 (* assumption: given file name ends with ".gz" or output is set *)
124 if (Pcre.pmatch ~rex:trailing_dot_gz_RE fname) then
125 Pcre.replace ~rex:trailing_dot_gz_RE fname
128 "Http_getter_misc.gunzip: unable to determine output file name"
129 | Some fname -> fname
131 debug_print (sprintf "gunzipping %s (keep: %b, output: %s)"
133 let (ic, oc) = (Gzip.open_in fname, open_out output) in
134 let buf = String.create bufsiz in
137 let bytes = Gzip.input ic buf 0 bufsiz in
138 if bytes = 0 then raise End_of_file else Pervasives.output oc buf 0 bytes
140 with End_of_file -> ());
141 Gzip.close_in ic; close_out oc;
142 if not keep then Sys.remove fname
145 let tempfile () = Filename.temp_file "http_getter_" ""
147 exception Mkdir_failure of string * string;; (* dirname, failure reason *)
150 let mkdir ?(parents = false) dirname =
153 let split = Pcre.split ~rex:dir_sep_RE dirname in
154 if Pcre.pmatch ~rex:heading_slash_RE dirname then
163 sprintf "%s%s%s" pre (match pre with "/" | "" -> "" | _ -> "/") dir
166 (match (Unix.stat next_dir).Unix.st_kind with
167 | Unix.S_DIR -> () (* dir component already exists, go on! *)
168 | _ -> (* dir component already exists but isn't a dir, abort! *)
170 (Mkdir_failure (dirname,
171 sprintf "'%s' already exists but is not a dir" next_dir)))
172 with Unix.Unix_error (Unix.ENOENT, "stat", _) ->
173 (* dir component doesn't exists, create it and go on! *)
174 Unix.mkdir next_dir dir_perm);
178 if parents then mkdirhier () else Unix.mkdir dirname dir_perm
180 let string_of_proc_status = function
181 | Unix.WEXITED code -> sprintf "[Exited: %d]" code
182 | Unix.WSIGNALED sg -> sprintf "[Killed: %d]" sg
183 | Unix.WSTOPPED sg -> sprintf "[Stopped: %d]" sg
186 if Pcre.pmatch ~rex:file_scheme_RE url then begin
187 (* file:// URL. Read data from file system *)
188 let fname = Pcre.replace ~rex:file_scheme_RE url in
190 let size = (Unix.stat fname).Unix.st_size in
191 let buf = String.create size in
192 let ic = open_in fname in
193 really_input ic buf 0 size;
196 with Unix.Unix_error (Unix.ENOENT, "stat", _) -> None
197 end else (* other URL, pass it to Http_client *)
199 Some (Http_client.http_get url)
201 prerr_endline (sprintf
202 "Warning: Http_client failed on url %s with exception: %s"
203 url (Printexc.to_string e));
207 let blank_line_RE = Pcre.regexp "(^#)|(^\\s*$)" in
209 Pcre.pmatch ~rex:blank_line_RE line