Results from #search are not marked as UTF-8 in Ruby 1.9.3

[keltia]

When trying to convert into YAML the values I get from my LDAP server through Treequel, strings are not marked as UTF-8 in Ruby 1.9.3 resulting in everything marked as "BINARY":

522 [16:29] roberto@roberto-aw:gems/ruby-ldap-0.9.12> curl http://localhost:3000/user/roberto.yaml

---
!binary "a3Nu":
- !binary |-
  Uk9CRVJU
!binary "a2dpdmVubmFtZQ==":
- !binary |-
  T2xsaXZpZXI=
!binary "dGVsZXBob25lbnVtYmVy":
- !binary |-
  KzMzICgwKTEgNjk4ODcyOTA=
!binary "ZXVyb2NvbnRyb2xyb29taWQ=":
- !binary |-
  Q0MuMjE=
!binary "bWFpbA==":
- !binary |-
  b2xsaXZpZXIucm9iZXJ0QGV1cm9jb250cm9sLmludA==
!binary "dWlk":
- !binary |-
  cm9iZXJ0bw==
!binary "ZXVyb2NvbnRyb2x1bml0Y29kZQ==":
- !binary |-
  RFIvUE0vRlJN
!binary "ZXVyb2NvbnRyb2xncmFkZQ==":
- !binary |-
  QUQxMA==
!binary "ZXVyb2NvbnRyb2xzdGFydGRhdGU=":
- !binary |-
  MTk5Ni8wNy8wMQ==
!binary "ZW1wbG95ZWVudW1iZXI=":
- !binary |-
  MzAzNQ==
!binary "ZXVyb2NvbnRyb2xzaXRlaWQ=":
- !binary |-
  QnJldGlnbnk=

instead of the UTF8 strings.

JSON output through again Treequel is fine:

524 [16:29] roberto@roberto-aw:gems/ruby-ldap-0.9.12> curl http://localhost:3000/user/roberto.json
{"persons":[{"person":["roberto",{"ksn":["ROBERT"],"kgivenname":["Ollivier"],"telephonenumber":["+33 (0)1 69887290"],"eurocontrolroomid":["CC.21"],"mail":["ollivier.robert@eurocontrol.int"],"uid":["roberto"],"eurocontrolunitcode":["DR/PM/FRM"],"eurocontrolgrade":["AD10"],"eurocontrolstartdate":["1996/07/01"],"employeenumber":["3035"],"eurocontrolsiteid":["Bretigny"]}]}]}%          

NOTE: what is running on port 3000 is my own LDAP proxy, built with Sinatra/Treequel/ruby-ldap: https://bitbucket.org/keltia/ldap-proxy

[ged]

I've confirmed that data comes out of Ruby-LDAP as ASCII-8BIT, which is why Treequel's strings have that encoding:

[1] pry(main)> require 'ldap'
=> true
[2] pry(main)> c = LDAP::SSLConn.new( 'ldap.example.com', 389, true )
=> #<LDAP::SSLConn:0x007fc7c30f8420
 @args=["ldap.example.com", 389, true],
 @sasl_quiet=false>
[3] pry(main)> res = c.search_ext2( 'dc=example,dc=com', LDAP::LDAP_SCOPE_SUBTREE, '(uid=mgranger)' )
=> [{"dn"=>["uid=mgranger,ou=People,dc=example,dc=com"], [...]}]
[4] pry(main)> res.first['cn']
=> ["Michael Granger"]
[5] pry(main)> res.first['cn'].first.encoding
=> #<Encoding:ASCII-8BIT>

[ghost]

hi guys,

I found example of solving encoding problem here and I think that it can be applied to str here

but problem is that I'm not a C programmer and I'm not sure that it is correct way for solving this problem.

what do you think?

[ghost]

looks like it is bad idea... :)

1.9.3p125 :001 > require 'ldap'
 => true 
1.9.3p125 :002 > c = LDAP::Conn.new( 'localhost', 389 )
 => #<LDAP::Conn:0x007fafcb06f810> 
1.9.3p125 :003 > res = c.search_ext2( 'dc=localhost,dc=com', LDAP::LDAP_SCOPE_SUBTREE, '(cn=admin)' )
 => [{"dn"=>["cn=admin,dc=localhost,dc=com"], "objectClass"=>["\u0005\x81\u0004"], "cn"=>["\u0005A\u0001"]}] 
1.9.3p125 :004 > res.first['cn']
 => ["\u0005A\u0001"] 
1.9.3p125 :005 > res.first['cn'].first.encoding
 => #<Encoding:UTF-8>