This directory contains encoding maps for some selected encodings. These maps were generated by a perl script, make_encmap, from mapping information available from the Internet at ftp://ftp.unicode.org/Public/MAPPINGS. If you edit the generated XML file to add the "expat='yes'" to the encmap start tag, then you can use the compile_encoding script to check whether the map meets expat requirements (and also create the corresponding binary encmap file.) The file encmap.dtd is the document type declaration for these files and contains information about the semantics. This should give you sufficient information to build your own encoding map. I can't vouch for the validity of the DTD, since I haven't processed it. It is provided for informational purposes only. As mentioned in the DTD, there are some restrictions on what kinds of encodings can be loaded due to restrictions that the expat library places on us for efficiency reasons. One of those restrictions is that the encoding must represent the ASCII set of characters with a single byte and that byte must be equal to the equivalent Unicode scalar value with the exception of a few punctuation characters. This distribution contains four contributed encodings from MURATA Makoto that are variations on the encoding commonly called Shift_JIS: x-sjis-cp932.xml x-sjis-jdk117.xml x-sjis-jisx0221.xml x-sjis-unicode.xml (This is the same encoding as the shift_jis.xml that was distributed with this module in version 1.00) Please read his message (Japanese_Encodings.msg) about why these are here and why I've removed the shift_jis.xml encoding. We also have two contributed encodings that are variations of the EUC-JP encoding from Yoshida Masato : x-euc-jp-jisx0221.xml x-euc-jp-unicode.xml The comments that MURATA Makoto made in his message apply to these encodings too. I've taken the liberty of breaking out common sections for these two as external entities. So the fault of the uninformative names of these four external entites is mine. Clark Cooper December 26, 1998