You could also do it like this,
which does not use dir? and does not
use any extra functions:
dir: %./ ; current directory
result: copy []
dirs: copy []
files: copy []
foreach file read dir [
append either #"/" = last file [dirs][files] file
]
result: compose [(sort dirs) (sort files)]
I have found it handy to have the files and dirs
in separate blocks for other processing, so that
may make this way preferable.
Anton.
> << How do I get the code, listing: sort read %., to sort alphabetically by
> directories first then by filenames? So, the result looks like this:
> [%dir1/ %dir2/ %dir3/ %file1 %file2 %file3] >>
>
> I don't know that there's a single switch to do something like that. Maybe
> someone else knows a nifty trick I don't. In any case, I whipped this up
> real quick to show how you *could* do it. It uses a function I
> have to split
> blocks into sub-blocks based on predicates.
>
> dir-file-sort: func [
> {Returns the block of files with directories first, followed by
> files, with each group sorted.}
> block [any-block!]
> /local result
> ][
> result: copy []
> foreach blk split block [:dir?] none [
> append result sort blk
> ]
> ]
>
> split: func [
> {Applies each predicate to each element in blk and returns
> a block of blocks with items partitioned by which predicate
> they match. If an item doesn't match any predicate, it will
> be in the last block. Only handles simple args right now.}
> series [series!]
> predicates [block!]
> args
> /local match result p
> ][
> result: copy []
> loop add length? predicates 1 [append/only result copy []]
> repeat el series [
> match: false
> repeat i length? predicates [
> ; Have to use a temp var for the predicate here,
> in addition
> to
> ; the result of the predicate call.
> p: get predicates/:i
> match: p :el either block? args [args/:i][args]
> if match [
> append result/:i :el
> break
> ]
> ]
> if not match [
> append last result :el
> ]
> ]
> result
> ]
> ;split [1 2 3.4 5.6 7 8.9] [integer? decimal?] none
> ;split [1 2 3.4 5.6 7 8.9 0 100] [lesser? greater?] [3 7]
> ;split [1 2 3.4 5.6 7 8.9 0 100] [lesser? greater?] 3
>
> I only tested this briefly but it should give you a starting point if
> nothing else.
>
> BIG CAVEAT: DIR? needs full path info if you're working on a list of files
> that aren't in the current directory.
>
> HTH!
>
> --Gregg
--
To unsubscribe from this list, please send an email to
[EMAIL PROTECTED] with "unsubscribe" in the
subject, without the quotes.
