Dear list, I have the following sample: \starttext \startluacode io.write(' Name? ') document.name = io.read() or '' \stopluacode \cldcontext{document.name} is the name. \stoptext Running it on Linux, I can input non-ascii characters. When running in Windows, if the input chars contains a non-ascii one, document.name is empty. I have no problem passing arguments from cmd with Unicode characters (such as in '--arguments="name={αβγ}"'. I’m using current latest (ConTeXt MkIV 2020.11.08 12:42). I’m not sure what I am missing or whether I have hit a bug. Many thanks for your help, Pablo -- http://www.ousia.tk
On 11/12/2020 10:39 PM, Pablo Rodriguez wrote:
Dear list,
I have the following sample:
\starttext \startluacode io.write(' Name? ') document.name = io.read() or '' \stopluacode \cldcontext{document.name} is the name. \stoptext
Running it on Linux, I can input non-ascii characters.
When running in Windows, if the input chars contains a non-ascii one, document.name is empty.
I have no problem passing arguments from cmd with Unicode characters (such as in '--arguments="name={αβγ}"'.
I’m using current latest (ConTeXt MkIV 2020.11.08 12:42).
I’m not sure what I am missing or whether I have hit a bug.
This has been discussed before I think. It has to do with how you configured your system (what encoding) and how consisteltly you take that into account. If you have some mixed setup, just don't use non-ascii. Anyway, in LMTX all file, commandline and systemn operations are utf 8 and on windows get translated into wide system calls so there it should work ok if you use utf8. Hans ----------------------------------------------------------------- Hans Hagen | PRAGMA ADE Ridderstraat 27 | 8061 GH Hasselt | The Netherlands tel: 038 477 53 69 | www.pragma-ade.nl | www.pragma-pod.nl -----------------------------------------------------------------
On 11/12/20 10:55 PM, Hans Hagen wrote:
On 11/12/2020 10:39 PM, Pablo Rodriguez wrote:
Dear list,
I have the following sample:
\starttext \startluacode io.write(' Name? ') document.name = io.read() or '' \stopluacode \cldcontext{document.name} is the name. \stoptext [...] I’m using current latest (ConTeXt MkIV 2020.11.08 12:42).
I’m not sure what I am missing or whether I have hit a bug.
This has been discussed before I think. It has to do with how you configured your system (what encoding) and how consisteltly you take that into account. If you have some mixed setup, just don't use non-ascii.
Many thanks for your reply, Hans. I use chcp 65001. I have no problem displaying UTF-8 messages and passing arguments with UTF-8 characters.
Anyway, in LMTX all file, commandline and systemn operations are utf 8 and on windows get translated into wide system calls so there it should work ok if you use utf8.
Well, according to https://ss64.com/nt/cmd.html the console uses UTF-16LE. Or UCS-2 (as described in https://devblogs.microsoft.com/commandline/windows-command-line-unicode-and-...). Since console input is UTF-16LE, I guess I may need something in Lua similar to 'regimes.translate(str, "utf16le")'. How can I convert a string from UTF-16LE (into UTF-8)? Many thanks for your help, Pablo -- http://www.ousia.tk
On 11/12/20 10:55 PM, Hans Hagen wrote:
On 11/12/2020 10:39 PM, Pablo Rodriguez wrote:
Dear list,
I have the following sample:
\starttext \startluacode io.write(' Name? ') document.name = io.read() or '' \stopluacode \cldcontext{document.name} is the name. \stoptext [...] I’m using current latest (ConTeXt MkIV 2020.11.08 12:42).
I’m not sure what I am missing or whether I have hit a bug.
This has been discussed before I think. It has to do with how you configured your system (what encoding) and how consisteltly you take that into account. If you have some mixed setup, just don't use non-ascii.
Many thanks for your reply, Hans.
I use chcp 65001. I have no problem displaying UTF-8 messages and passing arguments with UTF-8 characters.
Anyway, in LMTX all file, commandline and systemn operations are utf 8 and on windows get translated into wide system calls so there it should work ok if you use utf8.
Well, according to https://ss64.com/nt/cmd.html the console uses UTF-16LE. Or UCS-2 (as described in https://devblogs.microsoft.com/commandline/windows-command-line-unicode-and-...).
Since console input is UTF-16LE, I guess I may need something in Lua similar to 'regimes.translate(str, "utf16le")'.
On 11/13/2020 2:49 PM, Pablo Rodriguez wrote: the console uses whatever code page you have configured and it also relates to the code page used for filenames .. the 16 bit values are used deep down and what you use gets translated into that (often folks still use some 8 bit code page so that then gets mapped) .. there is no way the system can know if what you provide is as it's bytes in whatever encoding used so, you need to look what your system is configured for (this is independent of the output to the console which is what the 65001 does) Hans ----------------------------------------------------------------- Hans Hagen | PRAGMA ADE Ridderstraat 27 | 8061 GH Hasselt | The Netherlands tel: 038 477 53 69 | www.pragma-ade.nl | www.pragma-pod.nl -----------------------------------------------------------------
On 11/13/20 3:15 PM, Hans Hagen wrote:
On 11/13/2020 2:49 PM, Pablo Rodriguez wrote:
[...] Since console input is UTF-16LE, I guess I may need something in Lua similar to 'regimes.translate(str, "utf16le")'.
the console uses whatever code page you have configured and it also relates to the code page used for filenames .. the 16 bit values are used deep down and what you use gets translated into that (often folks still use some 8 bit code page so that then gets mapped) .. there is no way the system can know if what you provide is as it's bytes in whatever encoding used
so, you need to look what your system is configured for
I think that codepage is named locale codepage (as different from the one in the console [or whatever chcp should change]). I wonder whether this is named codeset in this Windows tool:
wmic os get locale, oslanguage, codeset CodeSet Locale OSLanguage 1252 0c0a 3082
(this is independent of the output to the console which is what the 65001 does)
It seems that UTF-8 as locale codepage is only available in Win10. An easy workaround is to rewrite the batch file to: chcp 65001 set /P "name=Name? " context --arguments="name={%name%}" document.tex Many thanks for your help, Pablo -- http://www.ousia.tk
participants (2)
-
Hans Hagen -
Pablo Rodriguez