psyced/world/net/place/threads.c

581 lines
16 KiB
C

// $Id: threads.c,v 1.41 2008/01/05 12:42:17 lynx Exp $ // vim:syntax=lpc
//
#include <net.h>
#include <person.h>
#include <status.h>
inherit NET_PATH "place/owned";
#ifndef DEFAULT_BACKLOG
# define DEFAULT_BACKLOG 5
#endif
// datenstruktur für threads?
//
// bestehende struktur ist: großes array von entries.
//
// wie wärs mit mapping mit key=threadname und value=array-of-entries
// subjects werden abgeschafft: sie sind der name des threads
// wer einen thread in seinem reply umnennen will legt in wirklichkeit
// einen neuen thread an, meinetwegen mit "was: old thread"
//
// der nachteil an solch einer struktur wäre, dass man neue comments
// in alten threads nicht so schnell findet - man ist auf die notification
// angewiesen, was andererseits die stärke von psycblogs ist.
// man könnte die notifications zudem noch in die history einspeisen..
//
// nachteile an der bestehenden struktur ist: 1. threadname in jeder
// entry, 2. threads nur mittels durchlauf des ganzen blogs darstellbar
//
// momentmal.. das was du "comments" nennst sind doch schon die threads!
protected mapping* _thread;
volatile int last_modified;
volatile string webact;
create() {
::create();
unless (pointerp(_thread)) _thread = ({ });
}
cmd(a, args, b, source, vars) {
// TODO: multiline-sachen irgendwie
mapping entry;
int i = 0;
int num_entries;
//unless (source) source = previous_object();
switch (a) {
case "entries":
num_entries = sizeof(args) >= 2 ? to_int(args[1]) : DEFAULT_BACKLOG;
// _thread[<5..]
foreach( entry : _thread[<num_entries..] ) {
sendmsg(source, "_list_thread_item",
// hier auch noch den text?
"([_number]) \"[_thread]\", [_author]",
([
"_thread" : entry["thread"],
"_text" : entry["text"],
"_author" : entry["author"],
"_date" : entry["date"],
"_number" : i++,
"_nick_place" : MYNICK ]) );
}
return 1;
case "thread":
unless (sizeof(args) > 2){ // num + thread
sendmsg(source, "_warning_usage_thread",
"Usage: /thread <threadid> <title>", ([ ]));
return 1;
}
return setSubject(to_int(args[1]), ARGS(2));
case "blog":
case "submit":
case "addentry":
unless (qAide(SNICKER)) return;
unless (sizeof(args) >= 2) {
sendmsg(source, "_warning_usage_submit",
"Usage: /submit <title> <text>", ([ ]));
return 1;
}
return addEntry(ARGS(2), SNICKER, args[1]);
// TODO: append fuer multiline-sachen
#if 0
case "iterator":
unless (qAide(SNICKER)) return;
sendmsg(source, "_notice_thread_iterator",
"[_iterator] blog entries have been requested "
"since creation.", ([
// i suppose this wasn't intentionally using
// MMP _count so i rename it to _iterator
"_iterator" : v("iterator")
]) );
return 1;
#endif
case "deblog":
case "delentry":
unless (qAide(SNICKER)) return;
// ist das ein typecheck ob args ein int is?
if (sizeof(regexp( ({ args[1] }) , "^[0-9][0-9]*$"))) {
unless (delEntry(to_int(args[1]), source, vars)) {
sendmsg(source,"_error_invalid_thread_item",
"There is no such thread item.", ([ ]));
} else {
sendmsg(source, "_notice_thread_item_removed",
"Thread item [_number] has been removed.",
([ "_number" : ARGS(1) ]) );
}
}
return 1;
}
return ::cmd(a, args, b, source, vars);
}
msg(source, mc, data, mapping vars){
P2(("blog.c's msg: mc %O, source %O\n", mc, source))
// TODO: die source muss hierbei uebereinstimmen mit dem autor
if (abbrev("_notice_authentication", mc)){
sendmsg(source, "_notice_place_blog_authentication_success", "([_entry]) has been authenticated",
([ "_entry" : "1" ]) );
return;
}
if (abbrev("error_invalid_authentication", mc)) {
sendmsg(source, "_notice_place_blog_authentication_failure", "Warning, someone pretends to blog as you", ([ ]) );
return;
}
return ::msg(source, mc, data, vars);
}
#if 0
listLastEntries(number) {
mapping* entries;
int i;
entries = _thread || ({ });
unless (sizeof(entries)) return 1;
i = v("iterator") || 0;
vSet("iterator", i + 1);
return entries[<number..];
}
#endif
#if 0
allEntries(source) {
mapping* entries;
mapping ar;
int i = 0;
entries = _thread || ({ });
unless (sizeof(entries)) return 1;
foreach (ar : entries) {
sendmsg(source, "_message_", "([_number]) \"[_topic]\", [_author]", ([ // ??
"_topic" : ar["topic"],
"_text" : ar["text"],
"_author" : ar["author"],
"_date" : ar["date"],
"_number" : i++,
"_nick_place" : MYNICK ]) );
}
return 1;
}
#endif
#if 0
addForum() {
// suggested protocol message for the buha forum
// (creation of a new thread)
:_target psyc://psyced.org/@buha
:_encoding utf-8
:_nick_forum morpheus
:_category Test-Forum
:_thread Test-Thread_
:_page_thread https://www.buha.info/board/showthread.php?t=1
_notice_thread
[_nick_forum] hat einen neuen Thread in [_category] erstellt: [_thread] ([_page_thread])
.
}
#endif
setSubject(num, thread) {
mapping* entries;
entries = _thread || ({ });
// TODO: das hier muss sicherer
entries[num]["thread"] = thread;
_thread = entries;
save();
return 1;
}
// TODO: topic uebergeben
addEntry(text, unick, thread) {
mapping* entries;
mapping newentry = ([ "text" : text,
"author" : unick,
"date" : isotime(ctime(), 1),
"thread" : thread || "",
]);
entries = _thread || ({ });
entries += ({ newentry });
_thread = entries;
save();
castmsg(ME, "_notice_thread_item",
"[_nick] adds an entry in \"[_thread]\" of [_nick_place].", ([
"_entry" : text,
"_thread" : thread,
"_nick" : unick,
]) );
return 1;
}
addComment(text, unick, entry_id) {
mapping entry;
if (sizeof(_thread) > entry_id) {
entry = _thread[entry_id];
unless (entry["comments"]) {
entry["comments"] = ({ });
}
entry["comments"] += ({ (["text" : text, "nick" : unick ]) });
// vSet("entries", entries);
castmsg(ME, "_notice_thread_comment",
"[_nick] adds a comment in \"[_thread]\" of [_nick_place].", ([
"_entry" : entry["text"],
"_thread" : entry["thread"],
"_comment" : text,
"_nick" : unick,
]) );
return save();
}
return -1;
}
delEntry(int number, source, vars) {
array(string) entries, authors, a;
string unick;
int size;
entries = _thread || ({ });
unless (size = sizeof(entries)) return 0;
if (number >= size) return 0;
if (qAide(unick = lower_case(SNICKER))) {
unless (lower_case(entries[number]["author"]) == unick) return 0;
}
_thread = entries[0..number-1] + entries[number+1..];
save();
return 1;
}
htget(prot, query, headers, qs, data) {
mapping entrymap;
mixed target;
string nick;
int i;
int a;
mapping* entries;
int num_entries = query["last"] ? to_int(query["last"]) : DEFAULT_BACKLOG;
unless (webact) webact = PLACE_PATH + MYLOWERNICK;
// shouldnt it be "html" here?
sTextPath(query["layout"] || MYNICK, query["lang"], "ht");
// Kommentare anzeigen
if (query["comments"]) {
htok(prot);
// kommentare + urspruengliche Nachricht anzeigen
displayHeader();
displayComments(_thread[to_int(query["comments"])]);
// eingabeformular ohne betreff
write("<form action='" + webact + "' method='GET'>\n"
"<input type='hidden' name='request' value='post'>\n"
"PSYC Uni: <input type='text' name='uni'><br>\n"
"<input type='hidden' name='reply' value='" + query["comments"] +"'>\n"
"<textarea name='text' rows='14' cols='80'>Enter your text here</textarea><br>\n"
"<input type='submit' value='submit'>\n"
"</form>\n");
write("<br><hr><br>");
logView(a < 24 ? a : 12, "html", 15);
displayFooter();
return 1;
}
// formularbehandlung
if (query["request"] == "post" && query["uni"]) {
htok(prot);
/*
sendmsg uni -> _request_authentication mit thread und text drin
dann auf die antwort warten die nen vars mapping mit thread + text hat wieder
*/
if (nick = legal_name(target = query["uni"])) {
target = summon_person(nick);
nick = target->qNick();
} else {
nick = target;
// write("Hello " + query["uni"] + "<br>\n");
// write("Remote auth doesn't work yet. TODO!!!\n");
// return 1;
}
#ifdef OWNED
if (qAide(nick)) {
#endif
#if 0
sendmsg(target, "_request_authentication", "please auth me!",
(["_host_IP" : query_ip_number(),
"_blog_thread" : query["thread"],
"_blog_text" : query["text"] ]));
write("your submit is avaiting authentication by " + query["uni"] + "<br>\n");
#endif // 0
if (target->checkAuthentication(ME, ([ "_host_IP" : query_ip_number() ]) ) > 0) {
// check ob reply auf irgendwas ist...
if (query["reply"]) {
addComment(query["text"], query["uni"], to_int(query["reply"]));
} else {
addEntry(query["text"], query["uni"], query["thread"]);
}
write("authentication successful!\n");
} else {
write("not authenticated!\n");
}
#ifdef OWNED
} else {
write("You are not owner or aide of this place.\n");
}
#endif
return 1;
}
// neuen Eintrag verfassen
if (query["request"] == "form") {
htok(prot);
displayHeader();
write("<form action='" + webact + "' method='GET'>\n"
"<input type='hidden' name='request' value='post'>\n"
"PSYC Identity:<br><input type='text' name='uni' size=60><br>\n"
"Thread:<br><input type='text' name='thread' size=60><br>\n"
"Text:<br>\n<textarea name='text' rows='14' cols='64'></textarea><br>\n"
"<input type='submit' value='CREATE MESSAGE'>\n"
"</form>\n");
displayFooter();
return 1;
}
::htget(prot, query, headers, qs, data, 1); // no processing, just info
// javascript-export
if (query["export"] == "javascript") {
// check If-Modified-Since header
htok3(prot, "application/x-javascript", "Cache-Control: no-cache\n");
jscriptExport(num_entries);
} else if (query["export"] == "rss" || query["export"] == "rdf") {
// export als RSS
// scheinbar gibt es ein limit von 15 items / channel
// htquote auch hier anwenden
// check If-Modified-Since header
htok3(prot, "text/xml", "");
rssExport(num_entries);
} else {
// normaler Export
P2(("all entries: %O\n", _thread))
htok3(prot, "text/html", "Cache-Control: no-cache\n");
displayHeader();
// display the blog
displayMain(num_entries);
// display the chatlog
logView(a < 24 ? a : 12, "html", 15);
displayFooter();
}
return 1;
}
rssExport(last) {
int i;
int len;
len = sizeof(_thread);
if (last > len) last = len;
write("<?xml version=\"1.0\" encoding=\"" SYSTEM_CHARSET "\" ?>\n"
"<rdf:RDF\n"
"xmlns:rdf=\"http://www.w3.org/1999/02/22-rdf-syntax-ns#\"\n"
"xmlns=\"http://purl.org/rss/1.0/\">\n\n"
"<channel>\n"
"\t<title>PSYC - Protocol for Synchronous Conferencing</title>\n"
"\t<link>http://www.psyc.eu</link>\n"
"\t<description>News about the PSYC project</description>\n"
"</channel>\n");
for (i = len - last; i < len; i++) {
write("\n<item>\n"
"\t<title>"+ _thread[i]["thread"] +"</title>\n"
"\t<link>http://" + SERVER_HOST + ":33333" + webact + "?comments=" + i + "</link>\n"
"\t<description>" + _thread[i]["text"] + "</description>\n"
"\t<dc:date>" + _thread[i]["date"] + "</dc:date>\n"
"\t<dc:creator>" + _thread[i]["author"] + "</dc:creator>\n");
write("</item>\n");
}
write("</rdf:RDF>\n");
}
jscriptExport(last) {
mapping item;
string buf = "";
// htok3(prot, "application/x-javascript", "Cache-Control: no-cache\n");
write("function Entry(thread, author, date, text) {\n"
"\tthis.thread = thread;\n"
"\tthis.author = author;\n"
"\tthis.date = date;\n"
"\tthis.text = text;\n"
"}\n\n"
"document.blogentries = new Array(\n");
foreach (item : _thread[<last..]) {
buf += "new Entry(\"" + item["thread"] + "\", \""
+ item["author"] + "\", \""
+ item["date"] + "\", \""
+ item["text"] + "\"),\n";
}
buf = buf[..<3] + ");";
write(buf);
}
#if 0
displayMain(last) {
int i;
int len;
len = sizeof(_thread);
if (last > len) last = len;
P2(("len %d, last %d\n", len, last))
for (i = len - last; i < len; i++) {
write("<table><tr><td class='blogthread'>" + _thread[i]["thread"]
+ "</td>"
"<td class='blogauthor'>" + _thread[i]["author"] + "</td>"
"<td class='blogdate'>" + _thread[i]["date"] + "</td></tr>"
"<tr><td class='blogtext' colspan=3>" + _thread[i]["text"]
+ "</td></tr>"
"<tr><td colspan=3 align='right'>");
write("<a href='" + webact + "?comments=" + i + "'>there are " + sizeof(_thread[i]["comments"]) + " comments</a>");
write("</td></tr></table>\n");
}
}
#endif
displayMain(last) {
int i;
int len;
string t;
len = sizeof(_thread);
if (last > len) last = len;
// reverse order
for (i = len-1; i >= len - last; i--) {
t = htquote(_thread[i]["text"]);
t = replace(t, "\n", "<br>\n");
write("<table class='newsTable' width='560px'>\n"
"<tr>"
"<td class='newsAuthor' width='20%'>" + _thread[i]["author"] + "</td>"
"<td class='newsTime' nowrap width='20%'>" + _thread[i]["date"] + "</td>"
"<td class='newsSubject' width='60%'>" + htquote(_thread[i]["thread"]) + "</td>"
"</tr>\n"
"<tr>"
"<td colspan='3' class='newsText'>" + t + "</td>"
"</tr>\n"
"<tr>"
"<td colspan='3' class='newsLink'><a href='" + webact + "?comments"
"=" + i + "'>" + sizeof(_thread[i]["comments"]) + " comments</a></td>"
"</tr>\n"
"</table>\n");
}
}
displayComments(data) {
mapping item;
write("Ursprüngliche Nachricht:<br>"
"<b>" + data["author"] + "</b>: " + data["text"] + "<br><br>\n");
if (data["comments"]) {
foreach(item : data["comments"]) {
write("<b>" + item["nick"] + "</b>: " + item["text"] + "<br>\n");
}
} else {
write("no comments...<br>\n");
}
}
nntpget(cmd, args) {
mapping item;
int i;
P2(("calling nntpget %s with %O\n", cmd, args))
switch(cmd) {
case "LIST":
write(MYNICK + " 0 1 n\n");
break;
case "ARTICLE":
i = to_int(args) - 1;
P2(("i is: %d\n", i))
P2(("entries: %O\n", _thread))
item = _thread[i];
write(S("220 %d <%s%d@%s> article\n",
i + 1, MYNICK, i + 1, SERVER_HOST));
write(S("From: %s\n", item["author"]));
write(S("Newsgroups: %s\n", MYNICK));
write(S("Subject: %s\n", item["thread"]));
write(S("Date: %s\n", item["date"]));
write(S("Xref: %s %s:%d\n", SERVER_HOST, MYNICK, i + 1));
write(S("Message-ID: <%s$%d@%s>\n", MYNICK, i+1, SERVER_HOST));
write("\n");
write(item["text"]);
write("\n.\n");
break;
case "GROUP":
write(S("211 %d 1 %d %s\n", sizeof(_thread),
sizeof(_thread), MYNICK));
break;
case "XOVER":
for (i = 0; i < sizeof(_thread); i++) {
item = _thread[i];
P2(("item: %O\n", item))
write(S("%d\t%s\t%s\t%s <%s%d@%s>\t1609\t22\tXref: news.t-online.com\t%s:%d\n",
i+1, item["thread"],
item["author"], item["date"],
MYNICK, i+1,
SERVER_HOST, MYNICK, i+1));
}
break;
default:
P2(("unimplemented nntp command: %s\n", cmd))
}
}
#ifndef STYLESHEET
# define STYLESHEET (v("_uniform_style") || "http://www.psyc.eu/blog.css")
#endif
// wir können zwei strategien fahren.. die technisch einfachere ist es
// die reihenfolge der elemente festzulegen und für jedes ein w(_HTML_xy
// auszuspucken. flexibler wär's stattdessen wenn jede seite ein einziges
// w(_PAGES_xy ausgeben würde in dem es per [_HTML_list_threads] oder
// ähnlichem die blog-elemente per psyctext-vars übergibt ... dann kann
// es immernoch per {_HTML_head_threads} header und footer einheitlich
// halten. womöglich kann man auch nachträglich plan A in plan B
// umwandeln..... hmmm -lynX
//
displayHeader() {
w("_HTML_head_threads",
"<link rel='stylesheet' type='text/css' href='"+
STYLESHEET +"'>\n"+
"<style type='text/css'>\n"
"<!--\n"
"body { font-family: lucida,verdana,geneva; }\n"
"td { font-family: lucida,verdana,geneva; }\n"
"table.newsTable {border: 1px solid #6f6; padding:4px; margin:6px;}\n"
"table td.newsAuthor {color:#fff; font-weight:bold;"
" border: 1px solid #6f6;"
" background-color:#041;}\n"
"table td.newsTime {color:#ddd;border: 1px solid #6f6;"
" background-color:#030;}\n"
"table td.newsSubject {color:#ddd; border: 1px solid #6f6;"
" background-color:#030;}\n"
"table td.newsLink {background-color:#030;}\n"
"table td.newsText {}\n"
"//-->\n"
"</style>\n"
"<body bgcolor='#002200' text='#33ff33' link='#ffdf00'"
"vlink='#ffaf00' alink='#00ff00'>\n\n");
}
displayFooter() {
w("_HTML_tail_threads", "</body>");
}