NewsDownloader: use feed.description as news context instead download full web page (#3426)

Fixes #3425.

plugins/newsdownloader.koplugin/feed_config.lua

@@ -1,11 +1,27 @@
 return {
 	-- list your feeds here:
 
-    { "http://feeds.reuters.com/Reuters/worldNews?format=xml", limit = 2},
-    -- set 'limit' to change number of 'news' to be downloaded from source
-    -- 'limit' equal "0" means no limit.
-    { "http://www.pcworld.com/index.rss", limit = 1 },
+    { "http://feeds.reuters.com/Reuters/worldNews?format=xml", limit = 2, download_full_article=false},
+
+    { "https://www.pcworld.com/index.rss", limit = 7 , download_full_article=true},
+
+		-- comment out line ("--" at line start) to stop downloading source
+    --{ "http://www.football.co.uk/international/rss.xml", limit = 0 , download_full_article=true},
+
+
+
+
+
+
+
+		--HELP:
+		-- use syntax: {"your_url", limit= max_number_of_items_to_be_created, download_full_article=true/false}
+
+		-- set 'limit' to change number of 'news' to be created
+    -- 'limit' equal "0" means no limit.
+
+		-- 'download_full_article=false' - means download full article using feed link (may not always work correctly)
+		-- 'download_full_article=true' - means use only feed description to create feeds (usually only part of the article)
+
 
-    -- comment out line to stop downloading source
-    --{ "http://www.football.co.uk/international/rss.xml", limit = 0 },
 }

plugins/newsdownloader.koplugin/main.lua

@@ -160,12 +160,13 @@ function NewsDownloader:loadConfigAndProcessFeeds()
     for idx, feed in ipairs(feed_config) do
         local url = feed[1]
         local limit = feed.limit
+        local download_full_article = feed.download_full_article
         if url and limit then
             info = InfoMessage:new{ text = T(_("Processing: %1"), url) }
             UIManager:show(info)
             -- processFeedSource is a blocking call, so manually force a UI refresh beforehand
             UIManager:forceRePaint()
-            self:processFeedSource(url, tonumber(limit), unsupported_feeds_urls)
+            self:processFeedSource(url, tonumber(limit), unsupported_feeds_urls, download_full_article)
             UIManager:close(info)
         else
             logger.warn('NewsDownloader: invalid feed config entry', feed)
@@ -191,7 +192,7 @@ function NewsDownloader:loadConfigAndProcessFeeds()
     end
 end
 
-function NewsDownloader:processFeedSource(url, limit, unsupported_feeds_urls)
+function NewsDownloader:processFeedSource(url, limit, unsupported_feeds_urls, download_full_article)
     local resp_lines = {}
     local parsed = socket_url.parse(url)
     local httpRequest = parsed.scheme == 'http' and http.request or https.request
@@ -207,9 +208,9 @@ function NewsDownloader:processFeedSource(url, limit, unsupported_feeds_urls)
     local is_atom = feeds.feed and feeds.feed.title and feeds.feed.entry[1] and feeds.feed.entry[1].title and feeds.feed.entry[1].link
 
     if is_atom then
-        self:processAtom(feeds, limit)
+        self:processAtom(feeds, limit, download_full_article)
     elseif is_rss then
-        self:processRSS(feeds, limit)
+        self:processRSS(feeds, limit, download_full_article)
     else
         table.insert(unsupported_feeds_urls, url)
         return
@@ -233,7 +234,7 @@ function NewsDownloader:deserializeXMLString(xml_str)
     return xmlhandler.root
 end
 
-function NewsDownloader:processAtom(feeds, limit)
+function NewsDownloader:processAtom(feeds, limit, download_full_article)
     local feed_output_dir = string.format("%s%s/",
                                           news_download_dir_path,
                                           util.replaceInvalidChars(getFeedTitle(feeds.feed.title)))
@@ -245,11 +246,15 @@ function NewsDownloader:processAtom(feeds, limit)
         if limit ~= 0 and index - 1 == limit then
             break
         end
-        self:downloadFeed(feed, feed_output_dir)
+        if download_full_article then
+            self:downloadFeed(feed, feed_output_dir)
+        else
+            self:createFromDescription(feed, feed.context, feed_output_dir)
+        end
     end
 end
 
-function NewsDownloader:processRSS(feeds, limit)
+function NewsDownloader:processRSS(feeds, limit, download_full_article)
     local feed_output_dir = ("%s%s/"):format(
         news_download_dir_path, util.replaceInvalidChars(feeds.rss.channel.title))
     if not lfs.attributes(feed_output_dir, "mode") then
@@ -260,7 +265,11 @@ function NewsDownloader:processRSS(feeds, limit)
         if limit ~= 0 and index - 1 == limit then
             break
         end
-        self:downloadFeed(feed, feed_output_dir)
+        if download_full_article then
+            self:downloadFeed(feed, feed_output_dir)
+        else
+            self:createFromDescription(feed, feed.description, feed_output_dir)
+        end
     end
 end
 
@@ -276,6 +285,25 @@ function NewsDownloader:downloadFeed(feed, feed_output_dir)
     httpRequest({ url = link, sink = ltn12.sink.file(io.open(news_dl_path, 'w')), })
 end
 
+function NewsDownloader:createFromDescription(feed, context, feed_output_dir)
+    local news_file_path = ("%s%s%s"):format(feed_output_dir,
+                                           util.replaceInvalidChars(getFeedTitle(feed.title)),
+                                           file_extension)
+    logger.dbg("NewsDownloader: News file will be created :", news_file_path)
+    local file = io.open(news_file_path, "w")
+    local footer = _("This is just description of the feed. To download full article go to News Downloader settings and change 'download_full_article' to 'true'")
+
+    local html = string.format([[<!DOCTYPE html>
+<html>
+<head><meta charset='UTF-8'><title>%s</title></head>
+<body><header><h2>%s</h2></header><article>%s</article>
+<br><footer><small>%s</small></footer>
+</body>
+</html>]], feed.title, feed.title, context, footer)
+    file:write(html)
+    file:close()
+end
+
 function NewsDownloader:removeNewsButKeepFeedConfig()
     logger.dbg("NewsDownloader: Removing news from :", news_download_dir_path)
     for entry in lfs.dir(news_download_dir_path) do