--- /dev/null
+(* COCANWIKI - a wiki written in Objective CAML.
+ * Written by Richard W.M. Jones <rich@merjis.com>.
+ * Copyright (C) 2004 Merjis Ltd.
+ * $Id: cocanwiki_cgi_args.ml,v 1.1 2004/09/24 10:44:55 rich Exp $
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; see the file COPYING. If not, write to
+ * the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
+ * Boston, MA 02111-1307, USA.
+ *
+ * This badly needs to become a standard part of the Cgi library.
+ *)
+
+let split1_re = Pcre.regexp "\\?"
+let split2_re = Pcre.regexp "&"
+let split3_re = Pcre.regexp "="
+
+let parse full_url =
+ let url, qs = match Pcre.split ~rex:split1_re ~max:2 full_url with
+ | [url] | [url;""] -> url, None
+ | [url;qs] -> url, Some qs
+ | _ -> assert false in
+ let args = match qs with
+ None -> []
+ | Some qs ->
+ let args = Pcre.split ~rex:split2_re qs in
+ let f arg =
+ match Pcre.split ~rex:split3_re ~max:2 arg with
+ | [] -> ("", "")
+ | [key] -> (key, "")
+ | [key;value] -> (key, value)
+ | _ -> assert false
+ in
+ List.map f args in
+ url, qs, args
(* COCANWIKI - a wiki written in Objective CAML.
* Written by Richard W.M. Jones <rich@merjis.com>.
* Copyright (C) 2004 Merjis Ltd.
- * $Id: page.ml,v 1.19 2004/09/23 15:16:21 rich Exp $
+ * $Id: page.ml,v 1.20 2004/09/24 10:44:55 rich Exp $
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
| FPRedirect of string
| FPNotFound
+(* Referer strings which help us decide if the user came from
+ * a search engine and highlight terms in the page appropriately.
+ *)
+let search_engines = [
+ Pcre.regexp "^http://.*google\\.", [ "q"; "as_q"; "as_epq"; "as_oq" ];
+ Pcre.regexp "^http://.*yahoo\\.", [ "p" ];
+ Pcre.regexp "^http://.*msn\\.", [ "q"; "MT" ]
+]
+let split_words = Pcre.regexp "\\W+"
+
+let xhtml_re = Pcre.regexp "<.*?>|[^<>]+"
+
let run r (q : cgi) (dbh : Dbi.connection) hostid {edit_anon=edit_anon} user =
let template_page = get_template dbh hostid "page.html" in
let template_404 = get_template dbh hostid "page_404.html" in
(* Do we have a stats page set up? *)
let has_stats = server_settings_stats_page dbh <> None in
+ (* Given the referer string, return the list of search terms. If none
+ * can be found, then throws Not_found.
+ *)
+ let search_terms_from_referer referer =
+ let _, argnames =
+ List.find (fun (rex, _) -> Pcre.pmatch ~rex referer) search_engines in
+ let _, _, args = Cocanwiki_cgi_args.parse referer in
+ let argname =
+ List.find (fun argname -> List.mem_assoc argname args) argnames in
+ let search_string = List.assoc argname args in
+ Pcre.split ~rex:split_words search_string
+ in
+
+ (* Given a full page of XHTML, highlight search terms found in the
+ * <body> part of the page.
+ *)
+ let highlight_search_terms xhtml search_terms span_class =
+ (* Split the original XHTML into strings and tags. For example if
+ * the original string is: "This is some <b>bold</b> text.<br/>", then
+ * after this step we will have the following list:
+ * [ "This is some "; "<b>"; "bold"; "</b>"; " text."; "<br/>" ]
+ *)
+ let xhtml = Pcre.extract_all ~rex:xhtml_re xhtml in
+ let xhtml = Array.to_list xhtml in
+ let xhtml = List.map (fun matches -> matches.(0)) xhtml in
+
+ (* Find the <body> ... </body> tags. We only want to apply
+ * highlighting to tags within this area.
+ *)
+ let rec list_split f acc = function
+ | [] -> List.rev acc, []
+ | ((x :: _) as xs) when f x -> List.rev acc, xs
+ | x :: xs ->
+ let acc = x :: acc in
+ list_split f acc xs
+ in
+ let head, body =
+ list_split (fun str -> String.starts_with str "<body") [] xhtml in
+ let body, tail =
+ list_split ((=) "</body>") [] body in
+ (* NB: Hopefully, xhtml = head @ body @ tail. *)
+
+ (* The search terms are a list of simple words. Turn into a big
+ * regular expression, because we want to substitute for each. We
+ * end up with a regexp like '(word1|word2|word3)'.
+ *)
+ let rex =
+ Pcre.regexp ~flags:[`CASELESS]
+ ("(" ^ String.concat "|" search_terms ^ ")") in
+
+ (* Do the substitution, but only on text, not elements! *)
+ let body =
+ let subst text =
+ "<span class=\"" ^ span_class ^ "\">" ^ text ^ "</span>"
+ in
+ List.map (fun str ->
+ if String.length str > 0 && str.[0] != '<' then
+ Pcre.substitute ~rex ~subst str
+ else
+ str) body in
+
+ (* Join the XHTML fragments back together again. *)
+ String.concat "" (List.concat [ head ; body ; tail ])
+ in
+
(* This code generates ordinary pages. *)
let make_page title description pageid last_modified_date has_page_css
version page page' =
t#conditional "user_logged_in" true;
t#set "username" username);
- q#template t
+ (* If we are coming from a search engine then we want to highlight
+ * search terms throughout the whole page ...
+ *)
+ try
+ let referer = Table.get (Request.headers_in r) "Referer" in
+ let search_terms = search_terms_from_referer referer in
+
+ (* Highlight the search terms. *)
+ let xhtml = t#to_string in
+ let xhtml = highlight_search_terms xhtml search_terms "search_term" in
+
+ (* Deliver the page. *)
+ q#header ();
+ print_string r xhtml
+ with
+ Not_found ->
+ (* No referer / no search terms / not a search engine referer. *)
+ q#template t
in
(* This code generates 404 pages. *)
git.annexia.org / cocanwiki.git / commitdiff