Re: Input methods at the age of Unicode

From: Marcel Schneider <>
Date: Thu, 16 Jul 2015 11:21:23 +0200 (CEST)

On 16 Jul 2015, at 10:35, Hans Aberg wrote:

> One still has to figure out a good map.
> Using Unicode helps the readability of the input file, though. One can use for example ConTeXt with LuaLaTeX, which comes with the TeX live installation.

Thank you very much for these hints, I'll try to apply them. Actually I stick with a rather common set of characters on the key map except that I've added U+2610, which is very useful, even more when it's a part of the dead lists as a base character, and several additional exotic currency symbols as a mark of respect. Backwards compatibility leads to limit the number of key positions. From eight per key I've come back down to four, and from a dozen or more dead keys (and a maximum of about twenty-five or thirty) back to five plus the Compose key (one key with four dead key positions: Compose, AltGr, Greek, Secondary group—with respect to ISO 9995). But with one Compose key we've potentially as many dead keys as there are key positions on the rest of the keyboard, and each one of them can give access to as many again. I believe that the future of keyboards is as well in the Compose tree as in the key map, or even more.

The file format of my source files is UTF-8, however the compiler admits clear characters only up to U+008F. From U+00A0 upwards, we must use code points. For readability I add Unicode characters in the trailing comments, as well as automatically added Unicode character identifiers (names), along with as much comments as we want. Doing all in spreadsheets allows to semi-automatically derive HTML tables without needing any other software than a text editor.

Now I've just downloaded the two versions of ConTEXT, which might well be the enhanced text editor I'm looking for since a while. LuaLaTeX will be very interesting too if I can edit source files with (however the bulk job is done in spreadsheet software which is Unicode; actual versions include even the UNICAR and UNICODE functions). I'll try if ConTeXt recognizes the Kana shift states (Gedit seemingly doesn't).

Have a great day,

Received on Thu Jul 16 2015 - 04:23:11 CDT

This archive was generated by hypermail 2.2.0 : Thu Jul 16 2015 - 04:23:11 CDT