1 (* COCANWIKI - a wiki written in Objective CAML.
2 * Written by Richard W.M. Jones <rich@merjis.com>.
3 * Copyright (C) 2004 Merjis Ltd.
4 * $Id: page.ml,v 1.22 2004/09/27 12:37:54 rich Exp $
6 * This program is free software; you can redistribute it and/or modify
7 * it under the terms of the GNU General Public License as published by
8 * the Free Software Foundation; either version 2 of the License, or
9 * (at your option) any later version.
11 * This program is distributed in the hope that it will be useful,
12 * but WITHOUT ANY WARRANTY; without even the implied warranty of
13 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
14 * GNU General Public License for more details.
16 * You should have received a copy of the GNU General Public License
17 * along with this program; see the file COPYING. If not, write to
18 * the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
19 * Boston, MA 02111-1307, USA.
30 open Cocanwiki_template
33 open Cocanwiki_server_settings
35 (* Maximum level of redirection. *)
38 type fp_status = FPOK of int * string * string * Dbi.datetime * bool
39 | FPRedirect of string
42 (* Referer strings which help us decide if the user came from
43 * a search engine and highlight terms in the page appropriately.
45 let search_engines = [
46 Pcre.regexp "^http://.*google\\.", [ "q"; "as_q"; "as_epq"; "as_oq" ];
47 Pcre.regexp "^http://.*yahoo\\.", [ "p" ];
48 Pcre.regexp "^http://.*msn\\.", [ "q"; "MT" ]
50 let split_words = Pcre.regexp "\\W+"
52 let xhtml_re = Pcre.regexp "<.*?>|[^<>]+"
54 let run r (q : cgi) (dbh : Dbi.connection) hostid {edit_anon=edit_anon} user =
55 let template_page = get_template dbh hostid "page.html" in
56 let template_404 = get_template dbh hostid "page_404.html" in
58 let page = q#param "page" in
59 let page = if page = "" then "index" else page in
61 (* Host-specific fields. *)
62 let sth = dbh#prepare_cached "select css is not null,
63 feedback_email is not null,
65 from hosts where id = ?" in
66 sth#execute [`Int hostid];
67 let has_host_css, has_feedback_email, mailing_list =
68 match sth#fetch1 () with
69 | [ `Bool has_host_css; `Bool has_feedback_email; `Bool mailing_list ] ->
70 has_host_css, has_feedback_email, mailing_list
71 | _ -> assert false in
73 (* Can the user edit? Manage users? etc. *)
74 let can_edit = can_edit edit_anon user in
75 let can_manage_users = can_manage_users user in
76 let can_manage_contacts = can_manage_contacts user in
77 let can_manage_site = can_manage_site user in
78 let can_edit_global_css = can_edit_global_css user in
80 (* Do we have a stats page set up? *)
81 let has_stats = server_settings_stats_page dbh <> None in
83 (* Given the referer string, return the list of search terms. If none
84 * can be found, then throws Not_found.
86 let search_terms_from_referer referer =
88 List.find (fun (rex, _) -> Pcre.pmatch ~rex referer) search_engines in
89 let args = Cgi_args.parse referer in
91 List.find (fun argname -> List.mem_assoc argname args) argnames in
92 let search_string = List.assoc argname args in
93 Pcre.split ~rex:split_words search_string
96 (* Given a full page of XHTML, highlight search terms found in the
97 * <body> part of the page.
99 let highlight_search_terms xhtml search_terms span_class =
100 (* Split the original XHTML into strings and tags. For example if
101 * the original string is: "This is some <b>bold</b> text.<br/>", then
102 * after this step we will have the following list:
103 * [ "This is some "; "<b>"; "bold"; "</b>"; " text."; "<br/>" ]
105 let xhtml = Pcre.extract_all ~rex:xhtml_re xhtml in
106 let xhtml = Array.to_list xhtml in
107 let xhtml = List.map (fun matches -> matches.(0)) xhtml in
109 (* Find the <body> ... </body> tags. We only want to apply
110 * highlighting to tags within this area.
112 let rec list_split f acc = function
113 | [] -> List.rev acc, []
114 | ((x :: _) as xs) when f x -> List.rev acc, xs
116 let acc = x :: acc in
120 list_split (fun str -> String.starts_with str "<body") [] xhtml in
122 list_split ((=) "</body>") [] body in
123 (* NB: Hopefully, xhtml = head @ body @ tail. *)
125 (* The search terms are a list of simple words. Turn into a big
126 * regular expression, because we want to substitute for each. We
127 * end up with a regexp like '(word1|word2|word3)'.
130 Pcre.regexp ~flags:[`CASELESS]
131 ("(" ^ String.concat "|" search_terms ^ ")") in
133 (* Do the substitution, but only on text, not elements! *)
136 "<span class=\"" ^ span_class ^ "\">" ^ text ^ "</span>"
139 if String.length str > 0 && str.[0] != '<' then
140 Pcre.substitute ~rex ~subst str
144 (* Join the XHTML fragments back together again. *)
145 String.concat "" (List.concat [ head ; body ; tail ])
148 (* This code generates ordinary pages. *)
149 let make_page title description pageid last_modified_date has_page_css
151 let t = template_page in
153 t#set "description" description;
154 t#set "pageid" (string_of_int pageid);
155 t#set "last_modified_date" (printable_date last_modified_date);
157 if page <> page' then (* redirection *) (
159 t#set "original_page" page; (* XXX title - get it from database *)
160 t#conditional "redirected" true
163 t#conditional "redirected" false
166 t#conditional "has_host_css" has_host_css;
167 t#conditional "has_page_css" has_page_css;
169 t#conditional "has_feedback_email" has_feedback_email;
170 t#conditional "mailing_list" mailing_list;
172 t#conditional "can_edit" can_edit;
173 t#conditional "can_manage_users" can_manage_users;
174 t#conditional "can_manage_contacts" can_manage_contacts;
175 t#conditional "can_manage_site" can_manage_site;
176 t#conditional "can_edit_global_css" can_edit_global_css;
178 t#conditional "has_stats" has_stats;
180 (* Pull out the sections in this page. *)
181 let sth = dbh#prepare_cached
182 "select ordering, sectionname, content, divname
185 order by ordering" in
186 sth#execute [`Int pageid];
190 (function [`Int ordering;
191 (`Null | `String _) as sectionname;
193 (`Null | `String _) as divname] ->
194 let divname, has_divname =
197 | `String divname -> divname, true in
198 let sectionname, has_sectionname =
199 match sectionname with
201 | `String sectionname -> sectionname, true in
202 let linkname = linkname_of_sectionname sectionname in
203 [ "ordering", Template.VarString (string_of_int ordering);
204 "has_sectionname", Template.VarConditional has_sectionname;
205 "sectionname", Template.VarString sectionname;
206 "linkname", Template.VarString linkname;
209 (Wikilib.xhtml_of_content dbh hostid content);
210 "has_divname", Template.VarConditional has_divname;
211 "divname", Template.VarString divname ]
212 | _ -> assert false) in
214 t#table "sections" sections;
216 (* Are we showing an old version of the page? If so, warn. *)
219 t#conditional "is_old_version" false
221 t#conditional "is_old_version" true;
222 t#set "old_version" (string_of_int pageid));
227 t#conditional "user_logged_in" false
228 | User (_, username, _) ->
229 t#conditional "user_logged_in" true;
230 t#set "username" username);
232 (* If we are coming from a search engine then we want to highlight
233 * search terms throughout the whole page ...
236 let referer = Table.get (Request.headers_in r) "Referer" in
237 let search_terms = search_terms_from_referer referer in
239 (* Highlight the search terms. *)
240 let xhtml = t#to_string in
241 let xhtml = highlight_search_terms xhtml search_terms "search_term" in
243 (* Deliver the page. *)
248 (* No referer / no search terms / not a search engine referer. *)
252 (* This code generates 404 pages. *)
254 Request.set_status r 404; (* Return a 404 error code. *)
256 let t = template_404 in
262 ('a'..'z' | 'A'..'Z' | '0'..'9') as c -> c
265 t#set "search_terms" search_terms;
267 t#conditional "has_host_css" has_host_css;
269 t#conditional "can_edit" can_edit;
270 t#conditional "can_manage_users" can_manage_users;
271 t#conditional "can_manage_contacts" can_manage_contacts;
272 t#conditional "can_manage_site" can_manage_site;
273 t#conditional "can_edit_global_css" can_edit_global_css;
275 t#conditional "has_stats" has_stats;
280 (* Fetch a page by name. This function can give three answers:
281 * (1) Page fetched OK (fetches some details of the page).
282 * (2) Page is a redirect (fetches the name of the redirect page).
283 * (3) Page not found in database, ie. 404 error.
285 (* XXX Should do a case-insensitive matching of URLs, and if the URL differs
286 * in case only should redirect to the lowercase version.
288 let fetch_page page version allow_redirect =
291 if allow_redirect then (
294 "select redirect, id, title, description, last_modified_date,
296 from pages where hostid = ? and url = ?" in
297 sth#execute [`Int hostid; `String page];
299 (match sth#fetch1 () with
300 | [ `Null; `Int id; `String title; `String description;
301 `Timestamp last_modified_date; `Bool has_page_css ] ->
302 FPOK (id, title, description, last_modified_date,
304 | `String redirect :: _ ->
308 Not_found -> FPNotFound)
309 ) else (* redirects not allowed ... *) (
312 "select id, title, description, last_modified_date,
314 from pages where hostid = ? and url = ?" in
315 sth#execute [`Int hostid; `String page];
317 (match sth#fetch1 () with
318 | [ `Int id; `String title; `String description;
319 `Timestamp last_modified_date; `Bool has_page_css ] ->
320 FPOK (id, title, description, last_modified_date,
324 Not_found -> FPNotFound)
329 "select id, title, description, last_modified_date,
332 where hostid = ? and id = ? and
333 (url = ? or url_deleted = ?)" in
334 sth#execute [`Int hostid; `Int version;
335 `String page; `String page];
337 (match sth#fetch1 () with
338 | [ `Int id; `String title; `String description;
339 `Timestamp last_modified_date; `Bool has_page_css ] ->
340 FPOK (id, title, description, last_modified_date,
344 Not_found -> FPNotFound)
347 (* Here we deal with the complex business of redirects and versions. *)
348 (* Only allow the no_redirect and version syntax for editors. *)
349 let allow_redirect, version =
351 not (q#param_true "no_redirect"),
352 try Some (int_of_string (q#param "version")) with Not_found -> None
356 let rec loop page' i =
357 if i > max_redirect then (
358 error ~title:"Too many redirections" ~back_button:true
359 q ("Too many redirects between pages. This may happen because " ^
360 "of a cycle of redirections.");
363 match fetch_page page' version allow_redirect with
364 | FPOK (pageid, title, description, last_modified_date, has_page_css)->
365 make_page title description pageid last_modified_date has_page_css
367 | FPRedirect page' ->