jonbell-mongo 1.3.1.2
Sign up to get free protection for your applications and to get access to all the features.
- data/LICENSE.txt +190 -0
- data/README.md +333 -0
- data/Rakefile +215 -0
- data/bin/mongo_console +21 -0
- data/docs/CREDITS.md +123 -0
- data/docs/FAQ.md +116 -0
- data/docs/GridFS.md +158 -0
- data/docs/HISTORY.md +263 -0
- data/docs/RELEASES.md +33 -0
- data/docs/REPLICA_SETS.md +72 -0
- data/docs/TUTORIAL.md +247 -0
- data/docs/WRITE_CONCERN.md +28 -0
- data/lib/mongo.rb +97 -0
- data/lib/mongo/collection.rb +895 -0
- data/lib/mongo/connection.rb +926 -0
- data/lib/mongo/cursor.rb +474 -0
- data/lib/mongo/db.rb +617 -0
- data/lib/mongo/exceptions.rb +71 -0
- data/lib/mongo/gridfs/grid.rb +107 -0
- data/lib/mongo/gridfs/grid_ext.rb +57 -0
- data/lib/mongo/gridfs/grid_file_system.rb +146 -0
- data/lib/mongo/gridfs/grid_io.rb +485 -0
- data/lib/mongo/gridfs/grid_io_fix.rb +38 -0
- data/lib/mongo/repl_set_connection.rb +356 -0
- data/lib/mongo/util/conversions.rb +89 -0
- data/lib/mongo/util/core_ext.rb +60 -0
- data/lib/mongo/util/pool.rb +177 -0
- data/lib/mongo/util/server_version.rb +71 -0
- data/lib/mongo/util/support.rb +82 -0
- data/lib/mongo/util/uri_parser.rb +185 -0
- data/mongo.gemspec +34 -0
- data/test/auxillary/1.4_features.rb +166 -0
- data/test/auxillary/authentication_test.rb +68 -0
- data/test/auxillary/autoreconnect_test.rb +41 -0
- data/test/auxillary/fork_test.rb +30 -0
- data/test/auxillary/repl_set_auth_test.rb +58 -0
- data/test/auxillary/slave_connection_test.rb +36 -0
- data/test/auxillary/threaded_authentication_test.rb +101 -0
- data/test/bson/binary_test.rb +15 -0
- data/test/bson/bson_test.rb +654 -0
- data/test/bson/byte_buffer_test.rb +208 -0
- data/test/bson/hash_with_indifferent_access_test.rb +38 -0
- data/test/bson/json_test.rb +17 -0
- data/test/bson/object_id_test.rb +154 -0
- data/test/bson/ordered_hash_test.rb +210 -0
- data/test/bson/timestamp_test.rb +24 -0
- data/test/collection_test.rb +910 -0
- data/test/connection_test.rb +324 -0
- data/test/conversions_test.rb +119 -0
- data/test/cursor_fail_test.rb +75 -0
- data/test/cursor_message_test.rb +43 -0
- data/test/cursor_test.rb +483 -0
- data/test/db_api_test.rb +738 -0
- data/test/db_connection_test.rb +15 -0
- data/test/db_test.rb +315 -0
- data/test/grid_file_system_test.rb +259 -0
- data/test/grid_io_test.rb +209 -0
- data/test/grid_test.rb +258 -0
- data/test/load/thin/load.rb +24 -0
- data/test/load/unicorn/load.rb +23 -0
- data/test/replica_sets/connect_test.rb +112 -0
- data/test/replica_sets/connection_string_test.rb +32 -0
- data/test/replica_sets/count_test.rb +35 -0
- data/test/replica_sets/insert_test.rb +53 -0
- data/test/replica_sets/pooled_insert_test.rb +55 -0
- data/test/replica_sets/query_secondaries.rb +108 -0
- data/test/replica_sets/query_test.rb +51 -0
- data/test/replica_sets/replication_ack_test.rb +66 -0
- data/test/replica_sets/rs_test_helper.rb +27 -0
- data/test/safe_test.rb +68 -0
- data/test/support/hash_with_indifferent_access.rb +186 -0
- data/test/support/keys.rb +45 -0
- data/test/support_test.rb +18 -0
- data/test/test_helper.rb +102 -0
- data/test/threading/threading_with_large_pool_test.rb +90 -0
- data/test/threading_test.rb +87 -0
- data/test/tools/auth_repl_set_manager.rb +14 -0
- data/test/tools/repl_set_manager.rb +266 -0
- data/test/unit/collection_test.rb +130 -0
- data/test/unit/connection_test.rb +85 -0
- data/test/unit/cursor_test.rb +109 -0
- data/test/unit/db_test.rb +94 -0
- data/test/unit/grid_test.rb +49 -0
- data/test/unit/pool_test.rb +9 -0
- data/test/unit/repl_set_connection_test.rb +59 -0
- data/test/unit/safe_test.rb +125 -0
- data/test/uri_test.rb +91 -0
- metadata +224 -0
@@ -0,0 +1,71 @@
|
|
1
|
+
# encoding: UTF-8
|
2
|
+
|
3
|
+
#
|
4
|
+
# --
|
5
|
+
# Copyright (C) 2008-2011 10gen Inc.
|
6
|
+
#
|
7
|
+
# Licensed under the Apache License, Version 2.0 (the "License");
|
8
|
+
# you may not use this file except in compliance with the License.
|
9
|
+
# You may obtain a copy of the License at
|
10
|
+
#
|
11
|
+
# http://www.apache.org/licenses/LICENSE-2.0
|
12
|
+
#
|
13
|
+
# Unless required by applicable law or agreed to in writing, software
|
14
|
+
# distributed under the License is distributed on an "AS IS" BASIS,
|
15
|
+
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
16
|
+
# See the License for the specific language governing permissions and
|
17
|
+
# limitations under the License.
|
18
|
+
# ++
|
19
|
+
|
20
|
+
module Mongo
|
21
|
+
# Generic Mongo Ruby Driver exception class.
|
22
|
+
class MongoRubyError < StandardError; end
|
23
|
+
|
24
|
+
# Raised when MongoDB itself has returned an error.
|
25
|
+
class MongoDBError < RuntimeError; end
|
26
|
+
|
27
|
+
# Raised when configuration options cause connections, queries, etc., to fail.
|
28
|
+
class ConfigurationError < MongoRubyError; end
|
29
|
+
|
30
|
+
# Raised on fatal errors to GridFS.
|
31
|
+
class GridError < MongoRubyError; end
|
32
|
+
|
33
|
+
# Raised on fatal errors to GridFS.
|
34
|
+
class GridFileNotFound < GridError; end
|
35
|
+
|
36
|
+
# Raised on fatal errors to GridFS.
|
37
|
+
class GridMD5Failure < GridError; end
|
38
|
+
|
39
|
+
# Raised when invalid arguments are sent to Mongo Ruby methods.
|
40
|
+
class MongoArgumentError < MongoRubyError; end
|
41
|
+
|
42
|
+
# Raised on failures in connection to the database server.
|
43
|
+
class ConnectionError < MongoRubyError; end
|
44
|
+
|
45
|
+
# Raised on failures in connection to the database server.
|
46
|
+
class ReplicaSetConnectionError < ConnectionError; end
|
47
|
+
|
48
|
+
# Raised on failures in connection to the database server.
|
49
|
+
class ConnectionTimeoutError < MongoRubyError; end
|
50
|
+
|
51
|
+
# Raised when a connection operation fails.
|
52
|
+
class ConnectionFailure < MongoDBError; end
|
53
|
+
|
54
|
+
# Raised when authentication fails.
|
55
|
+
class AuthenticationError < MongoDBError; end
|
56
|
+
|
57
|
+
# Raised when a database operation fails.
|
58
|
+
class OperationFailure < MongoDBError; end
|
59
|
+
|
60
|
+
# Raised when a socket read operation times out.
|
61
|
+
class OperationTimeout < ::Timeout::Error; end
|
62
|
+
|
63
|
+
# Raised when a client attempts to perform an invalid operation.
|
64
|
+
class InvalidOperation < MongoDBError; end
|
65
|
+
|
66
|
+
# Raised when an invalid collection or database name is used (invalid namespace name).
|
67
|
+
class InvalidNSName < RuntimeError; end
|
68
|
+
|
69
|
+
# Raised when the client supplies an invalid value to sort by.
|
70
|
+
class InvalidSortValueError < MongoRubyError; end
|
71
|
+
end
|
@@ -0,0 +1,107 @@
|
|
1
|
+
# encoding: UTF-8
|
2
|
+
|
3
|
+
# --
|
4
|
+
# Copyright (C) 2008-2011 10gen Inc.
|
5
|
+
#
|
6
|
+
# Licensed under the Apache License, Version 2.0 (the "License");
|
7
|
+
# you may not use this file except in compliance with the License.
|
8
|
+
# You may obtain a copy of the License at
|
9
|
+
#
|
10
|
+
# http://www.apache.org/licenses/LICENSE-2.0
|
11
|
+
#
|
12
|
+
# Unless required by applicable law or agreed to in writing, software
|
13
|
+
# distributed under the License is distributed on an "AS IS" BASIS,
|
14
|
+
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
15
|
+
# See the License for the specific language governing permissions and
|
16
|
+
# limitations under the License.
|
17
|
+
# ++
|
18
|
+
|
19
|
+
module Mongo
|
20
|
+
|
21
|
+
# Implementation of the MongoDB GridFS specification. A file store.
|
22
|
+
class Grid
|
23
|
+
include GridExt::InstanceMethods
|
24
|
+
|
25
|
+
DEFAULT_FS_NAME = 'fs'
|
26
|
+
|
27
|
+
# Initialize a new Grid instance, consisting of a MongoDB database
|
28
|
+
# and a filesystem prefix if not using the default.
|
29
|
+
#
|
30
|
+
# @core gridfs
|
31
|
+
#
|
32
|
+
# @see GridFileSystem
|
33
|
+
def initialize(db, fs_name=DEFAULT_FS_NAME)
|
34
|
+
raise MongoArgumentError, "db must be a Mongo::DB." unless db.is_a?(Mongo::DB)
|
35
|
+
|
36
|
+
@db = db
|
37
|
+
@files = @db["#{fs_name}.files"]
|
38
|
+
@chunks = @db["#{fs_name}.chunks"]
|
39
|
+
@fs_name = fs_name
|
40
|
+
|
41
|
+
# Ensure indexes only if not connected to slave.
|
42
|
+
unless db.connection.slave_ok?
|
43
|
+
@chunks.create_index([['files_id', Mongo::ASCENDING], ['n', Mongo::ASCENDING]], :unique => true)
|
44
|
+
end
|
45
|
+
end
|
46
|
+
|
47
|
+
# Store a file in the file store. This method is designed only for writing new files;
|
48
|
+
# if you need to update a given file, first delete it using Grid#delete.
|
49
|
+
#
|
50
|
+
# Note that arbitary metadata attributes can be saved to the file by passing
|
51
|
+
# them in as options.
|
52
|
+
#
|
53
|
+
# @param [String, #read] data a string or io-like object to store.
|
54
|
+
#
|
55
|
+
# @option opts [String] :filename (nil) a name for the file.
|
56
|
+
# @option opts [Hash] :metadata ({}) any additional data to store with the file.
|
57
|
+
# @option opts [ObjectId] :_id (ObjectId) a unique id for
|
58
|
+
# the file to be use in lieu of an automatically generated one.
|
59
|
+
# @option opts [String] :content_type ('binary/octet-stream') If no content type is specified,
|
60
|
+
# the content type will may be inferred from the filename extension if the mime-types gem can be
|
61
|
+
# loaded. Otherwise, the content type 'binary/octet-stream' will be used.
|
62
|
+
# @option opts [Integer] (262144) :chunk_size size of file chunks in bytes.
|
63
|
+
# @option opts [Boolean] :safe (false) When safe mode is enabled, the chunks sent to the server
|
64
|
+
# will be validated using an md5 hash. If validation fails, an exception will be raised.
|
65
|
+
#
|
66
|
+
# @return [Mongo::ObjectId] the file's id.
|
67
|
+
def put(data, opts={})
|
68
|
+
opts = opts.dup
|
69
|
+
filename = opts[:filename]
|
70
|
+
opts.merge!(default_grid_io_opts)
|
71
|
+
file = GridIO.new(@files, @chunks, filename, 'w', opts=opts)
|
72
|
+
file.write(data)
|
73
|
+
file.close
|
74
|
+
file.files_id
|
75
|
+
end
|
76
|
+
|
77
|
+
# Read a file from the file store.
|
78
|
+
#
|
79
|
+
# @param [] id the file's unique id.
|
80
|
+
#
|
81
|
+
# @return [Mongo::GridIO]
|
82
|
+
def get(id)
|
83
|
+
opts = {:query => {'_id' => id}}.merge!(default_grid_io_opts)
|
84
|
+
GridIO.new(@files, @chunks, nil, 'r', opts)
|
85
|
+
end
|
86
|
+
|
87
|
+
# Delete a file from the store.
|
88
|
+
#
|
89
|
+
# Note that deleting a GridFS file can result in read errors if another process
|
90
|
+
# is attempting to read a file while it's being deleted. While the odds for this
|
91
|
+
# kind of race condition are small, it's important to be aware of.
|
92
|
+
#
|
93
|
+
# @param [] id
|
94
|
+
#
|
95
|
+
# @return [Boolean]
|
96
|
+
def delete(id)
|
97
|
+
@files.remove({"_id" => id})
|
98
|
+
@chunks.remove({"files_id" => id})
|
99
|
+
end
|
100
|
+
|
101
|
+
private
|
102
|
+
|
103
|
+
def default_grid_io_opts
|
104
|
+
{:fs_name => @fs_name}
|
105
|
+
end
|
106
|
+
end
|
107
|
+
end
|
@@ -0,0 +1,57 @@
|
|
1
|
+
# encoding: UTF-8
|
2
|
+
|
3
|
+
# --
|
4
|
+
# Copyright (C) 2008-2011 10gen Inc.
|
5
|
+
#
|
6
|
+
# Licensed under the Apache License, Version 2.0 (the "License");
|
7
|
+
# you may not use this file except in compliance with the License.
|
8
|
+
# You may obtain a copy of the License at
|
9
|
+
#
|
10
|
+
# http://www.apache.org/licenses/LICENSE-2.0
|
11
|
+
#
|
12
|
+
# Unless required by applicable law or agreed to in writing, software
|
13
|
+
# distributed under the License is distributed on an "AS IS" BASIS,
|
14
|
+
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
15
|
+
# See the License for the specific language governing permissions and
|
16
|
+
# limitations under the License.
|
17
|
+
# ++
|
18
|
+
|
19
|
+
module Mongo
|
20
|
+
module GridExt
|
21
|
+
module InstanceMethods
|
22
|
+
|
23
|
+
# Check the existence of a file matching the given query selector.
|
24
|
+
#
|
25
|
+
# Note that this method can be used with both the Grid and GridFileSystem classes. Also
|
26
|
+
# keep in mind that if you're going to be performing lots of existence checks, you should
|
27
|
+
# keep an instance of Grid or GridFileSystem handy rather than instantiating for each existence
|
28
|
+
# check. Alternatively, simply keep a reference to the proper files collection and query that
|
29
|
+
# as needed. That's exactly how this methods works.
|
30
|
+
#
|
31
|
+
# @param [Hash] selector a query selector.
|
32
|
+
#
|
33
|
+
# @example
|
34
|
+
#
|
35
|
+
# # Check for the existence of a given filename
|
36
|
+
# @grid = GridFileSystem.new(@db)
|
37
|
+
# @grid.exist?(:filename => 'foo.txt')
|
38
|
+
#
|
39
|
+
# # Check for existence filename and content type
|
40
|
+
# @grid = GridFileSystem.new(@db)
|
41
|
+
# @grid.exist?(:filename => 'foo.txt', :content_type => 'image/jpg')
|
42
|
+
#
|
43
|
+
# # Check for existence by _id
|
44
|
+
# @grid = Grid.new(@db)
|
45
|
+
# @grid.exist?(:_id => BSON::ObjectId.from_string('4bddcd24beffd95a7db9b8c8'))
|
46
|
+
#
|
47
|
+
# # Check for existence by an arbitrary attribute.
|
48
|
+
# @grid = Grid.new(@db)
|
49
|
+
# @grid.exist?(:tags => {'$in' => ['nature', 'zen', 'photography']})
|
50
|
+
#
|
51
|
+
# @return [nil, Hash] either nil for the file's metadata as a hash.
|
52
|
+
def exist?(selector)
|
53
|
+
@files.find_one(selector)
|
54
|
+
end
|
55
|
+
end
|
56
|
+
end
|
57
|
+
end
|
@@ -0,0 +1,146 @@
|
|
1
|
+
# encoding: UTF-8
|
2
|
+
|
3
|
+
# --
|
4
|
+
# Copyright (C) 2008-2011 10gen Inc.
|
5
|
+
#
|
6
|
+
# Licensed under the Apache License, Version 2.0 (the "License");
|
7
|
+
# you may not use this file except in compliance with the License.
|
8
|
+
# You may obtain a copy of the License at
|
9
|
+
#
|
10
|
+
# http://www.apache.org/licenses/LICENSE-2.0
|
11
|
+
#
|
12
|
+
# Unless required by applicable law or agreed to in writing, software
|
13
|
+
# distributed under the License is distributed on an "AS IS" BASIS,
|
14
|
+
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
15
|
+
# See the License for the specific language governing permissions and
|
16
|
+
# limitations under the License.
|
17
|
+
# ++
|
18
|
+
|
19
|
+
module Mongo
|
20
|
+
|
21
|
+
# A file store built on the GridFS specification featuring
|
22
|
+
# an API and behavior similar to that of a traditional file system.
|
23
|
+
class GridFileSystem
|
24
|
+
include GridExt::InstanceMethods
|
25
|
+
|
26
|
+
# Initialize a new GridFileSystem instance, consisting of a MongoDB database
|
27
|
+
# and a filesystem prefix if not using the default.
|
28
|
+
#
|
29
|
+
# @param [Mongo::DB] db a MongoDB database.
|
30
|
+
# @param [String] fs_name A name for the file system. The default name, based on
|
31
|
+
# the specification, is 'fs'.
|
32
|
+
def initialize(db, fs_name=Grid::DEFAULT_FS_NAME)
|
33
|
+
raise MongoArgumentError, "db must be a Mongo::DB." unless db.is_a?(Mongo::DB)
|
34
|
+
|
35
|
+
@db = db
|
36
|
+
@files = @db["#{fs_name}.files"]
|
37
|
+
@chunks = @db["#{fs_name}.chunks"]
|
38
|
+
@fs_name = fs_name
|
39
|
+
|
40
|
+
@default_query_opts = {:sort => [['filename', 1], ['uploadDate', -1]], :limit => 1}
|
41
|
+
|
42
|
+
# Ensure indexes only if not connected to slave.
|
43
|
+
unless db.connection.slave_ok?
|
44
|
+
@files.create_index([['filename', 1], ['uploadDate', -1]])
|
45
|
+
@chunks.create_index([['files_id', Mongo::ASCENDING], ['n', Mongo::ASCENDING]], :unique => true)
|
46
|
+
end
|
47
|
+
end
|
48
|
+
|
49
|
+
# Open a file for reading or writing. Note that the options for this method only apply
|
50
|
+
# when opening in 'w' mode.
|
51
|
+
#
|
52
|
+
# Note that arbitary metadata attributes can be saved to the file by passing
|
53
|
+
# them is as options.
|
54
|
+
#
|
55
|
+
# @param [String] filename the name of the file.
|
56
|
+
# @param [String] mode either 'r' or 'w' for reading from
|
57
|
+
# or writing to the file.
|
58
|
+
# @param [Hash] opts see GridIO#new
|
59
|
+
#
|
60
|
+
# @option opts [Hash] :metadata ({}) any additional data to store with the file.
|
61
|
+
# @option opts [ObjectId] :_id (ObjectId) a unique id for
|
62
|
+
# the file to be use in lieu of an automatically generated one.
|
63
|
+
# @option opts [String] :content_type ('binary/octet-stream') If no content type is specified,
|
64
|
+
# the content type will may be inferred from the filename extension if the mime-types gem can be
|
65
|
+
# loaded. Otherwise, the content type 'binary/octet-stream' will be used.
|
66
|
+
# @option opts [Integer] (262144) :chunk_size size of file chunks in bytes.
|
67
|
+
# @option opts [Boolean] :delete_old (false) ensure that old versions of the file are deleted. This option
|
68
|
+
# only work in 'w' mode. Certain precautions must be taken when deleting GridFS files. See the notes under
|
69
|
+
# GridFileSystem#delete.
|
70
|
+
# @option opts [Boolean] :safe (false) When safe mode is enabled, the chunks sent to the server
|
71
|
+
# will be validated using an md5 hash. If validation fails, an exception will be raised.
|
72
|
+
#
|
73
|
+
# @example
|
74
|
+
#
|
75
|
+
# # Store the text "Hello, world!" in the grid file system.
|
76
|
+
# @grid = GridFileSystem.new(@db)
|
77
|
+
# @grid.open('filename', 'w') do |f|
|
78
|
+
# f.write "Hello, world!"
|
79
|
+
# end
|
80
|
+
#
|
81
|
+
# # Output "Hello, world!"
|
82
|
+
# @grid = GridFileSystem.new(@db)
|
83
|
+
# @grid.open('filename', 'r') do |f|
|
84
|
+
# puts f.read
|
85
|
+
# end
|
86
|
+
#
|
87
|
+
# # Write a file on disk to the GridFileSystem
|
88
|
+
# @file = File.open('image.jpg')
|
89
|
+
# @grid = GridFileSystem.new(@db)
|
90
|
+
# @grid.open('image.jpg, 'w') do |f|
|
91
|
+
# f.write @file
|
92
|
+
# end
|
93
|
+
#
|
94
|
+
# @return [Mongo::GridIO]
|
95
|
+
def open(filename, mode, opts={})
|
96
|
+
opts = opts.dup
|
97
|
+
opts.merge!(default_grid_io_opts(filename))
|
98
|
+
del = opts.delete(:delete_old) && mode == 'w'
|
99
|
+
file = GridIO.new(@files, @chunks, filename, mode, opts)
|
100
|
+
return file unless block_given?
|
101
|
+
result = nil
|
102
|
+
begin
|
103
|
+
result = yield file
|
104
|
+
ensure
|
105
|
+
id = file.close
|
106
|
+
if del
|
107
|
+
self.delete do
|
108
|
+
@files.find({'filename' => filename, '_id' => {'$ne' => id}}, :fields => ['_id'])
|
109
|
+
end
|
110
|
+
end
|
111
|
+
end
|
112
|
+
result
|
113
|
+
end
|
114
|
+
|
115
|
+
# Delete the file with the given filename. Note that this will delete
|
116
|
+
# all versions of the file.
|
117
|
+
#
|
118
|
+
# Be careful with this. Deleting a GridFS file can result in read errors if another process
|
119
|
+
# is attempting to read a file while it's being deleted. While the odds for this
|
120
|
+
# kind of race condition are small, it's important to be aware of.
|
121
|
+
#
|
122
|
+
# @param [String] filename
|
123
|
+
#
|
124
|
+
# @yield [] pass a block that returns an array of documents to be deleted.
|
125
|
+
#
|
126
|
+
# @return [Boolean]
|
127
|
+
def delete(filename=nil)
|
128
|
+
if block_given?
|
129
|
+
files = yield
|
130
|
+
else
|
131
|
+
files = @files.find({'filename' => filename}, :fields => ['_id'])
|
132
|
+
end
|
133
|
+
files.each do |file|
|
134
|
+
@files.remove({'_id' => file['_id']})
|
135
|
+
@chunks.remove({'files_id' => file['_id']})
|
136
|
+
end
|
137
|
+
end
|
138
|
+
alias_method :unlink, :delete
|
139
|
+
|
140
|
+
private
|
141
|
+
|
142
|
+
def default_grid_io_opts(filename=nil)
|
143
|
+
{:fs_name => @fs_name, :query => {'filename' => filename}, :query_opts => @default_query_opts}
|
144
|
+
end
|
145
|
+
end
|
146
|
+
end
|
@@ -0,0 +1,485 @@
|
|
1
|
+
# encoding: UTF-8
|
2
|
+
|
3
|
+
# --
|
4
|
+
# Copyright (C) 2008-2011 10gen Inc.
|
5
|
+
#
|
6
|
+
# Licensed under the Apache License, Version 2.0 (the "License");
|
7
|
+
# you may not use this file except in compliance with the License.
|
8
|
+
# You may obtain a copy of the License at
|
9
|
+
#
|
10
|
+
# http://www.apache.org/licenses/LICENSE-2.0
|
11
|
+
#
|
12
|
+
# Unless required by applicable law or agreed to in writing, software
|
13
|
+
# distributed under the License is distributed on an "AS IS" BASIS,
|
14
|
+
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
15
|
+
# See the License for the specific language governing permissions and
|
16
|
+
# limitations under the License.
|
17
|
+
# ++
|
18
|
+
|
19
|
+
require 'digest/md5'
|
20
|
+
|
21
|
+
module Mongo
|
22
|
+
|
23
|
+
# GridIO objects represent files in the GridFS specification. This class
|
24
|
+
# manages the reading and writing of file chunks and metadata.
|
25
|
+
class GridIO
|
26
|
+
DEFAULT_CHUNK_SIZE = 256 * 1024
|
27
|
+
DEFAULT_CONTENT_TYPE = 'binary/octet-stream'
|
28
|
+
PROTECTED_ATTRS = [:files_id, :file_length, :client_md5, :server_md5]
|
29
|
+
|
30
|
+
attr_reader :content_type, :chunk_size, :upload_date, :files_id, :filename,
|
31
|
+
:metadata, :server_md5, :client_md5, :file_length, :file_position
|
32
|
+
|
33
|
+
# Create a new GridIO object. Note that most users will not need to use this class directly;
|
34
|
+
# the Grid and GridFileSystem classes will instantiate this class
|
35
|
+
#
|
36
|
+
# @param [Mongo::Collection] files a collection for storing file metadata.
|
37
|
+
# @param [Mongo::Collection] chunks a collection for storing file chunks.
|
38
|
+
# @param [String] filename the name of the file to open or write.
|
39
|
+
# @param [String] mode 'r' or 'w' or reading or creating a file.
|
40
|
+
#
|
41
|
+
# @option opts [Hash] :query a query selector used when opening the file in 'r' mode.
|
42
|
+
# @option opts [Hash] :query_opts any query options to be used when opening the file in 'r' mode.
|
43
|
+
# @option opts [String] :fs_name the file system prefix.
|
44
|
+
# @option opts [Integer] (262144) :chunk_size size of file chunks in bytes.
|
45
|
+
# @option opts [Hash] :metadata ({}) any additional data to store with the file.
|
46
|
+
# @option opts [ObjectId] :_id (ObjectId) a unique id for
|
47
|
+
# the file to be use in lieu of an automatically generated one.
|
48
|
+
# @option opts [String] :content_type ('binary/octet-stream') If no content type is specified,
|
49
|
+
# the content type will may be inferred from the filename extension if the mime-types gem can be
|
50
|
+
# loaded. Otherwise, the content type 'binary/octet-stream' will be used.
|
51
|
+
# @option opts [Boolean] :safe (false) When safe mode is enabled, the chunks sent to the server
|
52
|
+
# will be validated using an md5 hash. If validation fails, an exception will be raised.
|
53
|
+
def initialize(files, chunks, filename, mode, opts={})
|
54
|
+
@files = files
|
55
|
+
@chunks = chunks
|
56
|
+
@filename = filename
|
57
|
+
@mode = mode
|
58
|
+
opts = opts.dup
|
59
|
+
@query = opts.delete(:query) || {}
|
60
|
+
@query_opts = opts.delete(:query_opts) || {}
|
61
|
+
@fs_name = opts.delete(:fs_name) || Grid::DEFAULT_FS_NAME
|
62
|
+
@safe = opts.delete(:safe) || false
|
63
|
+
@local_md5 = Digest::MD5.new if @safe
|
64
|
+
@custom_attrs = {}
|
65
|
+
|
66
|
+
case @mode
|
67
|
+
when 'r' then init_read
|
68
|
+
when 'w' then init_write(opts)
|
69
|
+
else
|
70
|
+
raise GridError, "Invalid file mode #{@mode}. Mode should be 'r' or 'w'."
|
71
|
+
end
|
72
|
+
end
|
73
|
+
|
74
|
+
def [](key)
|
75
|
+
@custom_attrs[key] || instance_variable_get("@#{key.to_s}")
|
76
|
+
end
|
77
|
+
|
78
|
+
def []=(key, value)
|
79
|
+
if PROTECTED_ATTRS.include?(key.to_sym)
|
80
|
+
warn "Attempting to overwrite protected value."
|
81
|
+
return nil
|
82
|
+
else
|
83
|
+
@custom_attrs[key] = value
|
84
|
+
end
|
85
|
+
end
|
86
|
+
|
87
|
+
# Read the data from the file. If a length if specified, will read from the
|
88
|
+
# current file position.
|
89
|
+
#
|
90
|
+
# @param [Integer] length
|
91
|
+
#
|
92
|
+
# @return [String]
|
93
|
+
# the data in the file
|
94
|
+
def read(length=nil)
|
95
|
+
return '' if @file_length.zero?
|
96
|
+
if length == 0
|
97
|
+
return ''
|
98
|
+
elsif length.nil? && @file_position.zero?
|
99
|
+
read_all
|
100
|
+
else
|
101
|
+
read_length(length)
|
102
|
+
end
|
103
|
+
end
|
104
|
+
alias_method :data, :read
|
105
|
+
|
106
|
+
# Write the given string (binary) data to the file.
|
107
|
+
#
|
108
|
+
# @param [String] string
|
109
|
+
# the data to write
|
110
|
+
#
|
111
|
+
# @return [Integer]
|
112
|
+
# the number of bytes written.
|
113
|
+
def write(io)
|
114
|
+
raise GridError, "file not opened for write" unless @mode[0] == ?w
|
115
|
+
if io.is_a? String
|
116
|
+
if @safe
|
117
|
+
@local_md5.update(io)
|
118
|
+
end
|
119
|
+
write_string(io)
|
120
|
+
else
|
121
|
+
length = 0
|
122
|
+
if @safe
|
123
|
+
while(string = io.read(@chunk_size))
|
124
|
+
@local_md5.update(string)
|
125
|
+
length += write_string(string)
|
126
|
+
end
|
127
|
+
else
|
128
|
+
while(string = io.read(@chunk_size))
|
129
|
+
length += write_string(string)
|
130
|
+
end
|
131
|
+
end
|
132
|
+
length
|
133
|
+
end
|
134
|
+
end
|
135
|
+
|
136
|
+
# Position the file pointer at the provided location.
|
137
|
+
#
|
138
|
+
# @param [Integer] pos
|
139
|
+
# the number of bytes to advance the file pointer. this can be a negative
|
140
|
+
# number.
|
141
|
+
# @param [Integer] whence
|
142
|
+
# one of IO::SEEK_CUR, IO::SEEK_END, or IO::SEEK_SET
|
143
|
+
#
|
144
|
+
# @return [Integer] the new file position
|
145
|
+
def seek(pos, whence=IO::SEEK_SET)
|
146
|
+
raise GridError, "Seek is only allowed in read mode." unless @mode == 'r'
|
147
|
+
target_pos = case whence
|
148
|
+
when IO::SEEK_CUR
|
149
|
+
@file_position + pos
|
150
|
+
when IO::SEEK_END
|
151
|
+
@file_length + pos
|
152
|
+
when IO::SEEK_SET
|
153
|
+
pos
|
154
|
+
end
|
155
|
+
|
156
|
+
new_chunk_number = (target_pos / @chunk_size).to_i
|
157
|
+
if new_chunk_number != @current_chunk['n']
|
158
|
+
save_chunk(@current_chunk) if @mode[0] == ?w
|
159
|
+
@current_chunk = get_chunk(new_chunk_number)
|
160
|
+
end
|
161
|
+
@file_position = target_pos
|
162
|
+
@chunk_position = @file_position % @chunk_size
|
163
|
+
@file_position
|
164
|
+
end
|
165
|
+
|
166
|
+
# The current position of the file.
|
167
|
+
#
|
168
|
+
# @return [Integer]
|
169
|
+
def tell
|
170
|
+
@file_position
|
171
|
+
end
|
172
|
+
alias :pos :tell
|
173
|
+
|
174
|
+
# Rewind the file. This is equivalent to seeking to the zeroth position.
|
175
|
+
#
|
176
|
+
# @return [Integer] the position of the file after rewinding (always zero).
|
177
|
+
def rewind
|
178
|
+
raise GridError, "file not opened for read" unless @mode[0] == ?r
|
179
|
+
seek(0)
|
180
|
+
end
|
181
|
+
|
182
|
+
# Return a boolean indicating whether the position pointer is
|
183
|
+
# at the end of the file.
|
184
|
+
#
|
185
|
+
# @return [Boolean]
|
186
|
+
def eof
|
187
|
+
raise GridError, "file not opened for read #{@mode}" unless @mode[0] == ?r
|
188
|
+
@file_position >= @file_length
|
189
|
+
end
|
190
|
+
alias :eof? :eof
|
191
|
+
|
192
|
+
# Return the next line from a GridFS file. This probably
|
193
|
+
# makes sense only if you're storing plain text. This method
|
194
|
+
# has a somewhat tricky API, which it inherits from Ruby's
|
195
|
+
# StringIO#gets.
|
196
|
+
#
|
197
|
+
# @param [String, Integer] separator or length. If a separator,
|
198
|
+
# read up to the separator. If a length, read the +length+ number
|
199
|
+
# of bytes. If nil, read the entire file.
|
200
|
+
# @param [Integer] length If a separator is provided, then
|
201
|
+
# read until either finding the separator or
|
202
|
+
# passing over the +length+ number of bytes.
|
203
|
+
#
|
204
|
+
# @return [String]
|
205
|
+
def gets(separator="\n", length=nil)
|
206
|
+
if separator.nil?
|
207
|
+
read_all
|
208
|
+
elsif separator.is_a?(Integer)
|
209
|
+
read_length(separator)
|
210
|
+
elsif separator.length > 1
|
211
|
+
read_to_string(separator, length)
|
212
|
+
else
|
213
|
+
read_to_character(separator, length)
|
214
|
+
end
|
215
|
+
end
|
216
|
+
|
217
|
+
# Return the next byte from the GridFS file.
|
218
|
+
#
|
219
|
+
# @return [String]
|
220
|
+
def getc
|
221
|
+
read_length(1)
|
222
|
+
end
|
223
|
+
|
224
|
+
# Creates or updates the document from the files collection that
|
225
|
+
# stores the chunks' metadata. The file becomes available only after
|
226
|
+
# this method has been called.
|
227
|
+
#
|
228
|
+
# This method will be invoked automatically when
|
229
|
+
# on GridIO#open is passed a block. Otherwise, it must be called manually.
|
230
|
+
#
|
231
|
+
# @return [BSON::ObjectId]
|
232
|
+
def close
|
233
|
+
if @mode[0] == ?w
|
234
|
+
if @current_chunk['n'].zero? && @chunk_position.zero?
|
235
|
+
warn "Warning: Storing a file with zero length."
|
236
|
+
end
|
237
|
+
@upload_date = Time.now.utc
|
238
|
+
id = @files.insert(to_mongo_object)
|
239
|
+
end
|
240
|
+
id
|
241
|
+
end
|
242
|
+
|
243
|
+
# Read a chunk of the data from the file and yield it to the given
|
244
|
+
# block.
|
245
|
+
#
|
246
|
+
# Note that this method reads from the current file position.
|
247
|
+
#
|
248
|
+
# @yield Yields on chunk per iteration as defined by this file's
|
249
|
+
# chunk size.
|
250
|
+
#
|
251
|
+
# @return [Mongo::GridIO] self
|
252
|
+
def each
|
253
|
+
return read_all unless block_given?
|
254
|
+
while chunk = read(chunk_size)
|
255
|
+
yield chunk
|
256
|
+
break if chunk.empty?
|
257
|
+
end
|
258
|
+
self
|
259
|
+
end
|
260
|
+
|
261
|
+
def inspect
|
262
|
+
"#<GridIO _id: #{@files_id}>"
|
263
|
+
end
|
264
|
+
|
265
|
+
private
|
266
|
+
|
267
|
+
def create_chunk(n)
|
268
|
+
chunk = BSON::OrderedHash.new
|
269
|
+
chunk['_id'] = BSON::ObjectId.new
|
270
|
+
chunk['n'] = n
|
271
|
+
chunk['files_id'] = @files_id
|
272
|
+
chunk['data'] = ''
|
273
|
+
@chunk_position = 0
|
274
|
+
chunk
|
275
|
+
end
|
276
|
+
|
277
|
+
def save_chunk(chunk)
|
278
|
+
@chunks.insert(chunk)
|
279
|
+
end
|
280
|
+
|
281
|
+
def get_chunk(n)
|
282
|
+
chunk = @chunks.find({'files_id' => @files_id, 'n' => n}).next_document
|
283
|
+
@chunk_position = 0
|
284
|
+
chunk
|
285
|
+
end
|
286
|
+
|
287
|
+
# Read a file in its entirety.
|
288
|
+
def read_all
|
289
|
+
buf = ''
|
290
|
+
if @current_chunk
|
291
|
+
buf << @current_chunk['data'].to_s
|
292
|
+
while buf.size < @file_length
|
293
|
+
@current_chunk = get_chunk(@current_chunk['n'] + 1)
|
294
|
+
break if @current_chunk.nil?
|
295
|
+
buf << @current_chunk['data'].to_s
|
296
|
+
end
|
297
|
+
@file_position = @file_length
|
298
|
+
end
|
299
|
+
buf
|
300
|
+
end
|
301
|
+
|
302
|
+
# Read a file incrementally.
|
303
|
+
def read_length(length)
|
304
|
+
cache_chunk_data
|
305
|
+
remaining = (@file_length - @file_position)
|
306
|
+
if length.nil?
|
307
|
+
to_read = remaining
|
308
|
+
else
|
309
|
+
to_read = length > remaining ? remaining : length
|
310
|
+
end
|
311
|
+
return nil unless remaining > 0
|
312
|
+
|
313
|
+
buf = ''
|
314
|
+
while to_read > 0
|
315
|
+
if @chunk_position == @chunk_data_length
|
316
|
+
@current_chunk = get_chunk(@current_chunk['n'] + 1)
|
317
|
+
cache_chunk_data
|
318
|
+
end
|
319
|
+
chunk_remainder = @chunk_data_length - @chunk_position
|
320
|
+
size = (to_read >= chunk_remainder) ? chunk_remainder : to_read
|
321
|
+
buf << @current_chunk_data[@chunk_position, size]
|
322
|
+
to_read -= size
|
323
|
+
@chunk_position += size
|
324
|
+
@file_position += size
|
325
|
+
end
|
326
|
+
buf
|
327
|
+
end
|
328
|
+
|
329
|
+
def read_to_character(character="\n", length=nil)
|
330
|
+
result = ''
|
331
|
+
len = 0
|
332
|
+
while char = getc
|
333
|
+
result << char
|
334
|
+
len += 1
|
335
|
+
break if char == character || (length ? len >= length : false)
|
336
|
+
end
|
337
|
+
result.length > 0 ? result : nil
|
338
|
+
end
|
339
|
+
|
340
|
+
def read_to_string(string="\n", length=nil)
|
341
|
+
result = ''
|
342
|
+
len = 0
|
343
|
+
match_idx = 0
|
344
|
+
match_num = string.length - 1
|
345
|
+
to_match = string[match_idx].chr
|
346
|
+
if length
|
347
|
+
matcher = lambda {|idx, num| idx < num && len < length }
|
348
|
+
else
|
349
|
+
matcher = lambda {|idx, num| idx < num}
|
350
|
+
end
|
351
|
+
while matcher.call(match_idx, match_num) && char = getc
|
352
|
+
result << char
|
353
|
+
len += 1
|
354
|
+
if char == to_match
|
355
|
+
while match_idx < match_num do
|
356
|
+
match_idx += 1
|
357
|
+
to_match = string[match_idx].chr
|
358
|
+
char = getc
|
359
|
+
result << char
|
360
|
+
if char != to_match
|
361
|
+
match_idx = 0
|
362
|
+
to_match = string[match_idx].chr
|
363
|
+
break
|
364
|
+
end
|
365
|
+
end
|
366
|
+
end
|
367
|
+
end
|
368
|
+
result.length > 0 ? result : nil
|
369
|
+
end
|
370
|
+
|
371
|
+
def cache_chunk_data
|
372
|
+
@current_chunk_data = @current_chunk['data'].to_s
|
373
|
+
if @current_chunk_data.respond_to?(:force_encoding)
|
374
|
+
@current_chunk_data.force_encoding("binary")
|
375
|
+
end
|
376
|
+
@chunk_data_length = @current_chunk['data'].length
|
377
|
+
end
|
378
|
+
|
379
|
+
def write_string(string)
|
380
|
+
# Since Ruby 1.9.1 doesn't necessarily store one character per byte.
|
381
|
+
if string.respond_to?(:force_encoding)
|
382
|
+
string.force_encoding("binary")
|
383
|
+
end
|
384
|
+
|
385
|
+
to_write = string.length
|
386
|
+
while (to_write > 0) do
|
387
|
+
if @current_chunk && @chunk_position == @chunk_size
|
388
|
+
next_chunk_number = @current_chunk['n'] + 1
|
389
|
+
@current_chunk = create_chunk(next_chunk_number)
|
390
|
+
end
|
391
|
+
chunk_available = @chunk_size - @chunk_position
|
392
|
+
step_size = (to_write > chunk_available) ? chunk_available : to_write
|
393
|
+
@current_chunk['data'] = BSON::Binary.new((@current_chunk['data'].to_s << string[-to_write, step_size]).unpack("c*"))
|
394
|
+
@chunk_position += step_size
|
395
|
+
to_write -= step_size
|
396
|
+
save_chunk(@current_chunk)
|
397
|
+
end
|
398
|
+
string.length - to_write
|
399
|
+
end
|
400
|
+
|
401
|
+
# Initialize the class for reading a file.
|
402
|
+
def init_read
|
403
|
+
doc = @files.find(@query, @query_opts).next_document
|
404
|
+
raise GridFileNotFound, "Could not open file matching #{@query.inspect} #{@query_opts.inspect}" unless doc
|
405
|
+
|
406
|
+
@files_id = doc['_id']
|
407
|
+
@content_type = doc['contentType']
|
408
|
+
@chunk_size = doc['chunkSize']
|
409
|
+
@upload_date = doc['uploadDate']
|
410
|
+
@aliases = doc['aliases']
|
411
|
+
@file_length = doc['length']
|
412
|
+
@metadata = doc['metadata']
|
413
|
+
@md5 = doc['md5']
|
414
|
+
@filename = doc['filename']
|
415
|
+
@custom_attrs = doc
|
416
|
+
|
417
|
+
@current_chunk = get_chunk(0)
|
418
|
+
@file_position = 0
|
419
|
+
end
|
420
|
+
|
421
|
+
# Initialize the class for writing a file.
|
422
|
+
def init_write(opts)
|
423
|
+
opts = opts.dup
|
424
|
+
@files_id = opts.delete(:_id) || BSON::ObjectId.new
|
425
|
+
@content_type = opts.delete(:content_type) || (defined? MIME) && get_content_type || DEFAULT_CONTENT_TYPE
|
426
|
+
@chunk_size = opts.delete(:chunk_size) || DEFAULT_CHUNK_SIZE
|
427
|
+
@metadata = opts.delete(:metadata)
|
428
|
+
@aliases = opts.delete(:aliases)
|
429
|
+
@file_length = 0
|
430
|
+
opts.each {|k, v| self[k] = v}
|
431
|
+
check_existing_file if @safe
|
432
|
+
|
433
|
+
@current_chunk = create_chunk(0)
|
434
|
+
@file_position = 0
|
435
|
+
end
|
436
|
+
|
437
|
+
def check_existing_file
|
438
|
+
if @files.find_one('_id' => @files_id)
|
439
|
+
raise GridError, "Attempting to overwrite with Grid#put. You must delete the file first."
|
440
|
+
end
|
441
|
+
end
|
442
|
+
|
443
|
+
def to_mongo_object
|
444
|
+
h = BSON::OrderedHash.new
|
445
|
+
h['_id'] = @files_id
|
446
|
+
h['filename'] = @filename if @filename
|
447
|
+
h['contentType'] = @content_type
|
448
|
+
h['length'] = @current_chunk ? @current_chunk['n'] * @chunk_size + @chunk_position : 0
|
449
|
+
h['chunkSize'] = @chunk_size
|
450
|
+
h['uploadDate'] = @upload_date
|
451
|
+
h['aliases'] = @aliases if @aliases
|
452
|
+
h['metadata'] = @metadata if @metadata
|
453
|
+
h['md5'] = get_md5
|
454
|
+
h.merge!(@custom_attrs)
|
455
|
+
h
|
456
|
+
end
|
457
|
+
|
458
|
+
# Get a server-side md5 and validate against the client if running in safe mode.
|
459
|
+
def get_md5
|
460
|
+
md5_command = BSON::OrderedHash.new
|
461
|
+
md5_command['filemd5'] = @files_id
|
462
|
+
md5_command['root'] = @fs_name
|
463
|
+
@server_md5 = @files.db.command(md5_command)['md5']
|
464
|
+
if @safe
|
465
|
+
@client_md5 = @local_md5.hexdigest
|
466
|
+
if @local_md5 == @server_md5
|
467
|
+
@server_md5
|
468
|
+
else
|
469
|
+
raise GridMD5Failure, "File on server failed MD5 check"
|
470
|
+
end
|
471
|
+
else
|
472
|
+
@server_md5
|
473
|
+
end
|
474
|
+
end
|
475
|
+
|
476
|
+
# Determine the content type based on the filename.
|
477
|
+
def get_content_type
|
478
|
+
if @filename
|
479
|
+
if types = MIME::Types.type_for(@filename)
|
480
|
+
types.first.simplified unless types.empty?
|
481
|
+
end
|
482
|
+
end
|
483
|
+
end
|
484
|
+
end
|
485
|
+
end
|