Have you tried to run source code in UTF8 with BOM at perl under windows? It will not run, even with use utf8 stuff;
It runs OK under *nix environments.
So, solution is using UTF-8 without BOM to make a multi-system portable script | [reply] |
perl foo.pl where foo.pl has bom will work on every platform
./foo.pl if .pl has been assoc-iated with perl will run on win32, and if foo.pl has bom it will be BOM-ed
./foo.pl won't work on unix if foo.pl has bom because shebang won't be detected and perl won't be called, the bom interferes with shebang
editors like notepad/scite... like the BOM ... I like the BOM even though I type in ASCII ... because I copy/paste stuff from the internet, and utf is the only way not to lose info ... most editors don't make files utf by default leading to corruption without BOM
| [reply]
[d/l] |
Funny you waited until now to mention this. You might have mentioned it earlier.
Yes, I run Perl scripts in UTF-8 with BOMs on Windows all the time. It works brilliantly. This is why I recommended it to you in response to your original inquiry: "[H]ow [do I] avoid writing 'use utf8' in every child script…?" Use BOMs instead, that's how. It's the documented way to let the perl interpreter know your script is in the UTF-8 character encoding scheme of Unicode.
If your Unicode Perl scripts don't run on Windows just because they have BOMs in them, then something's wrong. Could it be that you're trying to use Unicode with an ancient and inadequate version of Perl?
Jim
| [reply]
[d/l] |
Hm, according to my testings UTF-8 BOM (on Linux) does not work at all (does not affect how perl behaves). And this note in documentation is just wrong.
| [reply] |
