-# This one needs hardcoded support, since the inputenc package does not know
-# tis620-0, and thailatex sets up babel directly to use tis620-0, so the value
-# for inputenc is never output to .tex files (but needed for the hardcoded
-# tis620-0 support).
-Encoding tis620-0 tis620-0 TIS620-0
+Encoding pt154 pt154 "Cyrillic (pt 154)" PT154 fixed inputenc
+End
+
+Encoding pt254 pt254 "Cyrillic (pt 254)" PT254 fixed inputenc
+End
+
+# encodings used by CJK.sty
+
+# The following encodings that are supported by the CJK package are not
+# included here, because they are not widely used and lack proper iconv support:
+# Bg5+, GBt
+# See the NOTES file of libiconv for details.
+
+# For traditional chinese
+Encoding big5 Bg5 "Chinese (traditional) (Big5)" BIG5 variableunsafe CJK
+End
+
+# For japanese
+# We use CP932 and not SJIS, since it is a super set of SJIS except that it
+# translates SJIS 0x5c to U+005C (REVERSE SOLIDUS aka backslash) instead of
+# U+00A5 (YEN SIGN). Strictly speaking this is wrong, but there is a long
+# standing confusion which translation is correct, and using CP932 is also
+# consistent with the SJIS example file <texmf>/doc/latex/cjk/examples/SJIS.tex
+# of the CJK package. The CP932 extensions over SJIS do not matter here.
+Encoding shift-jis SJIS "Japanese (CJK) (SJIS)" CP932 variableunsafe CJK
+End
+
+# The following encodings need hardcoded support of the encodable unicode
+# range, but are known by iconv:
+
+# For simplified chinese
+Encoding euc-cn GB "Chinese (simplified) (EUC-CN)" EUC-CN variable CJK
+End
+
+# For simplified chinese
+Encoding gbk GBK "Chinese (simplified) (GBK)" GBK variable CJK
+End
+
+# For japanese
+Encoding jis JIS "Japanese (CJK) (JIS)" ISO-2022-JP variable CJK
+End
+
+# For korean
+Encoding euc-kr KS "Korean (EUC-KR)" EUC-KR variable CJK