«

New user question:

It seems to me when I run:
Find.find('/user/name/documents') {|path| puts path}

it returns all the directories in the reverse order. I was expecting
the directories to be returned in alphabetical order, but that doesn't
seem to be the case. Also, in one case it reads half of a directory's
files, then the sub dirs and then it finished reading the rest of the
directory it started in and finished writing them in the Puts
statement.

Am I missing something? How do you do get it to write out the
directories in alphabetical order?

Any and all help welcome.

Thank you.

Brad

On Mar 9, 2007, at 7:35 PM, Brad wrote:

> New user question:
>
> It seems to me when I run:
> Find.find('/user/name/documents') {|path| puts path}
>
> it returns all the directories in the reverse order. I was expecting
> the directories to be returned in alphabetical order, but that doesn't
> seem to be the case. Also, in one case it reads half of a directory's
> files, then the sub dirs and then it finished reading the rest of the
> directory it started in and finished writing them in the Puts
> statement.
>
> Am I missing something? How do you do get it to write out the
> directories in alphabetical order?
>
> Any and all help welcome.

That's just the way Fine.find works. My experience is that, for most
applications, the order in which Find.find traverses directories
doesn't cause problems. But it may not be the best way to get a deep
list of the contents of a directory. Two alternatives you might try are

BASE = '/Users/mg/Downloads'

paths = []
Find.find(BASE) { |path| paths << path }
puts paths.reverse

or

Dir.glob("#{BASE}/**/*") { |path| puts path }

Note: this form of Dir.glob doesn't find dotted (hidden) files and
Fine.find does (there is a flag you can set to get dotted files with
Dir.glob).

Regards, Morton

On 9 mar, 21:33, "Brad" <bradask...@gmail.com> wrote:
> New user question:
>
> Am I missing something? How do you do get it to write out the
> directories in alphabetical order?
>

You are probably using the wrong api. If you want to get a list of
files/directories in a single directory, you should use:
Dir.entries()
or
Dir.foreach() for more complex iterations.

Find does a recursive search across multiple directories (and
subdirectories).

That being said, files are read and returned as your OS does. So it
is up to you to sort them out into directories and not directories and
to sort them out. You can usually do that with arrays.

dir = Dir.pwd

#
# dirs and files sorted together
#
puts Dir.entries(dir).sort

#
# or... more complex...
#
dirs = []
files = []
Dir.foreach(dir) do |file|
next if file =~ /^\..?$/ # ignore . and ..
if File.directory?("#{dir}/#{file}")
dirs << file
else
files << file
end
end
dirs.sort!
files.sort!
puts dirs, "---", files

#
# or same, but simpler (more ruby-like)
#
all = Dir.entries(dir) - ['.', '..']
all.sort!
dirs = all.select { |x| File.directory?("#{dir}/#{x}") }
files = all - dirs
puts dirs, "---", files

You could also use a system call to the tools "which" or "locate" or
"find" on *nix systems and pipe that to where you need it. Very DRY!
On Mar 10, 2007, at 4:44 PM, Morton Goldberg wrote:

> On Mar 9, 2007, at 7:35 PM, Brad wrote:
>
>> New user question:
>>
>> It seems to me when I run:
>> Find.find('/user/name/documents') {|path| puts path}
>>
>> it returns all the directories in the reverse order. I was expecting
>> the directories to be returned in alphabetical order, but that
>> doesn't
>> seem to be the case. Also, in one case it reads half of a
>> directory's
>> files, then the sub dirs and then it finished reading the rest of the
>> directory it started in and finished writing them in the Puts
>> statement.
>>
>> Am I missing something? How do you do get it to write out the
>> directories in alphabetical order?
>>
>> Any and all help welcome.
>
> That's just the way Fine.find works. My experience is that, for
> most applications, the order in which Find.find traverses
> directories doesn't cause problems. But it may not be the best way
> to get a deep list of the contents of a directory. Two alternatives
> you might try are
>
> BASE = '/Users/mg/Downloads'
>
> paths = []
> Find.find(BASE) { |path| paths << path }
> puts paths.reverse
>
> or
>
> Dir.glob("#{BASE}/**/*") { |path| puts path }
>
> Note: this form of Dir.glob doesn't find dotted (hidden) files and
> Fine.find does (there is a flag you can set to get dotted files
> with Dir.glob).
>
> Regards, Morton
>

Here's a little something I happened to stumble upon yesterday though. I
think these are pretty sound arguments against calling external
commands, when you have the choice. Might want to take them into
consideration.

http://www.caliban.org/ruby/rubyguide.shtml#external



On Sat, 2007-03-10 at 18:10 +0900, John Joyce wrote:
> You could also use a system call to the tools "which" or "locate" or
> "find" on *nix systems and pipe that to where you need it. Very DRY!
> On Mar 10, 2007, at 4:44 PM, Morton Goldberg wrote:
>
> > On Mar 9, 2007, at 7:35 PM, Brad wrote:
> >
> >> New user question:
> >>
> >> It seems to me when I run:
> >> Find.find('/user/name/documents') {|path| puts path}
> >>
> >> it returns all the directories in the reverse order. I was expecting
> >> the directories to be returned in alphabetical order, but that
> >> doesn't
> >> seem to be the case. Also, in one case it reads half of a
> >> directory's
> >> files, then the sub dirs and then it finished reading the rest of the
> >> directory it started in and finished writing them in the Puts
> >> statement.
> >>
> >> Am I missing something? How do you do get it to write out the
> >> directories in alphabetical order?
> >>
> >> Any and all help welcome.
> >
> > That's just the way Fine.find works. My experience is that, for
> > most applications, the order in which Find.find traverses
> > directories doesn't cause problems. But it may not be the best way
> > to get a deep list of the contents of a directory. Two alternatives
> > you might try are
> >
> > BASE = '/Users/mg/Downloads'
> >
> > paths = []
> > Find.find(BASE) { |path| paths << path }
> > puts paths.reverse
> >
> > or
> >
> > Dir.glob("#{BASE}/**/*") { |path| puts path }
> >
> > Note: this form of Dir.glob doesn't find dotted (hidden) files and
> > Fine.find does (there is a flag you can set to get dotted files
> > with Dir.glob).
> >
> > Regards, Morton
> >
>
>

Indeed! Those should always be considerations when making calls to
external binaries. But given those considerations, no need to repeat
things. Fact is, many things are ported easily because of such
binaries being pretty standard on *nix systems.
You could also always bundle those with it (if the licensing allows).
Point is, if you don't have to write something, don't do it.
I'm talking about duct tape, pretty much like what Perl is often used
for. (no offense to anyone, but I couldn't care less what anyone does
with windows, this sort of thing will work on most *nixes and a
windows DOS command should exist as well, it's part of porting. At
some level there are dependencies to care about and you can't support
ever version of every platform, it's just impossible.)
As for being a security risk, well, if you have Ruby executables then
you already have the same level of risk as calling any binary on the
system by any other means. You can't protect people from themselves.
Installers do this stuff all the time.
On Mar 10, 2007, at 6:20 PM, Choong Wei Tjeng wrote:

> Here's a little something I happened to stumble upon yesterday
> though. I
> think these are pretty sound arguments against calling external
> commands, when you have the choice. Might want to take them into
> consideration.
>
> http://www.caliban.org/ruby/rubyguide.shtml#external

This link is for anyone new to Ruby. It's incredibly concise and well
written. Apparently it was an internal style guide at Google.
Whatever. It's a pretty good quick intro to what's in store for you
in Ruby. You can use it as a gage of what you need to know.
Especially useful if you've come from Perl.

http://www.caliban.org/ruby/rubyguide.shtml

(Choong Wei gave this link in a different thread)

On 10.03.2007 01:33, Brad wrote:
> It seems to me when I run:
> Find.find('/user/name/documents') {|path| puts path}
>
> it returns all the directories in the reverse order. I was expecting
> the directories to be returned in alphabetical order, but that doesn't
> seem to be the case. Also, in one case it reads half of a directory's
> files, then the sub dirs and then it finished reading the rest of the
> directory it started in and finished writing them in the Puts
> statement.
>
> Am I missing something? How do you do get it to write out the
> directories in alphabetical order?

As others said already, the order is dictated by how the OS returns
entries. Here's another solution:

puts Dir['/user/name/documents/**/*'].sort

Kind regards

robert

Hi --

On 3/10/07, John Joyce <> wrote:
> This link is for anyone new to Ruby. It's incredibly concise and well
> written. Apparently it was an internal style guide at Google.
> Whatever. It's a pretty good quick intro to what's in store for you
> in Ruby. You can use it as a gage of what you need to know.
> Especially useful if you've come from Perl.
>
> http://www.caliban.org/ruby/rubyguide.shtml

Thanks for the link. That really is a fine document. The only thing I
spotted that I'd change is the recommendation to put all require and
include statements at the top of a file, before class definitions.
With requires it usually (though not invariably) makes sense, but with
includes it doesn't, since they pertain to specific classes.


David

--
Q. What is THE Ruby book for Rails developers?
A. RUBY FOR RAILS by David A. Black (http://www.manning.com/black)
(See what readers are saying! http://www.rubypal.com/r4rrevs.pdf)
Q. Where can I get Ruby/Rails on-site training, consulting, coaching?
A. Ruby Power and Light, LLC (http://www.rubypal.com)

Thank you everyone! There were several posts that worked and did what
I needed to do.

Thanks again.

Brad