Re: Fileutils

2009/5/26 Norman Walsh <>:

> Ok. I also proposed a 'create temporary file' step in a follup-up message.

same as 'touch' except specifies a temp file?
I'm easy either way.

>> delete (file|directories [recurse])
>> copy (file|directory [recurse])
>> mv (file|directory)
> I go the file ones. I suppose the directory options make sense.

file or directory, easier if handles either?

>> (I've never used 'is readable|writable|exists - but I can see it being useful)
>> (rename? I'd prefer move)
> The file properties step gives you (a superset?) of what those steps
> return, doesn't it?

mv (as in Unix mv)? Is that included?

>>> Q: Should "file" be made absolute wrt to the current base URI, or left
>>> unchanged (effectively making it relative to the implementations
>>> notion of current working directory)?
>> Any use made of PATH variable?
> No. We're not talking about executables (in general).
>> Treat as per current OS, i.e. could be relative or absolute?
> It can always be absolute, the question is what to do if the author
> puts in a relative name.

Confict? XML/xproc has xml:base (or may do)
My 'OS' centric view expects you to use the current path?
Your choice Norm

>> Using base URI would make it 'odd' wrt other file utils?
> Odd wrt command line file utils, but *not* doing it is odd wrt other
> XProc steps.

Yes, as I said, your choice, but possibly aligned with rest of xproc
would annoy | catch out fewer users.


Dave Pawson
Docbook FAQ.

Received on Tuesday, 26 May 2009 13:50:44 UTC