Cannot find Document with Ruby and MongoDB when using ObjectId

Question

I have some code written in Ruby 1.9.2 patch level 136 and I'm have an issue where when I perform a find via the _id in the raw ruby mongo driver I get a nil when trying to use a value from a csv file. Here's the code:

require 'mongo'
require 'csv'
require 'bson'

# Games database
gamedb = Mongo::Connection.new("localhost", 27017).db("gamedb")
@games = gamedb.collection("games")

# Loop over CSV data.
CSV.foreach("/tmp/somedata.csv") do |row|

  puts row[0] # Puts the ObjectId

  @game = @games.find( { "_id" => row[0] } ).first  
  puts @game.inspect

end

The CSV file looks like this:

_id,game_title,platform,upc_db_match,upc
4ecdacc339c7d7a2a6000002,TMNT,PSP,TMNT,085391157663
4ecdacc339c7d7a2a6000004,Super Mario Galaxy,Wii,Super Mario Galaxy,045496900434
4ecdacc339c7d7a2a6000005,Beowulf,PSP,Beowulf,097363473046

The first column is the objectId in Mongo that I already have. If I perform a local find from the mongo command line the values in the first column, I get the data I want. However, the code above returns nil on the @game.inspect call.

I've tried the following variations, which all produce nil:

@game = @games.find( { "_id" => row[0].to_s } ).first
@game = @games.find( { "_id" => row[0].to_s.strip } ).first

I've even tried building the ObjectId with the BSON classes as such:

@game = @games.find( { "_id" => BSON::ObjectId(row[0]) } ).first

or

@game = @games.find( { "_id" => BSON::ObjectId("#{row[0]}") } ).first

Both of which output the following error:

/Users/donnfelker/.rvm/gems/ruby-1.9.2-p136@upc-etl/gems/bson-1.4.0/lib/bson/types/object_id.rb:126:in `from_string': illegal ObjectId format: _id (BSON::InvalidObjectId)
    from /Users/donnfelker/.rvm/gems/ruby-1.9.2-p136@upc-etl/gems/bson-1.4.0/lib/bson/types/object_id.rb:26:in `ObjectId'
    from migrate_upc_from_csv.rb:14:in `block in <main>'
    from /Users/donnfelker/.rvm/rubies/ruby-1.9.2-p136/lib/ruby/1.9.1/csv.rb:1768:in `each'
    from /Users/donnfelker/.rvm/rubies/ruby-1.9.2-p136/lib/ruby/1.9.1/csv.rb:1202:in `block in foreach'
    from /Users/donnfelker/.rvm/rubies/ruby-1.9.2-p136/lib/ruby/1.9.1/csv.rb:1340:in `open'
    from /Users/donnfelker/.rvm/rubies/ruby-1.9.2-p136/lib/ruby/1.9.1/csv.rb:1201:in `foreach'
    from migrate_upc_from_csv.rb:10:in `<main>'

The crazy thing is, if I manually create the BSON ObjectId by hand it works (as shown below):

@game = @games.find( { "_id" => BSON::ObjectId("4ecdacc339c7d7a2a6000004") } ).first

When I run @game.inspect I get my data back, as I would expect. However, If I change this to use row[0], I get nil.

Why? What am I doing wrong?

System Details

$ gem list

*** LOCAL GEMS ***

bson (1.4.0)
bson_ext (1.4.0)
mongo (1.4.0)

RVM Version: rvm 1.6.9

Ruby Version: ruby 1.9.2p136 (2010-12-25 revision 30365) [x86_64-darwin10.6.0]

Mongo Version:

[initandlisten] db version v1.8.2, pdfile version 4.5
[initandlisten] git version: 433bbaa14aaba6860da15bd4de8edf600f56501b

Again, why? What am I doing wrong here? Thanks!

Michael Kohl · Answer 1 · 2011-11-28T23:13:17.263

2

Are you sure your CSV parsing code isn't treating the headers as a first line of data and actually tries to do BSON::ObjectId("_id")? The error message kinda looks like it. Try with FasterCSV.foreach('/tmp/somedata.csv', :headers => true) and using row['_id'] (IIRC you'll still have to use BSON::ObjectID).

edited Nov 28 '11 at 23:13

answered Nov 28 '11 at 23:07

Michael Kohl

66,324
14
138
158

Thanks for the comment. I'm not using CSV, however your answer did help me get to the correct end result with the assistance of the correct syntax below. Thanks Michael! – Donn Felker Nov 29 '11 at 02:11

score 2 · Accepted Answer · answered Nov 28 '11 at 23:50

2

The first row is not being read as a header, to do that pass in :headers => true like this:

require 'csv'

# Loop over CSV data.
CSV.foreach("/tmp/somedata.csv", :headers => true) do |row|

  puts row[0] # Puts the ObjectId

end

If you do not pass the :headers parameter in you can see the first row[0] object is the string "_id":

_id
4ecdacc339c7d7a2a6000002
4ecdacc339c7d7a2a6000004
4ecdacc339c7d7a2a6000005

When you include it, you are golden:

4ecdacc339c7d7a2a6000002
4ecdacc339c7d7a2a6000004
4ecdacc339c7d7a2a6000005

answered Nov 28 '11 at 23:50

Tyler Brock

29,626
15
79
79

Isn't that essentially what I said? – Michael Kohl Nov 28 '11 at 23:56
Yeah sorry about that! I hadn't refreshed the page before I started writing. At least Donn now also has a definitive answer which might be a little more clear. I tried the above on my computer and can confirm our solution fixes the problem he is having and that he can accept the answer. Apologies if I stepped on any toes here. – Tyler Brock Nov 29 '11 at 00:04
1

Accepting this as the answer as the code sample shows what I need. I'm not using FasterCSV (which I know has been replaced in 1.9.2 etc etc etc). I still up voted the original though too. Thanks! – Donn Felker Nov 29 '11 at 02:10
Also, its good to note that the reason I never saw the _id header as the first puts is because my CSV file is > 1000 lines long. The screen output scrolled before I could notice the _id. – Donn Felker Nov 29 '11 at 02:18

Cannot find Document with Ruby and MongoDB when using ObjectId

2 Answers2