/*
- * Bring external RSS feeds into rooms.
+ * Bring external RSS and/or Atom feeds into rooms. This module implements a
+ * very loose parser that scrapes both kinds of feeds and is not picky about
+ * the standards compliance of the source data.
*
- * Copyright (c) 2007-2017 by the citadel.org team
+ * Copyright (c) 2007-2020 by the citadel.org team
*
- * This program is open source software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License version 3.
+ * This program is open source software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License version 3.
*
* 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
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*/
#include "parsedate.h"
#include "database.h"
#include "citadel_dirs.h"
-#include "md5.h"
#include "context.h"
#include "internet_addressing.h"
};
time_t last_run = 0L;
-struct CitContext rss_CC;
struct rssurl *rsstodo = NULL;
void rss_end_element(void *data, const char *el)
{
struct rssparser *r = (struct rssparser *)data;
+ StrBuf *encoded_field;
+
+ if (StrLength(r->CData) > 0) { // strip leading/trailing whitespace from field
+ StrBufTrim(r->CData);
+ }
if ( // end of a new item(rss) or entry(atom)
(!strcasecmp(el, "entry"))
|| (!strcasecmp(el, "item"))
) {
-
if (r->msg != NULL) { // Save the message to the rooms
// use the link as an item id if nothing else is available
long msgnum = (-1);
for (rr=r->rooms; rr!=NULL; rr=rr->next) {
if (rr == r->rooms) {
- msgnum = CtdlSubmitMsg(r->msg, NULL, rr->room, 0);
+ msgnum = CtdlSubmitMsg(r->msg, NULL, rr->room); // in first room, save msg
}
else {
- CtdlSaveMsgPointerInRoom(rr->room, msgnum, 0, NULL);
+ CtdlSaveMsgPointerInRoom(rr->room, msgnum, 0, NULL); // elsewhere, save a pointer
}
syslog(LOG_DEBUG, "rssclient: saved message %ld to %s", msgnum, rr->room);
}
else if (!strcasecmp(el, "title")) { // item subject (rss and atom)
if ((r->msg != NULL) && (CM_IsEmpty(r->msg, eMsgSubject))) {
- CM_SetField(r->msg, eMsgSubject, ChrPtr(r->CData), StrLength(r->CData));
- striplt(r->msg->cm_fields[eMsgSubject]);
+ encoded_field = NewStrBuf();
+ StrBufRFC2047encode(&encoded_field, r->CData);
+ CM_SetAsFieldSB(r->msg, eMsgSubject, &encoded_field);
}
}
- else if (!strcasecmp(el, "author")) { // author of item (rss and maybe atom)
+ else if (!strcasecmp(el, "creator")) { // <creator> can be used if <author> is not present
if ((r->msg != NULL) && (CM_IsEmpty(r->msg, eAuthor))) {
- CM_SetField(r->msg, eAuthor, ChrPtr(r->CData), StrLength(r->CData));
- striplt(r->msg->cm_fields[eAuthor]);
+ encoded_field = NewStrBuf();
+ StrBufRFC2047encode(&encoded_field, r->CData);
+ CM_SetAsFieldSB(r->msg, eAuthor, &encoded_field);
+ }
+ }
+
+ else if (!strcasecmp(el, "author")) { // <author> supercedes <creator> if both are present
+ if (r->msg != NULL) {
+ encoded_field = NewStrBuf();
+ StrBufRFC2047encode(&encoded_field, r->CData);
+ CM_SetAsFieldSB(r->msg, eAuthor, &encoded_field);
}
}
r->link = NULL;
}
r->link = strdup(ChrPtr(r->CData));
- striplt(r->link);
}
else if (
r->item_id = NULL;
}
r->item_id = strdup(ChrPtr(r->CData));
- striplt(r->item_id);
}
else if (
r->description = NULL;
}
r->description = strdup(ChrPtr(r->CData));
- striplt(r->description);
}
if (r->CData != NULL) {
thisone = r;
}
}
+
if (thisone == NULL) {
thisone = malloc(sizeof(struct rssurl));
thisone->url = strdup(rssurl);
return;
}
- become_session(&rss_CC);
syslog(LOG_DEBUG, "rssclient: started");
CtdlForEachRoom(rssclient_scan_room, NULL);
rss_pull_feeds();
syslog(LOG_INFO, "rssclient: using %s", curl_version());
CtdlRegisterSessionHook(rssclient_scan, EVT_TIMER, PRIO_AGGR + 300);
}
- else
- {
- CtdlFillSystemContext(&rss_CC, "rssclient");
- }
return "rssclient";
}