refactoring

CMakeLists.txt

@@ -1,6 +1,6 @@
 cmake_minimum_required (VERSION 3.7)
 project (mastorss
-         VERSION 0.3.7
+         VERSION 0.4.0
          LANGUAGES CXX
 )
 

README.md

@@ -1,7 +1,9 @@
 **mastorss** dumps RSS feeds into a mastodon account.
-It is hacked together and generally only extended/fixed when it fails.
+Supports RSS 2.0 but not RSS 0.92. Does not support Atom at the moment.
-Do NOT assume it follows any standards.
+
-Use at your own risk.
+<item>s in feeds must have <link>, <title> and <description>.
+
+The documentation is far from complete, sorry.
 
 # Install
 
@@ -13,7 +15,7 @@ Use at your own risk.
  * [boost](http://www.boost.org/) (tested: 1.63.0)
  * [libcurl](https://curl.haxx.se/) (tested: 7.58.0)
  * [curlpp](http://www.curlpp.org/) (tested: 0.8.1)
- * [mastodon-cpp](https://github.com/tastytea/mastodon-cpp) (at least: 0.2.13)
+ * [mastodon-cpp](https://github.com/tastytea/mastodon-cpp) (at least: 0.6.4)
 
 ## Get sourcecode
 
@@ -28,9 +30,7 @@ Use at your own risk.
     cmake ..
     make
 
-cmake options:
+## Install
-
- * `-DCMAKE_BUILD_TYPE=Debug` for a debug build
 
 Install with `make install`.
 
@@ -53,7 +53,16 @@ ${HOME}/.config/mastorss/config-example.json
             "feedurl": "https:\/\/example.com\/feed.rss",
             "access_token": "123abc",
             "max_size": "400",
-            "last_entry": "Example\n\nThis is an example.\n\nhttps:\/\/example.com\/12345.html\n\n#bot"
+            "skip":
+            [
+                "If the entry starts with this, skip it",
+                "Skip me too!"
+            ],
+            "fixes":
+            [
+                "delete this",
+                "[Rr]ead more(\.{3}|…)"
+            ]
         }
     }
 

src/config.cpp

@@ -31,7 +31,7 @@ using std::cerr;
 using std::cin;
 using std::string;
 
-std::uint16_t read_config(pt::ptree &config, const string &profile, string &instance, string &access_token, string &feedurl)
+std::uint16_t read_config(string &instance, string &access_token, string &feedurl)
 {
     bool config_changed = false;
 
@@ -63,8 +63,7 @@ std::uint16_t read_config(pt::ptree &config, const string &profile, string &inst
         cout << "No access token found.\n";
         string client_id, client_secret, url;
         Mastodon::API masto(instance, "");
-        std::uint16_t ret = masto.register_app1(instance,
+        std::uint16_t ret = masto.register_app1("mastorss",
-                                                "mastorss",
                                                 "urn:ietf:wg:oauth:2.0:oob",
                                                 "write",
                                                 "https://github.com/tastytea/mastorss",
@@ -78,8 +77,7 @@ std::uint16_t read_config(pt::ptree &config, const string &profile, string &inst
             cout << "Insert code: ";
             cin >> code;
 
-            masto.register_app2(instance,
+            masto.register_app2(client_id,
-                                client_id,
                                 client_secret,
                                 "urn:ietf:wg:oauth:2.0:oob",
                                 code,

src/mastorss.cpp

@@ -39,6 +39,8 @@ using std::string;
 // Initialize global variables
 std::uint16_t max_size = 500;
 const string filepath = string(getenv("HOME")) + "/.config/mastorss/";
+pt::ptree config;
+std::string profile;
 
 int main(int argc, char *argv[])
 {
@@ -53,29 +55,29 @@ int main(int argc, char *argv[])
         max_size = std::stoi(argv[2]);
     }
 
-    pt::ptree config;
     string instance = "";
     string access_token = "";
     string feedurl = "";
-    const string profile = argv[1];
+    profile = argv[1];
     std::uint16_t ret;
-
-    read_config(config, profile, instance, access_token, feedurl);
-    curlpp_init();
-
     string answer;
-    string last_entry = config.get(profile + ".last_entry", "");
     std::vector<string> entries;
 
+    read_config(instance, access_token, feedurl);
+    curlpp_init();
+
     ret = http_get(feedurl, answer, "mastorss/" + (string)global::version);
     if (ret != 0)
     {
         return ret;
     }
-    entries = parse_website(profile, answer);
+    entries = parse_website(answer);
 
+    string last_entry = config.get(profile + ".last_entry", "");
     if (last_entry.empty())
     {
+        // If no last_entry is stored in the config file,
+        // make last_entry the second-newest entry.
         last_entry = entries.at(1);
     }
     config.put(profile + ".last_entry", entries.front());
@@ -85,6 +87,8 @@ int main(int argc, char *argv[])
     {
         if (!new_content && (*rit).compare(last_entry) == 0)
         {
+            // If the last entry is found in entries,
+            // start tooting in the next loop.
             new_content = true;
             continue;
         }
@@ -113,14 +117,7 @@ int main(int argc, char *argv[])
         std::this_thread::sleep_for(std::chrono::seconds(2));
     }
 
-    // If the last entry is not in the current feed, set the oldest item as last entry
+    // Write the new last_entry only if no error happened.
-    // Could lead to spamming if an item gets deleted or changed.
-    // Update: It DID couse spamming :-(
-    // TODO: Think of something better
-    // if (!new_content)
-    // {
-    //     config.put(profile + ".last_entry", entries.at(entries.size() - 1));
-    // }
     pt::write_json(filepath + "config-" + profile + ".json", config);
 
     return 0;

src/mastorss.hpp

@@ -11,10 +11,12 @@ using std::string;
 
 extern std::uint16_t max_size;
 extern const string filepath;
+extern pt::ptree config;
+extern std::string profile;
 
-std::uint16_t read_config(pt::ptree &config, const string &profile, string &instance, string &access_token, string &feedurl);
+std::uint16_t read_config(string &instance, string &access_token, string &feedurl);
 
-std::vector<string> parse_website(const string &profile, const string &xml);
+std::vector<string> parse_website(const string &xml);
 void unescape_html(const string &str);
 void individual_fixes(string &str);
 

src/parse.cpp

@@ -64,7 +64,7 @@ void unescape_html(string &str)
     str = std::regex_replace(str, reapos, "\'");
 }
 
-std::vector<string> parse_website(const string &profile, const string &xml)
+std::vector<string> parse_website(const string &xml)
 {
     pt::ptree json;
     std::vector<string> watchwords;
@@ -83,6 +83,7 @@ std::vector<string> parse_website(const string &profile, const string &xml)
 
     try
     {
+        // Read profile-specific hashtags or fail silently
         for (const pt::ptree::value_type &value : json.get_child(profile + ".tags"))
         {
             watchwords.push_back(value.second.data());
@@ -94,6 +95,7 @@ std::vector<string> parse_website(const string &profile, const string &xml)
     }
     try
     {
+        // Read global hashtags or fail silently
         for (const pt::ptree::value_type &value : json.get_child("global.tags"))
         {
             watchwords.push_back(value.second.data());
@@ -120,14 +122,29 @@ std::vector<string> parse_website(const string &profile, const string &xml)
                 string desc = v.second.get_child("description").data();
                 string str = title + "\n\n" + desc;
 
-                // ANF News puts this always on top, causing us to think it's new
+                try
-                if (title.compare(0, 35, "Newsticker zu den Angriffen auf Efr") == 0)
                 {
-                    continue;
+                    // Skip entries beginning with this text
+                    for (const pt::ptree::value_type &v : config.get_child(profile + ".skip"))
+                    {
+                        const string skip = v.second.data();
+                        if (!skip.empty())
+                        {
+                            if (title.compare(0, skip.length(), skip) == 0)
+                            {
+                                continue;
+                            }
+                        }
+                    }
+                }
+                catch (const std::exception &e)
+                {
+                    // Node not found, no problem
                 }
 
                 unescape_html(str);
 
+                // Try to turn the HTML into human-readable text
                 std::regex reparagraph("</p><p>");
                 std::regex recdata1("<!\\[CDATA\\[");
                 std::regex recdata2("\\]\\]>");
@@ -166,16 +183,19 @@ std::vector<string> parse_website(const string &profile, const string &xml)
     return ret;
 }
 
+// Read regular expressions from the config file and delete all matches.
 void individual_fixes(string &str)
 {
-    // de.indymedia.org articles sometimes have CSS in the description
+    try
-    std::regex reindyfuckup("\\/\\* Style Definitions \\*\\/[.[:space:]]*$");
+    {
-    // Direkte Action closing
+         for (const pt::ptree::value_type &v : config.get_child(profile + ".fixes"))
-    std::regex redaclosing("Der Beitrag .* erschien zuerst auf Direkte Aktion.");
+        {
-    // GG/BO closing
+            std::regex refix(v.second.data());
-    std::regex reggboclosing("Die von den einzelnen AutorInnen .*$");
+            str = std::regex_replace(str, refix, "");
-
+        }
-    str = std::regex_replace(str, reindyfuckup, "");
+    }
-    str = std::regex_replace(str, redaclosing, "");
+    catch (const std::exception &e)
-    str = std::regex_replace(str, reggboclosing, "");
+    {
+        // Node not found, no problem
+    }
 }