longleaf 0.1.0 → 0.2.0.pre.1
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +4 -4
- data/.editorconfig +13 -0
- data/.gitignore +1 -0
- data/.rubocop.yml +4 -0
- data/.rubocop_todo.yml +755 -0
- data/README.md +29 -7
- data/lib/longleaf/candidates/file_selector.rb +107 -0
- data/lib/longleaf/candidates/service_candidate_filesystem_iterator.rb +99 -0
- data/lib/longleaf/candidates/service_candidate_locator.rb +18 -0
- data/lib/longleaf/cli.rb +102 -6
- data/lib/longleaf/commands/deregister_command.rb +50 -0
- data/lib/longleaf/commands/preserve_command.rb +45 -0
- data/lib/longleaf/commands/register_command.rb +24 -38
- data/lib/longleaf/commands/validate_config_command.rb +6 -2
- data/lib/longleaf/commands/validate_metadata_command.rb +49 -0
- data/lib/longleaf/errors.rb +19 -0
- data/lib/longleaf/events/deregister_event.rb +55 -0
- data/lib/longleaf/events/event_names.rb +9 -0
- data/lib/longleaf/events/event_status_tracking.rb +59 -0
- data/lib/longleaf/events/preserve_event.rb +71 -0
- data/lib/longleaf/events/register_event.rb +37 -26
- data/lib/longleaf/helpers/digest_helper.rb +50 -0
- data/lib/longleaf/helpers/service_date_helper.rb +51 -0
- data/lib/longleaf/logging.rb +1 -0
- data/lib/longleaf/logging/redirecting_logger.rb +9 -8
- data/lib/longleaf/models/app_fields.rb +2 -0
- data/lib/longleaf/models/file_record.rb +8 -3
- data/lib/longleaf/models/md_fields.rb +1 -0
- data/lib/longleaf/models/metadata_record.rb +16 -4
- data/lib/longleaf/models/service_definition.rb +4 -3
- data/lib/longleaf/models/service_fields.rb +2 -0
- data/lib/longleaf/models/service_record.rb +4 -1
- data/lib/longleaf/models/storage_location.rb +18 -1
- data/lib/longleaf/preservation_services/fixity_check_service.rb +121 -0
- data/lib/longleaf/preservation_services/rsync_replication_service.rb +183 -0
- data/lib/longleaf/services/application_config_deserializer.rb +4 -6
- data/lib/longleaf/services/application_config_manager.rb +4 -2
- data/lib/longleaf/services/application_config_validator.rb +1 -1
- data/lib/longleaf/services/configuration_validator.rb +1 -0
- data/lib/longleaf/services/metadata_deserializer.rb +47 -10
- data/lib/longleaf/services/metadata_serializer.rb +42 -6
- data/lib/longleaf/services/service_class_cache.rb +112 -0
- data/lib/longleaf/services/service_definition_manager.rb +5 -1
- data/lib/longleaf/services/service_definition_validator.rb +4 -4
- data/lib/longleaf/services/service_manager.rb +72 -9
- data/lib/longleaf/services/service_mapping_manager.rb +4 -3
- data/lib/longleaf/services/service_mapping_validator.rb +4 -4
- data/lib/longleaf/services/storage_location_manager.rb +26 -5
- data/lib/longleaf/services/storage_location_validator.rb +1 -1
- data/lib/longleaf/services/storage_path_validator.rb +2 -2
- data/lib/longleaf/specs/config_builder.rb +9 -5
- data/lib/longleaf/specs/custom_matchers.rb +9 -0
- data/lib/longleaf/specs/file_helpers.rb +60 -0
- data/lib/longleaf/version.rb +1 -1
- data/longleaf.gemspec +1 -0
- metadata +39 -7
- data/lib/longleaf/commands/abstract_command.rb +0 -37
@@ -0,0 +1,183 @@
|
|
1
|
+
require 'longleaf/events/event_names'
|
2
|
+
require 'longleaf/logging'
|
3
|
+
require 'longleaf/errors'
|
4
|
+
require 'longleaf/models/file_record'
|
5
|
+
require 'longleaf/models/service_fields'
|
6
|
+
require 'longleaf/events/register_event'
|
7
|
+
require 'open3'
|
8
|
+
|
9
|
+
module Longleaf
|
10
|
+
# Preservation service which performs replication of a file to one or more destinations using rsync.
|
11
|
+
#
|
12
|
+
# The service definition must contain one or more destinations, specified with the "to" property.
|
13
|
+
# These destinations must be either a known storage location name, a remote path, or absolute path.
|
14
|
+
#
|
15
|
+
# Optional service configuration properties:
|
16
|
+
# * replica_collision_policy = specifies the desired outcome if the service attempts to replicate
|
17
|
+
# a file which already exists at a destination. Default: "replace".
|
18
|
+
# * rsync_command = the command to invoke in order to execute rsync. Default: "rsync"
|
19
|
+
# * rsync_options = additional parameters that will be passed along to rsync. Cannot include options
|
20
|
+
# which change the target of the command or prevent its execution, such as "files-from", "dry-run",
|
21
|
+
# "help", etc. Command will always include "-R". Default "-a".
|
22
|
+
class RsyncReplicationService
|
23
|
+
include Longleaf::Logging
|
24
|
+
|
25
|
+
COLLISION_PROPERTY = "replica_collision_policy"
|
26
|
+
DEFAULT_COLLISION_POLICY = "replace"
|
27
|
+
VALID_COLLISION_POLICIES = ["replace"]
|
28
|
+
|
29
|
+
RSYNC_COMMAND_PROPERTY = "rsync_command"
|
30
|
+
DEFAULT_COMMAND = "rsync"
|
31
|
+
|
32
|
+
RSYNC_OPTIONS_PROPERTY = "rsync_options"
|
33
|
+
DEFAULT_OPTIONS = "-a"
|
34
|
+
DISALLOWED_OPTIONS = ["files-from", "n", "dry-run", "exclude", "exclude-from", "cvs-exclude",
|
35
|
+
"h", "help", "f", "F", "filter"]
|
36
|
+
|
37
|
+
attr_reader :command, :options, :collision_policy
|
38
|
+
|
39
|
+
# Initialize a RsyncReplicationService from the given service definition
|
40
|
+
#
|
41
|
+
# @param service_def [ServiceDefinition] the configuration for this service
|
42
|
+
# @param location_manager [StorageLocationManager] manager for configured storage locations
|
43
|
+
def initialize(service_def, app_manager)
|
44
|
+
@service_def = service_def
|
45
|
+
@app_manager = app_manager
|
46
|
+
|
47
|
+
@command = @service_def.properties[RSYNC_COMMAND_PROPERTY] || DEFAULT_COMMAND
|
48
|
+
|
49
|
+
# Validate rsync parameters
|
50
|
+
@options = @service_def.properties[RSYNC_OPTIONS_PROPERTY] || DEFAULT_OPTIONS
|
51
|
+
if contains_disallowed_option?(@options)
|
52
|
+
raise ArgumentError.new("Service #{service_def.name} specifies a disallowed rsync paramter," \
|
53
|
+
+ " rsync_options may not include the following: #{DISALLOWED_OPTIONS.join(" ")}")
|
54
|
+
end
|
55
|
+
|
56
|
+
# Add -R (--relative) in to command options to ensure full path gets replicated
|
57
|
+
@options = @options + " -R"
|
58
|
+
|
59
|
+
# Set and validate the replica collision policy
|
60
|
+
@collision_policy = @service_def.properties[COLLISION_PROPERTY] || DEFAULT_COLLISION_POLICY
|
61
|
+
if !VALID_COLLISION_POLICIES.include?(@collision_policy)
|
62
|
+
raise ArgumentError.new("Service #{service_def.name} received invalid #{COLLISION_PROPERTY}" \
|
63
|
+
+ " value #{collision_policy}")
|
64
|
+
end
|
65
|
+
|
66
|
+
# Store and validate destinations
|
67
|
+
replicate_to = @service_def.properties[ServiceFields::REPLICATE_TO]
|
68
|
+
if replicate_to.nil? || replicate_to.empty?
|
69
|
+
raise ArgumentError.new("Service #{service_def.name} must provide one or more replication destinations.")
|
70
|
+
end
|
71
|
+
|
72
|
+
loc_manager = app_manager.location_manager
|
73
|
+
# Build list of destinations, translating to storage locations when relevant
|
74
|
+
@destinations = Array.new
|
75
|
+
replicate_to.each do |dest|
|
76
|
+
# Assume that if destination contains a : or / it is a path rather than storage location
|
77
|
+
if dest =~ /[:\/]/
|
78
|
+
@destinations << dest
|
79
|
+
else
|
80
|
+
if loc_manager.locations.key?(dest)
|
81
|
+
@destinations << loc_manager.locations[dest]
|
82
|
+
else
|
83
|
+
raise ArgumentError.new("Service #{service_def.name} specifies unknown storage location '#{dest}'" \
|
84
|
+
+ " as a replication destination")
|
85
|
+
end
|
86
|
+
end
|
87
|
+
end
|
88
|
+
end
|
89
|
+
|
90
|
+
# During a replication event, perform replication of the specified file to all configured destinations
|
91
|
+
# as necessary.
|
92
|
+
#
|
93
|
+
# @param file_rec [FileRecord] record representing the file to perform the service on.
|
94
|
+
# @param event [String] name of the event this service is being invoked by.
|
95
|
+
# @raise [PreservationServiceError] if the rsync replication fails
|
96
|
+
def perform(file_rec, event)
|
97
|
+
@destinations.each do |destination|
|
98
|
+
dest_is_storage_loc = destination.is_a?(Longleaf::StorageLocation)
|
99
|
+
|
100
|
+
if dest_is_storage_loc
|
101
|
+
dest_path = destination.path
|
102
|
+
else
|
103
|
+
dest_path = destination
|
104
|
+
end
|
105
|
+
|
106
|
+
# Determine the path to the file being replicated relative to its storage location
|
107
|
+
rel_path = file_rec.path.sub(/\A#{file_rec.storage_location.path}/, "")
|
108
|
+
# source path with . so that rsync will only create destination directories starting from that point
|
109
|
+
source_path = File.join(file_rec.storage_location.path, "./#{rel_path}")
|
110
|
+
|
111
|
+
# Check that the destination is available because attempting to write
|
112
|
+
verify_destination_available(destination, file_rec)
|
113
|
+
|
114
|
+
logger.debug("Invoking rsync with command: #{@command} \"#{source_path}\" \"#{dest_path}\" #{@options}")
|
115
|
+
stdout, stderr, status = Open3.capture3("#{@command} \"#{source_path}\" \"#{dest_path}\" #{@options}")
|
116
|
+
raise PreservationServiceError.new("Failed to replicate #{file_rec.path} to #{dest_path}: #{stderr}") \
|
117
|
+
unless status.success?
|
118
|
+
|
119
|
+
logger.info("Replicated #{file_rec.path} to destination #{dest_path}")
|
120
|
+
|
121
|
+
# For destinations which are storage locations, register the replica with longleaf
|
122
|
+
if dest_is_storage_loc
|
123
|
+
register_replica(destination, rel_path, file_rec)
|
124
|
+
end
|
125
|
+
end
|
126
|
+
end
|
127
|
+
|
128
|
+
# Determine if this service is applicable for the provided event, given the configured service definition
|
129
|
+
#
|
130
|
+
# @param event [String] name of the event
|
131
|
+
# @return [Boolean] returns true if this service is applicable for the provided event
|
132
|
+
def is_applicable?(event)
|
133
|
+
case event
|
134
|
+
when EventNames::PRESERVE
|
135
|
+
true
|
136
|
+
else
|
137
|
+
false
|
138
|
+
end
|
139
|
+
end
|
140
|
+
|
141
|
+
private
|
142
|
+
def contains_disallowed_option?(options)
|
143
|
+
DISALLOWED_OPTIONS.each do |disallowed|
|
144
|
+
if disallowed.length == 1
|
145
|
+
if options =~ /(\A| )-[a-zA-Z0-9]*#{disallowed}[a-zA-Z0-9]*( |=|\z)/
|
146
|
+
return true
|
147
|
+
end
|
148
|
+
else
|
149
|
+
if options =~ /(\A| )--#{disallowed}( |=|\z)/
|
150
|
+
return true
|
151
|
+
end
|
152
|
+
end
|
153
|
+
end
|
154
|
+
|
155
|
+
false
|
156
|
+
end
|
157
|
+
|
158
|
+
def verify_destination_available(destination, file_rec)
|
159
|
+
if destination.is_a?(Longleaf::StorageLocation)
|
160
|
+
begin
|
161
|
+
destination.available?
|
162
|
+
rescue StorageLocationUnavailableError => e
|
163
|
+
raise StorageLocationUnavailableError.new("Cannot replicate #{file_rec.path} to destination" \
|
164
|
+
+ " storage location #{destination.name}, it is unavailable.") unless destination.available?
|
165
|
+
end
|
166
|
+
elsif destination.start_with?("/")
|
167
|
+
raise StorageLocationUnavailableError.new("Cannot replicate #{file_rec.path} to destination" \
|
168
|
+
+ " #{destination}, path does not exist.") unless Dir.exist?(destination)
|
169
|
+
end
|
170
|
+
end
|
171
|
+
|
172
|
+
def register_replica(destination, rel_path, file_rec)
|
173
|
+
dest_file_path = File.join(destination.path, rel_path)
|
174
|
+
dest_file_rec = FileRecord.new(dest_file_path, destination)
|
175
|
+
|
176
|
+
register_event = RegisterEvent.new(file_rec: dest_file_rec,
|
177
|
+
app_manager: @app_manager,
|
178
|
+
force: true,
|
179
|
+
checksums: file_rec.metadata_record.checksums)
|
180
|
+
register_event.perform
|
181
|
+
end
|
182
|
+
end
|
183
|
+
end
|
@@ -1,14 +1,14 @@
|
|
1
1
|
require 'longleaf/services/application_config_validator'
|
2
2
|
require 'longleaf/services/application_config_manager'
|
3
3
|
|
4
|
-
# Deserializer for application configuration files
|
5
4
|
module Longleaf
|
5
|
+
# Deserializer for application configuration files
|
6
6
|
class ApplicationConfigDeserializer
|
7
7
|
|
8
8
|
# Deserializes a valid application configuration file as a ApplicationConfigManager option
|
9
9
|
# @param config_path [String] file path to the application configuration file
|
10
10
|
# @param format [String] encoding format of the config file
|
11
|
-
# return [
|
11
|
+
# return [ApplicationConfigManager] manager for the loaded configuration
|
12
12
|
def self.deserialize(config_path, format: 'yaml')
|
13
13
|
config = load(config_path, format: format)
|
14
14
|
|
@@ -35,11 +35,9 @@ module Longleaf
|
|
35
35
|
YAML.load_file(config_path)
|
36
36
|
rescue Errno::ENOENT => err
|
37
37
|
raise Longleaf::ConfigurationError.new(
|
38
|
-
"
|
38
|
+
"Configuration file #{config_path} does not exist.")
|
39
39
|
rescue => err
|
40
|
-
raise Longleaf::ConfigurationError.new(
|
41
|
-
%Q(Failed to load application configuration due to the following reason:
|
42
|
-
#{err.message}))
|
40
|
+
raise Longleaf::ConfigurationError.new(err)
|
43
41
|
end
|
44
42
|
end
|
45
43
|
end
|
@@ -6,8 +6,8 @@ require_relative 'service_mapping_validator'
|
|
6
6
|
require_relative 'service_mapping_manager'
|
7
7
|
require_relative 'service_manager'
|
8
8
|
|
9
|
-
# Manager which loads and provides access to the configuration of the application
|
10
9
|
module Longleaf
|
10
|
+
# Manager which loads and provides access to the configuration of the application
|
11
11
|
class ApplicationConfigManager
|
12
12
|
attr_reader :service_manager
|
13
13
|
attr_reader :location_manager
|
@@ -18,7 +18,9 @@ module Longleaf
|
|
18
18
|
definition_manager = Longleaf::ServiceDefinitionManager.new(config)
|
19
19
|
mapping_manager = Longleaf::ServiceMappingManager.new(config)
|
20
20
|
@service_manager = Longleaf::ServiceManager.new(
|
21
|
-
definition_manager: definition_manager,
|
21
|
+
definition_manager: definition_manager,
|
22
|
+
mapping_manager: mapping_manager,
|
23
|
+
app_manager: self)
|
22
24
|
end
|
23
25
|
end
|
24
26
|
end
|
@@ -2,8 +2,8 @@ require_relative 'storage_location_validator'
|
|
2
2
|
require_relative 'service_definition_validator'
|
3
3
|
require_relative 'service_mapping_validator'
|
4
4
|
|
5
|
-
# Validator for Longleaf application configuration
|
6
5
|
module Longleaf
|
6
|
+
# Validator for Longleaf application configuration
|
7
7
|
class ApplicationConfigValidator
|
8
8
|
|
9
9
|
# Validates the application configuration provided. Will raise ConfigurationError
|
@@ -1,26 +1,28 @@
|
|
1
1
|
require 'yaml'
|
2
|
-
|
3
|
-
|
4
|
-
|
2
|
+
require 'longleaf/models/metadata_record'
|
3
|
+
require 'longleaf/models/md_fields'
|
4
|
+
require 'longleaf/errors'
|
5
|
+
require 'longleaf/logging'
|
5
6
|
|
6
|
-
# Service which deserializes metadata files into MetadataRecord objects
|
7
7
|
module Longleaf
|
8
|
+
# Service which deserializes metadata files into MetadataRecord objects
|
8
9
|
class MetadataDeserializer
|
9
|
-
|
10
|
+
extend Longleaf::Logging
|
11
|
+
MDF ||= MDFields
|
10
12
|
|
11
13
|
# Deserialize a file into a MetadataRecord object
|
12
14
|
#
|
13
15
|
# @param file_path [String] path of the file to read. Required.
|
14
16
|
# @param format [String] format the file is stored in. Default is 'yaml'.
|
15
|
-
def self.deserialize(file_path:, format: 'yaml')
|
17
|
+
def self.deserialize(file_path:, format: 'yaml', digest_algs: [])
|
16
18
|
case format
|
17
19
|
when 'yaml'
|
18
|
-
md = from_yaml(file_path)
|
20
|
+
md = from_yaml(file_path, digest_algs)
|
19
21
|
else
|
20
22
|
raise ArgumentError.new('Invalid deserialization format #{format} specified')
|
21
23
|
end
|
22
24
|
|
23
|
-
if !md || !md.key?(MDF::DATA) || !md.key?(MDF::SERVICES)
|
25
|
+
if !md || !md.is_a?(Hash) || !md.key?(MDF::DATA) || !md.key?(MDF::SERVICES)
|
24
26
|
raise Longleaf::MetadataError.new("Invalid metadata file, did not contain data or services fields: #{file_path}")
|
25
27
|
end
|
26
28
|
|
@@ -61,8 +63,43 @@ module Longleaf
|
|
61
63
|
last_modified: last_modified)
|
62
64
|
end
|
63
65
|
|
64
|
-
|
65
|
-
|
66
|
+
# Load configuration a yaml encoded configuration file
|
67
|
+
def self.from_yaml(file_path, digest_algs)
|
68
|
+
File.open(file_path, 'r:bom|utf-8') do |f|
|
69
|
+
contents = f.read
|
70
|
+
|
71
|
+
verify_digests(file_path, contents, digest_algs)
|
72
|
+
|
73
|
+
YAML.load(contents)
|
74
|
+
end
|
75
|
+
end
|
76
|
+
|
77
|
+
def self.verify_digests(file_path, contents, digest_algs)
|
78
|
+
return if digest_algs.nil? || digest_algs.empty?
|
79
|
+
|
80
|
+
digest_algs.each do |alg|
|
81
|
+
if file_path.respond_to?(:path)
|
82
|
+
path = file_path.path
|
83
|
+
else
|
84
|
+
path = file_path
|
85
|
+
end
|
86
|
+
digest_path = "#{path}.#{alg}"
|
87
|
+
unless File.exist?(digest_path)
|
88
|
+
logger.warn("Missing expected #{alg} digest for #{path}")
|
89
|
+
next
|
90
|
+
end
|
91
|
+
|
92
|
+
digest = DigestHelper::start_digest(alg)
|
93
|
+
result = digest.hexdigest(contents)
|
94
|
+
existing_digest = IO.read(digest_path)
|
95
|
+
|
96
|
+
if result == existing_digest
|
97
|
+
logger.info("Metadata fixity check using algorithm '#{alg}' succeeded for file #{path}")
|
98
|
+
else
|
99
|
+
raise ChecksumMismatchError.new("Metadata digest of type #{alg} did not match the contents of #{path}:" \
|
100
|
+
+ " expected #{existing_digest}, calculated #{result}")
|
101
|
+
end
|
102
|
+
end
|
66
103
|
end
|
67
104
|
end
|
68
105
|
end
|
@@ -1,20 +1,27 @@
|
|
1
1
|
require 'yaml'
|
2
|
-
|
3
|
-
|
2
|
+
require 'longleaf/models/metadata_record'
|
3
|
+
require 'longleaf/models/md_fields'
|
4
|
+
require 'longleaf/helpers/digest_helper'
|
5
|
+
require 'longleaf/errors'
|
6
|
+
require 'longleaf/logging'
|
7
|
+
require 'pathname'
|
4
8
|
|
5
|
-
# Service which serializes MetadataRecord objects
|
6
9
|
module Longleaf
|
10
|
+
# Service which serializes MetadataRecord objects
|
7
11
|
class MetadataSerializer
|
8
|
-
|
12
|
+
extend Longleaf::Logging
|
13
|
+
MDF ||= MDFields
|
9
14
|
|
10
15
|
# Serialize the contents of the provided metadata record to the specified path
|
11
16
|
#
|
12
17
|
# @param metadata [MetadataRecord] metadata record to serialize. Required.
|
13
18
|
# @param file_path [String] path to write the file to. Required.
|
14
19
|
# @param format [String] format to serialize the metadata in. Default is 'yaml'.
|
15
|
-
|
20
|
+
# @param digest_algs [Array] if provided, sidecar digest files for the metadata file
|
21
|
+
# will be generated for each algorithm.
|
22
|
+
def self.write(metadata:, file_path:, format: 'yaml', digest_algs: [])
|
16
23
|
raise ArgumentError.new('metadata parameter must be a MetadataRecord') \
|
17
|
-
unless metadata.class ==
|
24
|
+
unless metadata.class == MetadataRecord
|
18
25
|
|
19
26
|
case format
|
20
27
|
when 'yaml'
|
@@ -23,7 +30,12 @@ module Longleaf
|
|
23
30
|
raise ArgumentError.new('Invalid serialization format #{format} specified')
|
24
31
|
end
|
25
32
|
|
33
|
+
# Fill in parent directories if they do not exist
|
34
|
+
parent_dir = Pathname(file_path).parent
|
35
|
+
parent_dir.mkpath unless parent_dir.exist?
|
36
|
+
|
26
37
|
File.write(file_path, content)
|
38
|
+
write_digests(file_path, content, digest_algs)
|
27
39
|
end
|
28
40
|
|
29
41
|
# @param metadata [MetadataRecord] metadata record to transform
|
@@ -33,6 +45,8 @@ module Longleaf
|
|
33
45
|
props.to_yaml
|
34
46
|
end
|
35
47
|
|
48
|
+
# Create a hash representation of the given MetadataRecord file
|
49
|
+
# @param metadata [MetadataRecord] metadata record to transform into a hash
|
36
50
|
def self.to_hash(metadata)
|
37
51
|
props = Hash.new
|
38
52
|
|
@@ -59,6 +73,9 @@ module Longleaf
|
|
59
73
|
props
|
60
74
|
end
|
61
75
|
|
76
|
+
# @param format [String] encoding format used for metadata file
|
77
|
+
# @return [String] the suffix used to indicate that a file is a metadata file in the provided encoding
|
78
|
+
# @raise [ArgumentError] raised if the provided format is not a supported metadata encoding format
|
62
79
|
def self.metadata_suffix(format: 'yaml')
|
63
80
|
case format
|
64
81
|
when 'yaml'
|
@@ -67,5 +84,24 @@ module Longleaf
|
|
67
84
|
raise ArgumentError.new('Invalid serialization format #{format} specified')
|
68
85
|
end
|
69
86
|
end
|
87
|
+
|
88
|
+
private
|
89
|
+
def self.write_digests(file_path, content, digests)
|
90
|
+
return if digests.nil? || digests.empty?
|
91
|
+
|
92
|
+
digests.each do |alg|
|
93
|
+
digest_class = DigestHelper::start_digest(alg)
|
94
|
+
result = digest_class.hexdigest(content)
|
95
|
+
if file_path.respond_to?(:path)
|
96
|
+
digest_path = "#{file_path.path}.#{alg}"
|
97
|
+
else
|
98
|
+
digest_path = "#{file_path}.#{alg}"
|
99
|
+
end
|
100
|
+
|
101
|
+
File.write(digest_path, result)
|
102
|
+
|
103
|
+
self.logger.debug("Generated #{alg} digest for metadata file #{file_path}: #{result}")
|
104
|
+
end
|
105
|
+
end
|
70
106
|
end
|
71
107
|
end
|
@@ -0,0 +1,112 @@
|
|
1
|
+
require 'pathname'
|
2
|
+
|
3
|
+
module Longleaf
|
4
|
+
# Cache for loading and retrieving preservation service classes
|
5
|
+
class ServiceClassCache
|
6
|
+
STD_PRESERVATION_SERVICE_PATH = 'longleaf/preservation_services/'
|
7
|
+
|
8
|
+
def initialize(app_manager)
|
9
|
+
@app_manager = app_manager
|
10
|
+
# Cache storing per service definition instances of service classes
|
11
|
+
@service_instance_cache = Hash.new
|
12
|
+
# Cache storing per script path class of service
|
13
|
+
@class_cache = Hash.new
|
14
|
+
end
|
15
|
+
|
16
|
+
# Returns an instance of the preversation service defined for the provided service definition,
|
17
|
+
# based on the work_script and work_class properties provided.
|
18
|
+
#
|
19
|
+
# @param service_def [ServiceDefinition] definition of service to instantiate
|
20
|
+
# @return [PreservationService] Instance of the preservation service class for the definition.
|
21
|
+
def service_instance(service_def)
|
22
|
+
service_name = service_def.name
|
23
|
+
# Return the cached instance of the service
|
24
|
+
if @service_instance_cache.key?(service_name)
|
25
|
+
return @service_instance_cache[service_name]
|
26
|
+
end
|
27
|
+
|
28
|
+
clazz = service_class(service_def)
|
29
|
+
# Cache and return the class instance
|
30
|
+
@service_instance_cache[service_name] = clazz.new(service_def, @app_manager)
|
31
|
+
end
|
32
|
+
|
33
|
+
# Load and return the PreservationService class assigned to the provided service definition,
|
34
|
+
# based on the work_script and work_class properties provided.
|
35
|
+
#
|
36
|
+
# @param service_def [ServiceDefinition] definition of service to retrieve class for
|
37
|
+
# @return [Class] class of work_script
|
38
|
+
def service_class(service_def)
|
39
|
+
service_name = service_def.name
|
40
|
+
work_script = service_def.work_script
|
41
|
+
|
42
|
+
if work_script.include?('/')
|
43
|
+
expanded_path = Pathname.new(work_script).expand_path.to_s
|
44
|
+
if !from_permitted_path?(expanded_path)
|
45
|
+
raise ConfigurationError.new("Unable to load work_script for service #{service_name}, #{work_script} is not in a known library path.")
|
46
|
+
end
|
47
|
+
|
48
|
+
last_slash_index = work_script.rindex('/')
|
49
|
+
script_path = work_script[0..last_slash_index]
|
50
|
+
script_name = work_script[(last_slash_index + 1)..-1]
|
51
|
+
else
|
52
|
+
script_path = STD_PRESERVATION_SERVICE_PATH
|
53
|
+
script_name = work_script
|
54
|
+
end
|
55
|
+
|
56
|
+
# Strip off the extension
|
57
|
+
script_name.sub!('.rb', '')
|
58
|
+
|
59
|
+
require_path = File.join(script_path, script_name)
|
60
|
+
# Return the cached Class if this path has been encountered before
|
61
|
+
if @class_cache.key?(require_path)
|
62
|
+
return @class_cache[require_path]
|
63
|
+
end
|
64
|
+
|
65
|
+
# Load the script
|
66
|
+
begin
|
67
|
+
require require_path
|
68
|
+
rescue LoadError => e
|
69
|
+
raise ConfigurationError.new("Failed to load work_script '#{script_name}' for service #{service_name}")
|
70
|
+
end
|
71
|
+
|
72
|
+
# Generate the class name, either configured or from file naming convention if possible
|
73
|
+
if service_def.work_class
|
74
|
+
class_name = service_def.work_class
|
75
|
+
else
|
76
|
+
class_name = script_name.split('_').map(&:capitalize).join
|
77
|
+
# Assume the longleaf module for classes in the standard path
|
78
|
+
class_name = 'Longleaf::' + class_name if script_path == STD_PRESERVATION_SERVICE_PATH
|
79
|
+
end
|
80
|
+
|
81
|
+
begin
|
82
|
+
class_constant = constantize(class_name)
|
83
|
+
# cache the class for this work_script and return it
|
84
|
+
@class_cache[require_path] = class_constant
|
85
|
+
rescue NameError
|
86
|
+
raise ConfigurationError.new("Failed to load work_script '#{script_name}' for service #{service_name}, class name #{class_name} was not found.")
|
87
|
+
end
|
88
|
+
end
|
89
|
+
|
90
|
+
private
|
91
|
+
# Borrowed from sidekiq implementation
|
92
|
+
def constantize(str)
|
93
|
+
names = str.split('::')
|
94
|
+
names.shift if names.empty? || names.first.empty?
|
95
|
+
|
96
|
+
names.inject(Object) do |constant, name|
|
97
|
+
# the false flag limits search for name to under the constant namespace
|
98
|
+
# which mimics Rails' behaviour
|
99
|
+
constant.const_defined?(name, false) ? constant.const_get(name, false) : constant.const_missing(name)
|
100
|
+
end
|
101
|
+
end
|
102
|
+
|
103
|
+
def from_permitted_path?(script_path)
|
104
|
+
$LOAD_PATH.each do |lib_path|
|
105
|
+
if script_path.start_with?(lib_path)
|
106
|
+
return true
|
107
|
+
end
|
108
|
+
end
|
109
|
+
false
|
110
|
+
end
|
111
|
+
end
|
112
|
+
end
|