Don't know how much gulf this can take, but I sure would like as small
a version as I can get.
paths = File.expand_path(File.dirname(__FILE__)).split('/')
paths.size.downto(1) do |i|
f = (paths.slice(0..i)+['test']).join('/')
$TESTDIR = File.join(f,'FIXTURE') if File.directory?(f)
end
raise unless $TESTDIR
T.
I believe you mean golf? 
Here are a couple options (probably not the shortest though):
paths = File.expand_path(File.dirname(__FILE__)).split('/')
while paths.length > 1
f = (paths+['test/FIXTURE']).join('/')
$TESTDIR = f if File.directory?(f)
paths.pop
end
raise unless $TESTDIR
path = File.expand_path(File.dirname(__FILE__))
until path == '/'
f = File.join(path,'test/FIXTURE')
$TESTDIR = f if File.directory?(f)
path = File.dirname(path)
end
raise unless $TESTDIR
Ryan
On 10/22/05, Trans <transfire@gmail.com> wrote:
Don't know how much gulf this can take, but I sure would like as small
a version as I can get.paths = File.expand_path(File.dirname(__FILE__)).split('/')
paths.size.downto(1) do |i|
f = (paths.slice(0..i)+['test']).join('/')
$TESTDIR = File.join(f,'FIXTURE') if File.directory?(f)
end
raise unless $TESTDIR
Hi --
On Sun, 23 Oct 2005, Trans wrote:
Don't know how much gulf this can take, but I sure would like as small
a version as I can get.paths = File.expand_path(File.dirname(__FILE__)).split('/')
paths.size.downto(1) do |i|
f = (paths.slice(0..i)+['test']).join('/')
$TESTDIR = File.join(f,'FIXTURE') if File.directory?(f)
end
raise unless $TESTDIR
require'file/find'
raise unless$TESTDIR=Find.find(File.dirname(__FILE__)){|f|
break File.join(f,"FIXTURE")if test(?d,f)&&/\btest$/.match(f)}
David
--
David A. Black
dblack@wobblini.net
Trans wrote:
Don't know how much gulf this can take, but I sure would like as small
a version as I can get.paths = File.expand_path(File.dirname(__FILE__)).split('/')
paths.size.downto(1) do |i|
f = (paths.slice(0..i)+['test']).join('/')
$TESTDIR = File.join(f,'FIXTURE') if File.directory?(f)
end
raise unless $TESTDIR
The best balance between size and readability is probably the
following:
find_test_dir
Of course, the method "find_test_dir" must be implemented somewhere,
but you won't have to read it very often, and you can just use the six
lines you've written above.
However, I would use the 'pathname' library to clarify some of the
code.
Cheers,
Gavin
Oops I mean "GOLF" with an O not a U. GOLF. G...O...L...F.
(smtms i wsh nglsh ws mr lk hbrw)
T.
Gavin Sinclair wrote:
> paths = File.expand_path(File.dirname(__FILE__)).split('/')
> paths.size.downto(1) do |i|
> f = (paths.slice(0..i)+['test']).join('/')
> $TESTDIR = File.join(f,'FIXTURE') if File.directory?(f)
> end
> raise unless $TESTDIR[...]
However, I would use the 'pathname' library to clarify some of the
code.
To wit (untested):
require 'pathname'
def find_test_dir
dir = Pathname.new(__FILE__)
loop do
dir = dir.dirname
test_dir = dir + "test"
return test_dir if test_dir.directory?
raise "Test directory not found" if dir.mountpoint?
end
end
# ...
fixture = find_test_dir() + "FIXTURE"
Cheers,
Gavin
David A. Black wrote:
require'file/find'
raise unless$TESTDIR=Find.find(File.dirname(__FILE__)){|f|
break File.join(f,"FIXTURE")if test(?d,f)&&/\btest$/.match(f)}
Shorter than I thought possible! And I did not know about file/find.
Thanks,
T.
Gavin Sinclair wrote:
The best balance between size and readability is probably the
following:find_test_dir
Of course, the method "find_test_dir" must be implemented somewhere,
but you won't have to read it very often, and you can just use the six
lines you've written above.
That' a good point. But my trouble here is that the tests themsselves
can get run out of either of two locations: as attached to the bottom
of the file they test, which is in lib/ or standalone in test/ dir
(same code I just extract it). So there's no convenient place to store
this and still be able to require it without the same trouble, AFAIK.
However, I would use the 'pathname' library to clarify some of the
code.
Yes, in this case it would be good, but I recently discovered Pathname
is terribly slow, so I stopped using it.
Thanks,
T.
?nm y t'nD
?nm y t'nd)wrbh kl rm sw hslgn hsw i smtms(
On 10/22/05, Trans <transfire@gmail.com> wrote:
Oops I mean "GOLF" with an O not a U. GOLF. G...O...L...F.
(smtms i wsh nglsh ws mr lk hbrw)
T.
--
thanks,
-pate
-------------------------
pat eyler wrote:
?nm y t'nD
?nm y t'nd)wrbh kl rm sw hslgn hsw i smtms(
LOL 
pat eyler <pat.eyler@gmail.com> writes:
?nm y t'nD
?nm y t'nd)wrbh kl rm sw hslgn hsw i smtms(
.hght, drrfn b nc gnnm ht wh gnzm ylpmS
$ echo -e "[...snip...]" | ./2hebrew.rb
lc yw s rtlsnrt wrbhgp t hslgn n
?s knht y t'nD
$ cat 2hebrew.rb
#!/usr/bin/env ruby
ARGF.each do | line | puts line.chomp.gsub(/[aeiou]/i, '').reverse end
cheers,
Brian
On 23/10/05, Christian Neukirchen <chneukirchen@gmail.com> wrote:
pat eyler <pat.eyler@gmail.com> writes:
> ?nm y t'nD
>
> ?nm y t'nd)wrbh kl rm sw hslgn hsw i smtms(.hght, drrfn b nc gnnm ht wh gnzm ylpmS
> -pate
--
Christian Neukirchen <chneukirchen@gmail.com> http://chneukirchen.org
--
http://ruby.brian-schroeder.de/
Stringed instrument chords: http://chordlist.brian-schroeder.de/