|
steve donovan wrote:
On Tue, Nov 2, 2010 at 12:40 PM, Lorenzo Donati <lorenzodonatibz@interfree.it> wrote:Well, I never tried parsing dir /B, so I suppose there could be problem only if filenames contained 8-bit non-ASCII chars.And then you might get the mess Gmail made of your accented characters...
:-)
Well, at that time there wasn't (given the constraint I was working under). I had to process linguistic data stored in XML files whose content was briefly described by their name in an encoded fashion (not my choice). So the file name was utterly meaningful (it was a database entry in itself, whose fields were separated by dots!!!). Moreover the linguists working on those files (hand editing their content) expected to type those names as parameters of the command line tools. And to this add that "the management" refused to gave me allowance to create a small Java GUI to bypass the WinXP shell, and ease other problems caused by the interaction of a bunch of command line tools written in different languages (Perl, .CMD shell scripts, Java).with setting the codepage to a Reader class in Java, but discovering the codepage code was a matter of trial and error (had to try more than 20something codepages!!!)There must be a simpler solution! But, maybe there isn't?
You know: no time, no money ... sometimes no brain ;-) :-)
http://github.com/stevedonovan/LakeAnd does it allow installation without luarocks?No luarocks dependency, just a single monster 2,000 line script. But beta is a good description for now ... of course, it will not move out of beta until I can find users to kick it around the block and abuse it constructively.
well, If I find some spare time, I will gladly give it a try.