Extract and parse <item>s from RSS feeds.

src/document.cpp

@@ -19,19 +19,28 @@
 #include "version.hpp"
 
 #include <boost/log/trivial.hpp>
+#include <boost/property_tree/xml_parser.hpp>
+#include <boost/regex.hpp>
+#include <mastodon-cpp/mastodon-cpp.hpp>
 #include <restclient-cpp/connection.h>
 #include <restclient-cpp/restclient.h>
 
+#include <list>
+#include <sstream>
 #include <string>
 #include <utility>
 
 namespace mastorss
 {
+using boost::regex;
+using boost::regex_replace;
+using std::list;
+using std::istringstream;
 using std::string;
 using std::move;
 
-Document::Document(string uri)
-    : _uri{move(uri)}
+Document::Document(const ProfileData &data)
+    : _data{data}
 {
     RestClient::init();
 
@@ -45,7 +54,7 @@ Document::~Document()
 
 void Document::download()
 {
-    RestClient::Connection connection(_uri);
+    RestClient::Connection connection(_data.feedurl);
     connection.SetUserAgent(string("mastorss/").append(version));
     connection.FollowRedirects(true, 10);
 
@@ -56,7 +65,7 @@ void Document::download()
     case 200:
     {
         _raw_doc = response.body;
-        BOOST_LOG_TRIVIAL(debug) << "Downloaded feed: " << _uri;
+        BOOST_LOG_TRIVIAL(debug) << "Downloaded feed: " << _data.feedurl;
         break;
     }
     case 301:
@@ -76,4 +85,90 @@ void Document::download()
     }
     }
 }
+
+void Document::parse()
+{
+    pt::ptree tree;
+    istringstream iss{_raw_doc};
+    pt::read_xml(iss, tree);
+
+    if (tree.front().first == "rss")
+    {
+        parse_rss(tree);
+    }
+}
+
+void Document::parse_rss(const pt::ptree &tree)
+{
+    for (const auto &child : tree.get_child("rss.channel"))
+    {
+        if (child.first == "item")
+        {
+            const auto &rssitem = child.second;
+
+            string guid{rssitem.get<string>("guid")};
+            if (guid.empty())   // We hope either <guid> or <link> are present.
+            {
+                guid = rssitem.get<string>("link");
+            }
+            if (guid == _data.last_guid)
+            {
+                break;
+            }
+
+            bool skipthis{false};
+            string title = rssitem.get<string>("title");
+            for (const auto &skip : _data.skip)
+            {
+                if (title.substr(0, skip.length()) == skip)
+                {
+                    skipthis = true;
+                    break;
+                }
+            }
+            if (skipthis)
+            {
+                BOOST_LOG_TRIVIAL(debug) << "Skipped GUID: " << guid;
+                continue;
+            }
+
+            Item item;
+            item.description = remove_html(rssitem.get<string>("description"));
+            item.guid = move(guid);
+            item.link = rssitem.get<string>("link");
+            item.title = move(title);
+            new_items.push_back(item);
+
+            BOOST_LOG_TRIVIAL(debug) << "Found GUID: " << item.guid;
+        }
+    }
+}
+
+string Document::remove_html(string html) const
+{
+    html = Mastodon::unescape_html(html); // Decode HTML entities.
+
+    html = regex_replace(html, regex{"<p>"}, "\n\n");
+
+    const list re_list
+        {
+            regex{R"(<!\[CDATA\[)"},      // CDATA beginning.
+            regex{R"(\]\]>)"},            // CDATA end.
+            regex{"<[^>]+>"},             // HTML tags.
+            regex{R"(\r)"},               // Carriage return.
+            regex{"\\n[ \\t\u00a0]+\\n"}, // Whitespace between newlines.
+            regex{R"(^\n+)"}              // Newlines at the beginning.
+        };
+    for (const regex &re : re_list)
+    {
+        html = regex_replace(html, re, "");
+    }
+
+    // Remove excess newlines.
+    html = regex_replace(html, regex{R"(\n{3,})"}, "\n\n");
+    // Replace single newlines with spaces (?<= is lookbehind, ?= is lookahead).
+    html = regex_replace(html, regex{R"((?<=[^\n])\n(?=[^\n]))"}, " ");
+
+    return html;
+}
 } // namespace mastorss

src/document.hpp

@@ -17,27 +17,59 @@
 #ifndef MASTORSS_DOCUMENT_HPP
 #define MASTORSS_DOCUMENT_HPP
 
+#include "config.hpp"
+
+#include <boost/property_tree/ptree.hpp>
+
 #include <string>
+#include <vector>
 
 namespace mastorss
 {
+namespace pt = boost::property_tree;
 using std::string;
+using std::vector;
 
+/*!
+ *  @brief  An Item of a feed.
+ *
+ *  @since  0.10.0
+ */
+struct Item
+{
+    string description;
+    string guid;
+    string link;
+    string title;
+};
+
+/*!
+ *  @brief  A feed.
+ *
+ *  @since  0.10.0
+ */
 class Document
 {
 public:
-    explicit Document(string uri);
+    explicit Document(const ProfileData &data);
     ~Document();
     Document(const Document &other) = default;
     Document &operator=(const Document &other) = delete;
     Document(Document &&other) = default;
     Document &operator=(Document &&other) = delete;
 
+    vector<Item> new_items;
+
     void download();
+    void parse();
 
 private:
-    const string _uri;
+    const ProfileData &_data;
     string _raw_doc;
+
+    void parse_rss(const pt::ptree &tree);
+    [[nodiscard]]
+    string remove_html(string html) const;
 };
 } // namespace mastorss
 

src/main.cpp

@@ -85,13 +85,18 @@ int main(int argc, char *argv[])
         }
         else
         {
-            const string_view profile = args[1];
+            const string_view profile{args[1]};
             BOOST_LOG_TRIVIAL(debug) << "Using profile: " << profile;
 
             try
             {
-                Config cfg(profile.data());
-                Document doc(cfg.data.feedurl);
+                Config cfg{profile.data()};
+                Document doc{cfg.data};
+                doc.parse();
+                for (const auto &item : doc.new_items)
+                {
+                    cout << "--\n" << item.description.substr(0, 200) << "\n";
+                }
             }
             catch (const FileException &e)
             {