Matrix: add a proper URI parser instead of regex for opening links. Show related body item image when pressing enter (for showing replied to media)

14 files changed, 215 insertions, 82 deletions

diff --git a/include/NetUtils.hpp b/include/NetUtils.hpp
new file mode 100644
index 0000000..84b9d18
--- /dev/null
+++ b/include/NetUtils.hpp
@@ -0,0 +1,11 @@
+#pragma once
+
+#include <string>
+#include <vector>
+
+namespace QuickMedia {
+    void html_escape_sequences(std::string &str);
+    void html_unescape_sequences(std::string &str);
+    std::string url_param_encode(const std::string &param);
+    std::vector<std::string> extract_urls(const std::string &str);
+}
\ No newline at end of file
diff --git a/plugins/Plugin.hpp b/plugins/Plugin.hpp
index 1427233..d754445 100644
--- a/plugins/Plugin.hpp
+++ b/plugins/Plugin.hpp
@@ -3,7 +3,6 @@
 #include "../include/Body.hpp"
 #include "../include/DownloadUtils.hpp"
 #include <stddef.h>
-#include <string>
 
 namespace QuickMedia {
     enum class PluginResult {
@@ -36,10 +35,6 @@ namespace QuickMedia {
         size_t index;
     };
 
-    void html_escape_sequences(std::string &str);
-    void html_unescape_sequences(std::string &str);
-    std::string url_param_encode(const std::string &param);
-
     SuggestionResult download_result_to_suggestion_result(DownloadResult download_result);
     PluginResult download_result_to_plugin_result(DownloadResult download_result);
     SearchResult download_result_to_search_result(DownloadResult download_result);
diff --git a/src/NetUtils.cpp b/src/NetUtils.cpp
new file mode 100644
index 0000000..e87c42c
--- /dev/null
+++ b/src/NetUtils.cpp
@@ -0,0 +1,140 @@
+#include "../include/NetUtils.hpp"
+#include "../include/StringUtils.hpp"
+#include <array>
+#include <sstream>
+#include <iomanip>
+
+namespace QuickMedia {
+    struct HtmlEscapeSequence {
+        char unescape_char;
+        std::string escape_sequence;
+    };
+
+    void html_escape_sequences(std::string &str) {
+        const std::array<HtmlEscapeSequence, 6> escape_sequences = {
+            HtmlEscapeSequence { '&', "&amp;" }, // This should be first, to not accidentally replace a new sequence caused by replacing this
+            HtmlEscapeSequence { '"', "&quot;" },
+            HtmlEscapeSequence { '\'', "&#39;" },
+            HtmlEscapeSequence { '<', "&lt;" },
+            HtmlEscapeSequence { '>', "&gt;" },
+            HtmlEscapeSequence { '\n', "<br>" }
+        };
+
+        for(const HtmlEscapeSequence &escape_sequence : escape_sequences) {
+            string_replace_all(str, escape_sequence.unescape_char, escape_sequence.escape_sequence);
+        }
+    }
+
+    struct HtmlUnescapeSequence {
+        std::string escape_sequence;
+        std::string unescaped_str;
+    };
+
+    void html_unescape_sequences(std::string &str) {
+        const std::array<HtmlUnescapeSequence, 6> unescape_sequences = {
+            HtmlUnescapeSequence { "&quot;", "\"" },
+            HtmlUnescapeSequence { "&#039;", "'" },
+            HtmlUnescapeSequence { "&#39;", "'" },
+            HtmlUnescapeSequence { "&lt;", "<" },
+            HtmlUnescapeSequence { "&gt;", ">" },
+            HtmlUnescapeSequence { "&amp;", "&" } // This should be last, to not accidentally replace a new sequence caused by replacing this
+        };
+
+        for(const HtmlUnescapeSequence &unescape_sequence : unescape_sequences) {
+            string_replace_all(str, unescape_sequence.escape_sequence, unescape_sequence.unescaped_str);
+        }
+    }
+
+    std::string url_param_encode(const std::string &param) {
+        std::ostringstream result;
+        result.fill('0');
+        result << std::hex;
+
+        for(char c : param) {
+            if(isalnum(c) || c == '-' || c == '_' || c == '.' || c == '~') {
+                result << c;
+            } else {
+                result << std::uppercase;
+                result << "%" << std::setw(2) << (int)(unsigned char)(c);
+            }
+        }
+
+        return result.str();
+    }
+
+    static bool is_alpha(char c) {
+        return (c >= 'a' && c <= 'z') || (c >= 'A' && c <= 'Z');
+    }
+
+    static bool is_digit(char c) {
+        return c >= '0' && c <= '9';
+    }
+
+    static bool is_url_character(char c) {
+        switch(c) {
+            case '%':
+            // Reserved
+            case ':':
+            case '/':
+            case '?':
+            case '#':
+            case '[':
+            case ']':
+            case '@':
+            case '!':
+            case '$':
+            case '&':
+            case '\'':
+            case '(':
+            case ')':
+            case '*':
+            case '+':
+            case ',':
+            case ';':
+            case '=':
+            // Unreserved:
+            case '-':
+            case '.':
+            case '_':
+            case '~':
+                return true;
+            default:
+                return is_alpha(c) || is_digit(c);
+        }
+    }
+
+    // Implementation follows URI standard: https://tools.ietf.org/html/rfc3986#section-2.2
+    // TODO: Maybe check if the TLD only contains valid characters (is_alpha)?
+    std::vector<std::string> extract_urls(const std::string &str) {
+        std::vector<std::string> urls;
+
+        size_t url_start = std::string::npos;
+        size_t url_dot_index = std::string::npos;
+        // str.size() is fine, we want to include the NULL character so we can extract url at the end of the string
+        for(size_t i = 0; i < (size_t)str.size() + 1; ++i) {
+            char c = str[i];
+            if(c == '.' && url_start != std::string::npos && url_dot_index == std::string::npos)
+                url_dot_index = i;
+            if(is_url_character(c)) {
+                if(url_start == std::string::npos)
+                    url_start = i;
+            } else {
+                if(url_start != std::string::npos) {
+                    // Its only an url if there is a dot and the dot is not the last character in the url, for example "example.com" is an url but "example." is not.
+                    if(url_dot_index != std::string::npos && url_dot_index != i - 1) {
+                        size_t url_length = i - url_start;
+                        char prev_char = str[i - 1];
+                        // We want to remove the last . or , because the string could contain for example "click on this like: example.com. There you can..."
+                        // and we want those links to work, I guess?
+                        if(prev_char == '.' || prev_char == ',')
+                            --url_length;
+                        urls.push_back(str.substr(url_start, url_length));
+                    }
+                    url_start = std::string::npos;
+                    url_dot_index = std::string::npos;
+                }
+            }
+        }
+        return urls;
+    }
+}
\ No newline at end of file
diff --git a/src/QuickMedia.cpp b/src/QuickMedia.cpp
index 070b8fe..e374dfb 100644
--- a/src/QuickMedia.cpp
+++ b/src/QuickMedia.cpp
@@ -18,12 +18,12 @@
 #include "../include/ImageUtils.hpp"
 #include "../include/base64_url.hpp"
 #include "../include/Entry.hpp"
+#include "../include/NetUtils.hpp"
 
 #include <assert.h>
 #include <cmath>
 #include <string.h>
 #include <signal.h>
-#include <regex>
 
 #include <SFML/Graphics/RectangleShape.hpp>
 #include <SFML/Window/Clipboard.hpp>
@@ -3225,8 +3225,6 @@ namespace QuickMedia {
         const float chat_input_padding_x = 15.0f;
         const float chat_input_padding_y = 15.0f;
 
-        std::regex url_extract_regex("(http(s?):\\/\\/)?([a-zA-Z0-9\\-_]+\\.)+[a-zA-Z]+[^\\s.,]+");
-
         Body url_selection_body(this, font.get(), bold_font.get(), cjk_font.get());
 
         sf::Clock read_marker_timer;
@@ -3476,36 +3474,40 @@ namespace QuickMedia {
                     if(tabs[selected_tab].type == ChatTabType::MESSAGES && event.key.code == sf::Keyboard::Enter) {
                         BodyItem *selected = tabs[selected_tab].body->get_selected();
                         if(selected) {
-                            if(!selected->url.empty()) {
-                                const char *content_type = link_get_content_type(selected->url);
+                            std::string selected_url = selected->url;
+                            if(selected_url.empty() && selected->embedded_item)
+                                selected_url = selected->embedded_item->url;
+                            if(!selected_url.empty()) {
+                                const char *content_type = link_get_content_type(selected_url);
                                 if(content_type && (strcmp(content_type, "audio") == 0 || strcmp(content_type, "video") == 0 || strcmp(content_type, "image") == 0)) {
                                     page_stack.push(PageType::CHAT);
                                     watched_videos.clear();
                                     current_page = PageType::VIDEO_CONTENT;
+                                    bool is_audio = strcmp(content_type, "audio") == 0;
+                                    bool prev_no_video = no_video;
+                                    no_video = is_audio;
                                     // TODO: Add title
-                                    video_content_page(video_page.get(), selected->url, "No title");
+                                    video_content_page(video_page.get(), selected_url, "No title");
+                                    no_video = prev_no_video;
                                     redraw = true;
                                     continue;
                                 }
 
-                                launch_url(selected->url.c_str());
+                                launch_url(selected_url);
                                 continue;
                             }
 
                             // TODO: If content type is a file, show file-manager prompt where it should be saved and asynchronously save it instead
                             
-                            // TODO: Change this when messages are not stored in the description
                             const std::string &message_str = selected->get_description();
-                            auto urls_begin = std::sregex_iterator(message_str.begin(), message_str.end(), url_extract_regex);
-                            auto urls_end = std::sregex_iterator();
-                            size_t num_urls = std::distance(urls_begin, urls_end);
-                            if(num_urls == 1) {
-                                launch_url(urls_begin->str());
-                            } else if(num_urls > 1) {
+                            std::vector<std::string> urls = extract_urls(message_str);
+                            if(urls.size() == 1) {
+                                launch_url(urls[0]);
+                            } else if(urls.size() > 1) {
                                 chat_state = ChatState::URL_SELECTION;
                                 url_selection_body.clear_items();
-                                for(auto it = urls_begin; it != urls_end; ++it) {
-                                    auto body_item = BodyItem::create(it->str());
+                                for(const std::string &url : urls) {
+                                    auto body_item = BodyItem::create(url);
                                     url_selection_body.items.push_back(std::move(body_item));
                                 }
                             }
diff --git a/src/plugins/Fourchan.cpp b/src/plugins/Fourchan.cpp
index 1d3681a..2938319 100644
--- a/src/plugins/Fourchan.cpp
+++ b/src/plugins/Fourchan.cpp
@@ -2,6 +2,7 @@
 #include "../../include/DataView.hpp"
 #include "../../include/Storage.hpp"
 #include "../../include/StringUtils.hpp"
+#include "../../include/NetUtils.hpp"
 #include <json/reader.h>
 #include <string.h>
 #include <tidy.h>
diff --git a/src/plugins/Mangadex.cpp b/src/plugins/Mangadex.cpp
index f0e163e..0ccecdd 100644
--- a/src/plugins/Mangadex.cpp
+++ b/src/plugins/Mangadex.cpp
@@ -2,6 +2,7 @@
 #include "../../include/Storage.hpp"
 #include "../../include/Notification.hpp"
 #include "../../include/StringUtils.hpp"
+#include "../../include/NetUtils.hpp"
 #include <rapidjson/document.h>
 #include <rapidjson/reader.h>
 #include <quickmedia/HtmlSearch.h>
diff --git a/src/plugins/Manganelo.cpp b/src/plugins/Manganelo.cpp
index e96bc65..b260dea 100644
--- a/src/plugins/Manganelo.cpp
+++ b/src/plugins/Manganelo.cpp
@@ -1,6 +1,7 @@
 #include "../../plugins/Manganelo.hpp"
 #include "../../include/Notification.hpp"
 #include "../../include/StringUtils.hpp"
+#include "../../include/NetUtils.hpp"
 #include <quickmedia/HtmlSearch.h>
 
 namespace QuickMedia {
diff --git a/src/plugins/Mangatown.cpp b/src/plugins/Mangatown.cpp
index 5d6f97f..992e0cc 100644
--- a/src/plugins/Mangatown.cpp
+++ b/src/plugins/Mangatown.cpp
@@ -1,6 +1,7 @@
 #include "../../plugins/Mangatown.hpp"
 #include "../../include/Notification.hpp"
 #include "../../include/StringUtils.hpp"
+#include "../../include/NetUtils.hpp"
 #include <quickmedia/HtmlSearch.h>
 
 static const std::string mangatown_url = "https://www.mangatown.com";
diff --git a/src/plugins/Matrix.cpp b/src/plugins/Matrix.cpp
index 5cf4611..d81c53c 100644
--- a/src/plugins/Matrix.cpp
+++ b/src/plugins/Matrix.cpp
@@ -1,6 +1,7 @@
 #include "../../plugins/Matrix.hpp"
 #include "../../include/Storage.hpp"
 #include "../../include/StringUtils.hpp"
+#include "../../include/NetUtils.hpp"
 #include <rapidjson/document.h>
 #include <rapidjson/writer.h>
 #include <rapidjson/stringbuffer.h>
diff --git a/src/plugins/NyaaSi.cpp b/src/plugins/NyaaSi.cpp
index dc6e19f..860eb71 100644
--- a/src/plugins/NyaaSi.cpp
+++ b/src/plugins/NyaaSi.cpp
@@ -3,6 +3,7 @@
 #include "../../include/Storage.hpp"
 #include "../../include/Notification.hpp"
 #include "../../include/StringUtils.hpp"
+#include "../../include/NetUtils.hpp"
 #include <quickmedia/HtmlSearch.h>
 
 namespace QuickMedia {
diff --git a/src/plugins/Plugin.cpp b/src/plugins/Plugin.cpp
index 3f76b4c..0b554ae 100644
--- a/src/plugins/Plugin.cpp
+++ b/src/plugins/Plugin.cpp
@@ -1,67 +1,6 @@
 #include "../../plugins/Plugin.hpp"
-#include "../../include/StringUtils.hpp"
-#include <sstream>
-#include <iomanip>
-#include <array>
 
 namespace QuickMedia {
-    struct HtmlEscapeSequence {
-        char unescape_char;
-        std::string escape_sequence;
-    };
-
-    void html_escape_sequences(std::string &str) {
-        const std::array<HtmlEscapeSequence, 6> escape_sequences = {
-            HtmlEscapeSequence { '&', "&amp;" }, // This should be first, to not accidentally replace a new sequence caused by replacing this
-            HtmlEscapeSequence { '"', "&quot;" },
-            HtmlEscapeSequence { '\'', "&#39;" },
-            HtmlEscapeSequence { '<', "&lt;" },
-            HtmlEscapeSequence { '>', "&gt;" },
-            HtmlEscapeSequence { '\n', "<br>" }
-        };
-
-        for(const HtmlEscapeSequence &escape_sequence : escape_sequences) {
-            string_replace_all(str, escape_sequence.unescape_char, escape_sequence.escape_sequence);
-        }
-    }
-
-    struct HtmlUnescapeSequence {
-        std::string escape_sequence;
-        std::string unescaped_str;
-    };
-
-    void html_unescape_sequences(std::string &str) {
-        const std::array<HtmlUnescapeSequence, 6> unescape_sequences = {
-            HtmlUnescapeSequence { "&quot;", "\"" },
-            HtmlUnescapeSequence { "&#039;", "'" },
-            HtmlUnescapeSequence { "&#39;", "'" },
-            HtmlUnescapeSequence { "&lt;", "<" },
-            HtmlUnescapeSequence { "&gt;", ">" },
-            HtmlUnescapeSequence { "&amp;", "&" } // This should be last, to not accidentally replace a new sequence caused by replacing this
-        };
-
-        for(const HtmlUnescapeSequence &unescape_sequence : unescape_sequences) {
-            string_replace_all(str, unescape_sequence.escape_sequence, unescape_sequence.unescaped_str);
-        }
-    }
-
-    std::string url_param_encode(const std::string &param) {
-        std::ostringstream result;
-        result.fill('0');
-        result << std::hex;
-
-        for(char c : param) {
-            if(isalnum(c) || c == '-' || c == '_' || c == '.' || c == '~') {
-                result << c;
-            } else {
-                result << std::uppercase;
-                result << "%" << std::setw(2) << (int)(unsigned char)(c);
-            }
-        }
-
-        return result.str();
-    }
-
     SuggestionResult download_result_to_suggestion_result(DownloadResult download_result) { return (SuggestionResult)download_result; }
     PluginResult download_result_to_plugin_result(DownloadResult download_result) { return (PluginResult)download_result; }
     SearchResult download_result_to_search_result(DownloadResult download_result) { return (SearchResult)download_result; }
diff --git a/src/plugins/Pornhub.cpp b/src/plugins/Pornhub.cpp
index afdd8fc..e8df9d7 100644
--- a/src/plugins/Pornhub.cpp
+++ b/src/plugins/Pornhub.cpp
@@ -1,5 +1,6 @@
 #include "../../plugins/Pornhub.hpp"
 #include "../../include/StringUtils.hpp"
+#include "../../include/NetUtils.hpp"
 #include <quickmedia/HtmlSearch.h>
 #include <string.h>
 
diff --git a/src/plugins/Youtube.cpp b/src/plugins/Youtube.cpp
index 1711e41..424a8d2 100644
--- a/src/plugins/Youtube.cpp
+++ b/src/plugins/Youtube.cpp
@@ -1,5 +1,6 @@
 #include "../../plugins/Youtube.hpp"
 #include "../../include/Storage.hpp"
+#include "../../include/NetUtils.hpp"
 #include <string.h>
 #include <unordered_set>
 
diff --git a/tests/main.cpp b/tests/main.cpp
new file mode 100644
index 0000000..38dd534
--- /dev/null
+++ b/tests/main.cpp
@@ -0,0 +1,38 @@
+#include <stdio.h>
+#include "../include/NetUtils.hpp"
+
+#define assert_fail(str) do { fprintf(stderr, "Assert failed on line %d, reason: %s\n", __LINE__, (str)); exit(1); } while(0)
+#define assert_equals(a, b) do { if((a) != (b)) { fprintf(stderr, "Assert failed on line %d, %s == %s\n", __LINE__, #a, #b); exit(1); } } while(0)
+
+int main() {
+    std::vector<std::string> urls;
+
+    urls = QuickMedia::extract_urls("example.com");
+    assert_equals(urls.size(), 1);
+    assert_equals(urls[0], "example.com");
+
+    urls = QuickMedia::extract_urls("example.com, is where I like to go");
+    assert_equals(urls.size(), 1);
+    assert_equals(urls[0], "example.com");
+
+    urls = QuickMedia::extract_urls("The website I like to go to is example.com");
+    assert_equals(urls.size(), 1);
+    assert_equals(urls[0], "example.com");
+
+    urls = QuickMedia::extract_urls("example.com. Is also a website");
+    assert_equals(urls.size(), 1);
+    assert_equals(urls[0], "example.com");
+
+    urls = QuickMedia::extract_urls("these. are. not. websites.");
+    assert_equals(urls.size(), 0);
+
+    urls = QuickMedia::extract_urls("This is not an url: example.");
+    assert_equals(urls.size(), 0);
+
+    urls = QuickMedia::extract_urls("the.se/~#423-_/2f.no/3df a.re considered sub.websit.es");
+    assert_equals(urls.size(), 3);
+    assert_equals(urls[0], "the.se/~#423-_/2f.no/3df");
+    assert_equals(urls[1], "a.re");
+    assert_equals(urls[2], "sub.websit.es");
+    return 0;
+}