Example: MongoDB Ruby Driver with GridFS

Load Example Script/Class

1. Load Example Script with GridfsLoader class

   $ irb
   > require './gridfs_loader'
    => true 

2. Get Connection to MongoDB

   > GridfsLoader.mongo_client
   creating connection mongodb://localhost:27017 test
    => #<Mongo::Client:0x16705380 cluster=localhost:27017> 

Import a File (into GridFS)

1. Locate Example File to Import

   $ ls -ltrh
   total 7.5M
   -rw-rw-r--. 1 jim jim 313K Nov  9 04:11 image1.jpg
   -rw-rw-r--. 1 jim jim 7.2M Nov  9 04:11 image2.jpg
   -rw-rw-r--. 1 jim jim 1.3K Nov 11 01:14 gridfs_loader.rb

2. Load the Operating System (OS) File

   > os_file=File.open("./image1.jpg", "rb")
    => #<File:./image1.jpg> 

3. Create a GridFS File from OS File

   Note that we are not passing in any file description properties at this time. All info will be derived form the raw file itself.

   > grid_file = Mongo::Grid::File.new(os_file.read)
     => #<Mongo::Grid::File:0x16281840 filename=> 

4. Inspect the GridFS File

   > grid_file.methods
    => [:chunk_size, :content_type, :filename, :id, :md5, :upload_date, :chunks, :data, :info

   MongoDB assigns an id we can use later to get the data from GridFS.

   > grid_file.id
    => BSON::ObjectId('5642e168e301d09ce9000000') 

   Content type defaults to binary/octet-stream

   > grid_file.content_type
    => "binary/octet-stream" 

   Filename is nil because this is not derived from the os_file or data and we did not pass in file description properties in when we created the file.

   > grid_file.filename
    => nil 

   MongoDB calculates an md5 hash from the data.

   > grid_file.md5
    => #<Digest::MD5: 3468ca1c23cc13ac6af493c4642cc72a> 

   MongoDB timestamps the creation of the file at the time the MongoDB File was created. It is not yet technically in MongoDB/GridFS yet.

   > grid_file.upload_date
    => 2015-11-11 06:32:29 UTC 

   Much of what was shown above enacsulated in the file's info object.

   > grid_file.info
    => #<Mongo::Grid::File::Info:0x21616940 
   chunk_size=261120 
   filename= 
   content_type=binary/octet-stream 
   id=5642e168e301d09ce9000000 
   md5=3468ca1c23cc13ac6af493c4642cc72a> 

   Chunks are broken up into ~255KB

   > grid_file.chunk_size
    => 261120 

   Our 313K OS File was broken up into two (2) chunks of up to 255K each.

   > grid_file.chunks.count
    => 2 
   > grid_file.chunks
    => [#<Mongo::Grid::File::Chunk:0x0000000293a6c8 ... "n"=>0}>, 
        #<Mongo::Grid::File::Chunk:0x0000000293a330 ... "n"=>1}>] 

5. Write the GridFS File into GridFS

   > c=GridfsLoader.mongo_client
   creating connection mongodb://localhost:27017 test
    => #<Mongo::Client:0x20670940 cluster=localhost:27017> 
   > r=c.database.fs.insert_one(grid_file)
    => BSON::ObjectId('5642e168e301d09ce9000000') 
   

Export File (from GridFS)

1. Find the GridFS File (by ID)

   > stored_file=c.database.fs.find_one(:_id => BSON::ObjectId('5642e168e301d09ce9000000'))
    => #<Mongo::Grid::File:0x19799600 filename=> 

2. Create an Output File to Write To

   > os_file2=File.open("./exported_copy.jpg","wb")
    => #<File:./exported_copy.jpg>

3. Write data to File

   > stored_file.chunks.size
    => 2 
   > stored_file.chunks.reduce([]) { |x,chunk| os_file2 << chunk.data.data }
    => #<File:./exported_copy.jpg> 

4. Locate New OS File Copy

   -rw-rw-r--. 1 jim jim 313K Nov  9 04:11 image1.jpg
   -rw-rw-r--. 1 jim jim 7.2M Nov  9 04:11 image2.jpg
   -rw-rw-r--. 1 jim jim 1.3K Nov 11 01:14 gridfs_loader.rb
   -rw-rw-r--. 1 jim jim 313K Nov 11 02:17 exported_copy.jpg

Adding File Info/Description

1. Create a description hash to pass into GridFS initialize

   > description={}
    => {} 

2. Assign a filename (this is a standard property)

   > description[:filename]="myfile.jpg"
    => "myfile.jpg" 

3. Assign a content_type (this is a standard property)

   > description[:content_type]="image/jpeg"
    => "image/jpeg" 

4. Assign some custom properties. This must go in metadata (metadata is a standard property but what is within metadata is custom)

   > description[:metadata]={:author=>"kiran", :topic=>"nice spot"}
    => {:author=>"kiran", :topic=>"nice spot"} 

5. Pass the file properties into initialize with OS File with data bytes

   > grid_file = Mongo::Grid::File.new(os_file.read, description)
    => #<Mongo::Grid::File:0x22644620 filename=myfile.jpg> 

6. Notice the standard properties showing up in the info object

   > grid_file.info
    => #<Mongo::Grid::File::Info:0x22644480 
         chunk_size=261120 
         filename=myfile.jpg 
         content_type=image/jpeg 
         id=5642f149e301d09ce9000009 
         md5=3468ca1c23cc13ac6af493c4642cc72a> 

7. Write data and file properties into GridFS

   > r=c.database.fs.insert_one(grid_file)
    => BSON::ObjectId('5642f149e301d09ce9000009') 

Find Files (Grid::File)

• Query standard properties at document root

  > c.database.fs.find_one(:contentType=>'image/jpeg', :filename=>'myfile.jpg')
   => #<Mongo::Grid::File:0x20470500 filename=myfile.jpg> 

• Query custom metadata properties using nested property dot (".") syntax

  > c.database.fs.find_one(:"metadata.author"=>"kiran", :"metadata.topic"=>{:$regex=>"spot"})
   => #<Mongo::Grid::File:0x18411820 filename=myfile.jpg> 

Find File Properties (Ruby Hash)

• Query standard properties at document root

  > require 'pp'
  > pp c.database.fs.find(:contentType=>'image/jpeg', :filename => "myfile.jpg").first
  {"_id"=>BSON::ObjectId('5642f149e301d09ce9000009'),
   "chunkSize"=>261120,
   "uploadDate"=>2015-11-11 07:41:50 UTC,
   "contentType"=>"image/jpeg",
   "filename"=>"myfile.jpg",
   "metadata"=>{"author"=>"kiran", "topic"=>"nice spot"},
   "length"=>307797,
   "md5"=>"3468ca1c23cc13ac6af493c4642cc72a"}

• Query custom metadata properties using nested property dot (".") syntax

  > pp c.database.fs.find(:"metadata.author"=>"kiran", :"metadata.topic"=>{:$regex=>"spot"}).first
  {"_id"=>BSON::ObjectId('5642f149e301d09ce9000009'),
   "chunkSize"=>261120,
   "uploadDate"=>2015-11-11 07:41:50 UTC,
   "contentType"=>"image/jpeg",
   "filename"=>"myfile.jpg",
   "metadata"=>{"author"=>"kiran", "topic"=>"nice spot"},
   "length"=>307797,
   "md5"=>"3468ca1c23cc13ac6af493c4642cc72a"}

Deleting Files from GridFS

• Delete a specific File

  > id=c.database.fs.find(:"metadata.author"=>"kiran").first[:_id]
   => BSON::ObjectId('5642f149e301d09ce9000009') 
  > r=c.database.fs.find(:_id=>id).delete_one
   => #<Mongo::Operation::Result:22441840 documents=[{"ok"=>1, "n"=>1}]>
  > r.deleted_count
   => 1 
  > r.n
   => 1 

• Delete all Files

  > r=c.database.fs.find.delete_many
   => #<Mongo::Operation::Result:18398180 documents=[{"ok"=>1, "n"=>6}]> 

Last Updated: 2016-02-21