Re: [vhdl-200x] VHDL support for Unicode

On 11/08/11 1:50 AM, Evan Lavelle wrote:

> I did initially go for UTF-32, but it turned out to be just too much work
> to change all the string stuff for wide char handling, and to find all
> the dependencies.

This sounds to be implementation dependent. I was inspired by the C
compiler on a PDP-11 treating a char as an 8 bit value defined in a short
int (16 bit) and thinking about oriental alphabets.

> I had to strip out several hundred warning and error messages and make the
> whole messaging system table-driven, so that messages could be replaced with
> local versions. That was a *lot* of work, and difficult to test.

I'd claim someone not conversant in the target alphabet couldn't test it
adequately. It makes for a good argument for standard error message (and a
very large appendix, too).

I think you could also define a couple of dozen syntax error messages
besides semantic ones, too. I've got 19 in my VHDL-93 lexer.

Someone asked about enumeration names in one Martin's responses for
character types. That'd make one big table and an entry size is based on
the largest enumeration name string.

I do character classifiers algorithmically in macros instead of look up
tables for speed. Historically x86 cache sizes aren't there. any character
value over a certain range can have it's name constructed algorithmically
handling any additional separators or delimiters by exception, I think. You
can't count on character literals for alphabets that aren't required to have
resident fonts. Your error and warning messages would be optional by
supported language, too - which sounds familiar.

-- 
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.