Delete file without the directory object

[ha]

Hi guys,

I've been going through the docs and the documentation I find always gets the directory first and then get the file to be destroyed.

I got a simple script running and it works like a champ!

The script is like this:

require 'fog'

conn = Fog::Storage.new(

  provider: 'Rackspace',

  rackspace_username: 'username',

  rackspace_api_key:  'key'

)

directory = conn.directories.get('mycontainer')

f = directory.files.head("path/file1.jpg")

puts "Should return file: #{f.inspect}"

f.destroy if f

My big problem is that I know the path of the file "path/file1.jpg" (which is the key attribute to the file) but not the container name, and I have about 20 different containers.

Is there a method that works like a search file that receives the file key?!

I don't know why the container is not saved along with 'path/file1.jpg' but I'm already thinking on adding it, I just want to know if I will have to do that before or after cleaning my containers.

Anyway, if there is no search I'll have to brute force and iterate through all containers all files to delete the old files...

Thanks in advance,

./Helio

[geemus (Wesley Beary)]

You should be able to get a performance boost by avoiding the get (if you can assume the directory already exists) and instead using new.

So instead of:

directory = conn.directories.get('mycontainer')

You would have:

directory = conn.directories.new(:key => 'mycontainer')

Beyond that, however, I don't think there is a simple facility for searching across buckets. So I fear falling back to iterating over them and using head may be your best bet.

wes

--
You received this message because you are subscribed to the Google Groups "ruby-fog" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ruby-fog+u...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

[ha]

I was seeing that coming...

Do you know if rackspace has some kind of interface that I can search across buckets?!

I can spend sometime working on a search, if Rackspace has an way of doing that...

Thanks,

./Helio

[ha]

Wes,

I do have the container CDN URL.

Do you know if it is possible something like this?

directory = conn.directories.new(:key => 'CDN_URL')
directory = conn.directories.new(:cdn_name => 'CDN_URL')
directory = conn.directories.new(:public_url => 'CDN_URL')

Thanks,

./Helio

[geemus (Wesley Beary)]

Not that I know of, hopefully some of our Rackspace experts can chime in. It might be easier to engage with them if you open an issue actually.

Thanks!

wes

--

[Rupak Ganguly]

Helio,

  I am not a Racker but I work for HP Cloud, and we share the Openstack platform. From what I know, you cannot pass in a CDN url to get to a directory object. You need a directory 'name' to 'get'/'head' a directory object.

As far as I know, none of these are possible.

directory = conn.directories.new(:key => 'CDN_URL')

directory = conn.directories.new(:cdn_name => 'CDN_URL')
directory = conn.directories.new(:public_url => 'CDN_URL')

But, still as Wes mentioned, I would open an issue and one of the Rackers to comment.

Thanks,
Rupak Ganguly
Ph: 678-648-7434

[Kyle Rames]

Hi Helio,

Sorry for the late response! I have been out of the office for the birth of my daughter.

Unfortunately, you cannot use the CDN URL to get a handle to the directory. The only solution is to would be to get a list of all the CDN enabled containers and then head each container until you find the desired URL which is sub-optimal at best.

I am also unaware of a call that will search all of the containers for an object. I have forwarded your question onto our cloud files team. Hopefully they will have a better answer for you.

Kyle