From f57ff976e5d2de220a60dafc92af28ed3d5223e3 Mon Sep 17 00:00:00 2001
From: tastytea <tastytea@tastytea.de>
Date: Mon, 30 Apr 2018 00:56:55 +0200
Subject: [PATCH] support for hexadecimal numbered entities, 1-8 digits

---
 CMakeLists.txt |  2 +-
 src/parse.cpp  | 18 ++++++++++++++----
 2 files changed, 15 insertions(+), 5 deletions(-)

diff --git a/CMakeLists.txt b/CMakeLists.txt
index deac81a..e8b23a3 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -1,6 +1,6 @@
 cmake_minimum_required (VERSION 3.7)
 project (mastorss
-         VERSION 0.5.12
+         VERSION 0.5.13
          LANGUAGES CXX
 )
 
diff --git a/src/parse.cpp b/src/parse.cpp
index b3fee77..0ef7567 100644
--- a/src/parse.cpp
+++ b/src/parse.cpp
@@ -39,13 +39,23 @@ void unescape_html(string &str)
     string html = str;
     str = "";
     // Used to convert int to utf-8 char
-    std::wstring_convert<std::codecvt_utf8<char16_t>, char16_t> u8c;
-    std::regex reentity("&#(\\d{2,4});");
+    std::wstring_convert<std::codecvt_utf8<char32_t>, char32_t> u8c;
+    std::regex re_entity("&#(x)?(\\d{1,8});");
     std::smatch match;
     
-    while (std::regex_search(html, match, reentity))
+    while (std::regex_search(html, match, re_entity))
     {
-        str += match.prefix().str() + u8c.to_bytes(std::stoi(match[1].str()));
+        char32_t codepoint = 0;
+        // 'x' in front of the number means it's hexadecimal, else decimal.
+        if (match[1].length() == 1)
+        {
+            codepoint = std::stoi(match[2].str(), nullptr, 16);
+        }
+        else
+        {
+            codepoint = std::stoi(match[2].str(), nullptr, 10);
+        }
+        str += match.prefix().str() + u8c.to_bytes(codepoint);
         html = match.suffix().str();
     }
     str += html;