Kai Grossjohann <[EMAIL PROTECTED]> writes:

Hi Kai,

> Checking file-name-all-completions is quite a cool idea.  Just for
> completeness' sake, let me propose an opposite approach: If
> file-name-all-completions is invoked again with the same arguments very
> quickly, then reload the cache.

Something similar was my first try. But it is more complex: if you have
a call of (file-name-all-completions "f" "/remote.host:/bar/"), and the
next call is (file-name-all-completions "foo" "/remote.host:/bar/"), you
should obviously NOT flush the cache for "/bar/". You would need a
book-keeping depending on both parameters, instead of only the directory.

Additionally, two subsequent calls of file-name-all-completions with
both identical parameters do not happen. This is already prevented in
the superior file name completion function (checked with pcomplete.el,
but the other packages might behave similar).

That's why I gave it up this way, and I've offered the other solution.

> Kai

Best regards, Michael.


_______________________________________________
Tramp-devel mailing list
[email protected]
http://lists.gnu.org/mailman/listinfo/tramp-devel

Reply via email to