From a83a2548c3d1faab210c9e959a7dc23511a1a301 Mon Sep 17 00:00:00 2001
From: tastytea <tastytea@tastytea.de>
Date: Sat, 13 Jul 2019 18:20:37 +0200
Subject: [PATCH] Remove HTML-encoded newlines in descriptions.

We didn't catch newlines encoded as HTML entities before.
---
 CMakeLists.txt |  2 +-
 src/uri.cpp    | 19 +++++++++++++------
 2 files changed, 14 insertions(+), 7 deletions(-)

diff --git a/CMakeLists.txt b/CMakeLists.txt
index 85a7352..c99330c 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -1,6 +1,6 @@
 cmake_minimum_required (VERSION 3.2)
 project(remwharead
-  VERSION 0.4.2
+  VERSION 0.4.3
   LANGUAGES CXX
 )
 
diff --git a/src/uri.cpp b/src/uri.cpp
index 802c369..2dd9646 100644
--- a/src/uri.cpp
+++ b/src/uri.cpp
@@ -64,7 +64,7 @@ const html_extract URI::get()
             return
                 {
                     extract_title(answer),
-                    strip_html(extract_description(answer)),
+                    extract_description(answer),
                     strip_html(answer)
                 };
         }
@@ -106,7 +106,7 @@ const string URI::extract_description(const string &html)
         smatch match;
         const regex re("description\"[^>]+content=\"([^\"]+)", icase);
         regex_search(html, match, re);
-        return remove_newlines(match[1].str());
+        return remove_newlines(strip_html(match[1].str()));
     }
 
     return "";
@@ -509,11 +509,18 @@ const string URI::archive()
 
 const string URI::remove_newlines(string text)
 {
-    size_t pos = 0;
-    while ((pos = text.find("\n", pos)) != std::string::npos)
+    size_t posn = 0;
+    while ((posn = text.find('\n', posn)) != std::string::npos)
     {
-        text.replace(pos, 1, " ");
-        ++pos;
+        text.replace(posn, 1, " ");
+
+        size_t posr = posn - 1;
+        if (text[posr] == '\r')
+        {
+            text.replace(posr, 1, " ");
+        }
+        ++posn;
     }
+
     return text;
 }