diff --git a/src/adoc.cpp b/src/adoc.cpp
index 813cfcf..f59ca80 100644
--- a/src/adoc.cpp
+++ b/src/adoc.cpp
@@ -29,183 +29,186 @@ using std::regex;
 using std::regex_replace;
 using tagpair = std::pair<string,vector<Database::entry>>;
 
-void export_adoc(const vector<Database::entry> &entries, ostream &out)
+namespace Export
 {
-    try
+    void AsciiDoc::print() const
     {
-        out << "= Visited things\n"
-            << ":Author:    remwharead " << global::version << endl
-            << ":Date:      "
-            << timepoint_to_string(system_clock::now()) << endl
-            << ":TOC:       right" << endl
-            << ":TOCLevels: 2" << endl
-            << endl;
-
-        std::map<string,vector<Database::entry>> alltags;
-        string day;
-        for (const Database::entry &entry : entries)
+        try
         {
-            const string datetime = timepoint_to_string(entry.datetime);
-            const string newday = datetime.substr(0, datetime.find('T'));
-            const string time = datetime.substr(datetime.find('T') + 1);
-            if (newday != day)
-            {
-                day = newday;
-                out << "== " << day << endl << endl;
-            }
+            _out << "= Visited things\n"
+                 << ":Author:    remwharead " << global::version << endl
+                 << ":Date:      "
+                 << timepoint_to_string(system_clock::now()) << endl
+                 << ":TOC:       right" << endl
+                 << ":TOCLevels: 2" << endl
+                 << endl;
 
-            out << "[[dt_" << datetime << "]]\n";
-            out << "* link:" << entry.uri;
-            if (!entry.title.empty())
+            std::map<string,vector<Database::entry>> alltags;
+            string day;
+            for (const Database::entry &entry : _entries)
             {
-                out << '[' << entry.title << ']';
-            }
-            else
-            {
-                out << "[]";
-            }
-            out << " +" << endl;
-
-            out << '_' << time.substr(0, 5) << '_';
-            if (!entry.archive_uri.empty())
-            {
-                out << " (link:" << entry.archive_uri << "[archived version])";
-            }
-
-            bool separator = false;
-            for (const string &tag : entry.tags)
-            {
-                if (tag.empty())
+                const string datetime = timepoint_to_string(entry.datetime);
+                const string newday = datetime.substr(0, datetime.find('T'));
+                const string time = datetime.substr(datetime.find('T') + 1);
+                if (newday != day)
                 {
-                    continue;
-                }
-                if (!separator)
-                {
-                    out << "\n| ";
-                    separator = true;
+                    day = newday;
+                    _out << "== " << day << endl << endl;
                 }
 
-                auto globaltag = alltags.find(tag);
-                if (globaltag != alltags.end())
+                _out << "[[dt_" << datetime << "]]\n";
+                _out << "* link:" << entry.uri;
+                if (!entry.title.empty())
                 {
-                    globaltag->second.push_back(entry);
+                    _out << '[' << entry.title << ']';
                 }
                 else
                 {
-                    alltags.insert({ tag, { entry } });
+                    _out << "[]";
                 }
+                _out << " +" << endl;
 
-                out << "xref:t_" << replace_in_tags(tag) << "[" << tag << ']';
-                if (tag != *(entry.tags.rbegin()))
+                _out << '_' << time.substr(0, 5) << '_';
+                if (!entry.archive_uri.empty())
                 {
-                    out << ", ";
+                    _out << " (link:" << entry.archive_uri << "[archived version])";
                 }
+
+                bool separator = false;
+                for (const string &tag : entry.tags)
+                {
+                    if (tag.empty())
+                    {
+                        continue;
+                    }
+                    if (!separator)
+                    {
+                        _out << "\n| ";
+                        separator = true;
+                    }
+
+                    auto globaltag = alltags.find(tag);
+                    if (globaltag != alltags.end())
+                    {
+                        globaltag->second.push_back(entry);
+                    }
+                    else
+                    {
+                        alltags.insert({ tag, { entry } });
+                    }
+
+                    _out << "xref:t_" << replace_in_tags(tag) << "[" << tag << ']';
+                    if (tag != *(entry.tags.rbegin()))
+                    {
+                        _out << ", ";
+                    }
+                }
+
+                if (!entry.description.empty())
+                {
+                    _out << " +" << endl << entry.description;
+                }
+                _out << endl << endl;
             }
 
-            if (!entry.description.empty())
+            if (!alltags.empty())
             {
-                out << " +" << endl << entry.description;
+                print_tags(alltags);
             }
-            out << endl << endl;
         }
-
-        if (!alltags.empty())
+        catch (std::exception &e)
         {
-            adoc_print_tags(alltags, out);
+            cerr << "Error in " << __func__ << ": " << e.what() << endl;
         }
     }
-    catch (std::exception &e)
+
+    const string AsciiDoc::replace_in_tags(string text) const
     {
-        cerr << "Error in " << __func__ << ": " << e.what() << endl;
+        // TODO: Find a better solution.
+        const std::map<const string, const string> searchreplace =
+            {
+                { " ", "-" }, { "§", "-" },
+                { "$", "-" }, { "%", "-" },
+                { "&", "-" }, { "/", "-" },
+                { "=", "-" }, { "^", "-" },
+                { "!", "-" }, { "?", "-" },
+                { "₀", "0" }, { "⁰", "0" },
+                { "₁", "1" }, { "¹", "1" },
+                { "₂", "2" }, { "²", "2" },
+                { "₃", "3" }, { "³", "3" },
+                { "₄", "4" }, { "⁴", "4" },
+                { "₅", "5" }, { "⁵", "5" },
+                { "₆", "6" }, { "⁶", "6" },
+                { "₇", "7" }, { "⁷", "7" },
+                { "₈", "8" }, { "⁸", "8" },
+                { "₉", "9" }, { "⁹", "9" }
+            };
+        for (const std::pair<const string, const string> &sr : searchreplace)
+        {
+            size_t pos = 0;
+            while ((pos = text.find(sr.first, pos)) != std::string::npos)
+            {
+                text.replace(pos, sr.first.length(), sr.second);
+            }
+        }
+        return text;
+    }
+
+    void AsciiDoc::print_tags(
+        const std::map<string,vector<Database::entry>> &tags) const
+    {
+        _out << "== Tags\n\n";
+        vector<tagpair> sortedtags(tags.size());
+        std::move(tags.begin(), tags.end(), sortedtags.begin());
+        std::sort(sortedtags.begin(), sortedtags.end(),
+                  [](const tagpair &a, tagpair &b)
+                  {
+                      if (a.second.size() != b.second.size())
+                      {   // Sort by number of occurrences if they are different.
+                          return a.second.size() > b.second.size();
+                      }
+                      else
+                      {   // Sort by tag names otherwise.
+                          std::locale loc;
+                          const std::collate<char> &coll =
+                              std::use_facet<std::collate<char>>(loc);
+                          return (coll.compare(
+                                      a.first.data(), a.first.data()
+                                      + a.first.length(),
+                                      b.first.data(), b.first.data()
+                                      + b.first.length()) == -1);
+                      }
+                  });
+
+        bool othertags = false;
+        for (const auto &tag : sortedtags)
+        {
+            if (tag.second.size() == 1)
+            {
+                if (!othertags)
+                {
+                    _out << "=== Less used tags\n\n";
+                    othertags = true;
+                }
+                _out << "=";
+            }
+
+            _out << "=== [[t_" << replace_in_tags(tag.first) << "]]"
+                << tag.first << endl;
+            for (const Database::entry &entry : tag.second)
+            {
+                const string datetime = timepoint_to_string(entry.datetime);
+                const string date = datetime.substr(0, datetime.find('T'));
+                string title = entry.title;
+                if (title.empty())
+                {
+                    title = "++" + entry.uri + "++";
+                }
+                _out << endl << "* xref:dt_" << datetime
+                     << '[' << title << "] _(" << date << ")_" << endl;
+            }
+            _out << endl;
+        }
+        _out << endl;
     }
 }
-
-const string replace_in_tags(string text)
-{
-    // TODO: Find a better solution.
-    const std::map<const string, const string> searchreplace =
-        {
-            { " ", "-" }, { "§", "-" },
-            { "$", "-" }, { "%", "-" },
-            { "&", "-" }, { "/", "-" },
-            { "=", "-" }, { "^", "-" },
-            { "!", "-" }, { "?", "-" },
-            { "₀", "0" }, { "⁰", "0" },
-            { "₁", "1" }, { "¹", "1" },
-            { "₂", "2" }, { "²", "2" },
-            { "₃", "3" }, { "³", "3" },
-            { "₄", "4" }, { "⁴", "4" },
-            { "₅", "5" }, { "⁵", "5" },
-            { "₆", "6" }, { "⁶", "6" },
-            { "₇", "7" }, { "⁷", "7" },
-            { "₈", "8" }, { "⁸", "8" },
-            { "₉", "9" }, { "⁹", "9" }
-        };
-    for (const std::pair<const string, const string> &sr : searchreplace)
-    {
-        size_t pos = 0;
-        while ((pos = text.find(sr.first, pos)) != std::string::npos)
-        {
-            text.replace(pos, sr.first.length(), sr.second);
-        }
-    }
-    return text;
-}
-
-void adoc_print_tags(const std::map<string,vector<Database::entry>> &tags,
-                     ostream &out)
-{
-    out << "== Tags\n\n";
-    vector<tagpair> sortedtags(tags.size());
-    std::move(tags.begin(), tags.end(), sortedtags.begin());
-    std::sort(sortedtags.begin(), sortedtags.end(),
-              [](const tagpair &a, tagpair &b)
-              {
-                  if (a.second.size() != b.second.size())
-                  {   // Sort by number of occurrences if they are different.
-                      return a.second.size() > b.second.size();
-                  }
-                  else
-                  {   // Sort by tag names otherwise.
-                      std::locale loc;
-                      const std::collate<char> &coll =
-                          std::use_facet<std::collate<char>>(loc);
-                      return (coll.compare(
-                                  a.first.data(), a.first.data()
-                                  + a.first.length(),
-                                  b.first.data(), b.first.data()
-                                  + b.first.length()) == -1);
-                  }
-              });
-
-    bool othertags = false;
-    for (const auto &tag : sortedtags)
-    {
-        if (tag.second.size() == 1)
-        {
-            if (!othertags)
-            {
-                out << "=== Less used tags\n\n";
-                othertags = true;
-            }
-            out << "=";
-        }
-
-        out << "=== [[t_" << replace_in_tags(tag.first) << "]]"
-            << tag.first << endl;
-        for (const Database::entry &entry : tag.second)
-        {
-            const string datetime = timepoint_to_string(entry.datetime);
-            const string date = datetime.substr(0, datetime.find('T'));
-            string title = entry.title;
-            if (title.empty())
-            {
-                title = "++" + entry.uri + "++";
-            }
-            out << endl << "* xref:dt_" << datetime
-                << '[' << title << "] _(" << date << ")_" << endl;
-        }
-        out << endl;
-    }
-    out << endl;
-}
diff --git a/src/export.hpp b/src/export.hpp
index d50a947..6d4dd2e 100644
--- a/src/export.hpp
+++ b/src/export.hpp
@@ -45,18 +45,27 @@ namespace Export
         using ExportBase::ExportBase;
 
         void print() const;
+
+    private:
         //! replaces " with "".
         const string quote(string field) const;
     };
+
+    class AsciiDoc : protected ExportBase
+    {
+    public:
+        using ExportBase::ExportBase;
+
+        void print() const;
+
+    private:
+        //! Replaces characters in tags that asciidoctor doesn't like.
+        const string replace_in_tags(string text) const;
+        void print_tags(const std::map<string,vector<Database::entry>> &tags) const;
+    };
 }
 
 
-void export_adoc(const vector<Database::entry> &entries, ostream &out = cout);
-//! Replaces characters in tags that asciidoctor doesn't like.
-const string replace_in_tags(string text);
-void adoc_print_tags(const std::map<string,vector<Database::entry>> &tags,
-                     ostream &out = cout);
-
 
 //! Export as Netscape bookmark file.
 void export_bookmarks(const vector<Database::entry> &entries,
diff --git a/src/main.cpp b/src/main.cpp
index 23fee28..ddd9a46 100644
--- a/src/main.cpp
+++ b/src/main.cpp
@@ -111,12 +111,12 @@ int main(const int argc, const char *argv[])
         {
             if (file.is_open())
             {
-                export_adoc(entries, file);
+                Export::AsciiDoc(entries, file).print();
                 file.close();
             }
             else
             {
-                export_adoc(entries);
+                Export::AsciiDoc(entries).print();
             }
             break;
         }
diff --git a/tests/test_adoc.cpp b/tests/test_adoc.cpp
index 436c6de..0e25ba9 100644
--- a/tests/test_adoc.cpp
+++ b/tests/test_adoc.cpp
@@ -47,7 +47,7 @@ SCENARIO ("The AsciiDoc export works correctly")
         try
         {
             std::ostringstream output;
-            export_adoc({ entry }, output);
+            Export::AsciiDoc({ entry }, output).print();
             const string adoc = output.str();
 
             const regex re_header