1 files changed, 17 insertions, 55 deletions

diff --git a/src/charset.h b/src/charset.h
index 3b5f102..8476ee7 100644
--- a/src/charset.h
+++ b/src/charset.h
@@ -1,5 +1,5 @@
 /*
-    Copyright (C) 2015 Matthias P. Braendli (http://opendigitalradio.org)
+    Copyright (C) 2018 Matthias P. Braendli (http://opendigitalradio.org)
 
     This program is free software: you can redistribute it and/or modify
     it under the terms of the GNU General Public License as published by
@@ -16,75 +16,37 @@
 */
 /*!
     \file charset.h
-    \brief Define the EBU charset according to ETSI TS 101 756v1.8.1 for DLS encoding
+    \brief A converter for UTF-8 to EBU Latin charset according to
+           ETSI TS 101 756 Annex C, used for DLS and Labels.
 
-    \author Matthias P. Braendli <matthias@mpb.li>
+    \author Matthias P. Braendli
     \author Lindsay Cornell
 */
 
-#ifndef __CHARSET_H_
-#define __CHARSET_H_
+#pragma once
 
-#include "common.h"
-
-#include "utf8.h"
+#include <cstdint>
 #include <string>
 #include <vector>
-#include <algorithm>
-
-#define CHARSET_TABLE_OFFSET 1 // NUL at index 0 cannot be represented
-#define CHARSET_TABLE_ENTRIES (256 - CHARSET_TABLE_OFFSET)
-extern const char* utf8_encoded_EBU_Latin[CHARSET_TABLE_ENTRIES];
+#include "utf8.h"
 
 class CharsetConverter
 {
     public:
-        CharsetConverter() {
-            /*! Build the converstion table that contains the known code points,
-             * at the indices corresponding to the EBU Latin table
-             */
-            using namespace std;
-            for (size_t i = 0; i < CHARSET_TABLE_ENTRIES; i++) {
-                string table_entry(utf8_encoded_EBU_Latin[i]);
-                string::iterator it = table_entry.begin();
-                uint32_t code_point = utf8::next(it, table_entry.end());
-                m_conversion_table.push_back(code_point);
-            }
-        }
+        CharsetConverter();
 
-        /*! Convert a UTF-8 encoded text line into an EBU Latin encoded byte stream
+        /*! Convert a UTF-8 encoded text line into an EBU Latin encoded byte
+         *  stream. If up_to_first_error is set, convert as much text as possible.
+         *  If false, raise an utf8::exception in case of conversion errors.
          */
-        std::string convert(std::string line_utf8) {
-            using namespace std;
-
-            // check for invalid utf-8, we only convert up to the first error
-            string::iterator end_it = utf8::find_invalid(line_utf8.begin(), line_utf8.end());
-
-            // Convert it to utf-32
-            vector<uint32_t> utf32line;
-            utf8::utf8to32(line_utf8.begin(), end_it, back_inserter(utf32line));
+        std::string convert(std::string line_utf8, bool up_to_first_error = true);
 
-            string encoded_line(utf32line.size(), '0');
-
-            // Try to convert each codepoint
-            for (size_t i = 0; i < utf32line.size(); i++) {
-                vector<uint32_t>::iterator iter = find(m_conversion_table.begin(),
-                        m_conversion_table.end(), utf32line[i]);
-                if (iter != m_conversion_table.end()) {
-                    size_t index = std::distance(m_conversion_table.begin(), iter);
-
-                    encoded_line[i] = (char)(index + CHARSET_TABLE_OFFSET);
-                }
-                else {
-                    encoded_line[i] = ' ';
-                }
-            }
-            return encoded_line;
-        }
+        /*! Convert a EBU Latin byte stream to a UTF-8 encoded string.
+         *  Invalid input characters are converted to ⁇ (unicode U+2047).
+         */
+        std::string convert_ebu_to_utf8(const std::string& str);
 
     private:
-
+        // Representation of the table in 32-bit unicode
         std::vector<uint32_t> m_conversion_table;
 };
-
-#endif