On 5/3/24 04:34, jeremy ardley wrote:
On 3/5/24 19:06, Greg Wooledge wrote:
I would suggest that if you need to use a debugger to track down a bug
in your program, you should use filenames that don't require quoting
when you set up your tests.
1970's style static test cases are not relevant here.
In the real world... I download files generated by another system that
are constantly changing content and with names I don't control.
My workflow is to download a new file from a remote source and then run
my processor over it.
As a necessary consequence I need the fully quoted or escaped file name
of the new file to feed to the processor/debugger.
I can obviously add an extra step to the process to convert the new file
name to something acceptable before processing. However, my question was
how to avoid that extra step by getting fully quoted filenames to process.
So, you are copying and pasting file names via some clipboard? emacs(1)
might have a way to put a filter into that process, but I am unaware of
a similar feature using Xfce and Terminal (my platform).
I have tried renaming files in similar situations, but you will want to
rename them everywhere if you use rsync(1).
What if you downloaded files to a directory with a well-formed name and
added a feature to your script to process files that appear in that
directory?
David
