[Jprogramming] How to translate unicode into ANSI coding?

[bill lam]

Gilles Kirouac wrote:
> In J601 WinXP:
> 
>    a.i.'é'	      NB. e + acute accent
> 195 169
> 
>    a.i.7 u: 'é'
> 233	              NB. correct ANSI code for e acute
> 
>    a.i.7 u: 'àâèêéîïôûü'
> 224 226 232 234 233 238 239 244 251 252	NB.expected ANSI codes
> 
>    $7 u: 'é'
> 1		      NB. expected
>    3!:0 ] 7 u: 'é'
> 131072		      NB. translated, but still 2 bytes per char
> 
> So I am using 
> 
> JtoANSI=: {&a.@:(a.&i.@:(7&u:"1))
> 
> This appears ok for the french accentuated vowels of interest.
> Is it safe?

I'm not sure if code points are translated correctly for french, but you may use
1&u: to discard high byte from 2 byte unicodes. pls see u: documentation.

 a.i. (1&u:)@(7&u:) 'àâèêéîïôûü'
224 226 232 234 233 238 239 244 251 252
 3!:0 (1&u:)@(7&u:) 'àâèêéîïôûü'
2

-- 
regards,
bill