Put encoder/decoder tables for EFM in private unnamed namespace

Don't know why this wasn't done in the first place, they are implementation details and should not be exposed publicly (even though they're not in the header, I don't even want the symbols available at least not implicitly).

Author: saxbophone

dengr/eight_to_fourteen.cpp

@@ -99,11 +99,9 @@ namespace {
          */
         std::array<PresentByte, 0b10010010010010 + 1> lookup_table;
     };
-}
 
-namespace com::saxbophone::dengr::eight_to_fourteen {
     /**
-     * @brief Lookup table used to encode 8-bit bytes into 14-bit EFM codes
+     * Lookup table used to encode 8-bit bytes into 14-bit EFM codes
      */
     static constexpr std::array<ChannelByte, 256> ENCODING_TABLE = {
         0b01001000100000,
@@ -365,10 +363,12 @@ namespace com::saxbophone::dengr::eight_to_fourteen {
     };
 
     /**
-     * @brief Lookup table used to decode 14-bit EFM codes into 8-bit bytes
+     * Lookup table used to decode 14-bit EFM codes into 8-bit bytes
      */
     static constexpr DecoderLookupTable DECODING_TABLE(ENCODING_TABLE);
+}
 
+namespace com::saxbophone::dengr::eight_to_fourteen {
     ChannelByte encode(Byte byte) {
         /*
          * NOTE: no need to use the bounds-checking interface of std::array here