etna 0.1.28 → 0.1.29
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +4 -4
- data/lib/etna/clients/magma/workflows/create_project_workflow.rb +1 -1
- data/lib/etna/clients/magma/workflows/crud_workflow.rb +9 -0
- data/lib/etna/clients/magma/workflows/materialize_magma_record_files_workflow.rb +43 -28
- data/lib/etna/clients/magma/workflows/walk_model_tree_workflow.rb +7 -2
- data/lib/etna/clients/metis/client.rb +1 -1
- data/lib/etna/clients/metis/workflows/metis_download_workflow.rb +15 -11
- data/lib/etna/clients/metis/workflows/metis_upload_workflow.rb +83 -13
- data/lib/etna/clients/metis/workflows/sync_metis_data_workflow.rb +43 -79
- data/lib/etna/cwl.rb +4 -0
- data/lib/etna/directed_graph.rb +1 -1
- data/lib/etna/filesystem.rb +143 -15
- data/lib/etna/spec/auth.rb +6 -6
- data/lib/etna/user.rb +2 -6
- metadata +30 -2
checksums.yaml
CHANGED
@@ -1,7 +1,7 @@
|
|
1
1
|
---
|
2
2
|
SHA256:
|
3
|
-
metadata.gz:
|
4
|
-
data.tar.gz:
|
3
|
+
metadata.gz: 8d4c9465edd6fd280a9cd91b9c3abbc3ef08114a46e3ffb41e14a20ce612fe7f
|
4
|
+
data.tar.gz: bfd29114b8db331c92f79be1fe765fa6b64c7b5ee4a8543cd75f6c9fe7c38a84
|
5
5
|
SHA512:
|
6
|
-
metadata.gz:
|
7
|
-
data.tar.gz:
|
6
|
+
metadata.gz: 8ea7b21a563d743cdb6f8a06f92ca15fdeb56ecb291b27ff32a555a8f863cd6ce4400fc6954ae5620a9766bbafbed861609f52309f4de3d407f117d4f9b3f1b1
|
7
|
+
data.tar.gz: 6a9530ee4beb8e6a6d7940c30a2a2bc7412dc4420492e650cb731eac4e5e958a4ca630a966aede4591706ae79212e3290b1a080b5fafe5549df24cc302cbc82b
|
@@ -70,7 +70,7 @@ module Etna
|
|
70
70
|
puts "Creating Janus project."
|
71
71
|
create_janus_project!
|
72
72
|
puts "Done! Adding you as an administrator on the project."
|
73
|
-
add_janus_user(user['email'], "#{user['
|
73
|
+
add_janus_user(user['email'], "#{user['name']}", 'editor')
|
74
74
|
promote_to_administrator(user['email'])
|
75
75
|
update_magma_client_token!
|
76
76
|
|
@@ -36,8 +36,17 @@ module Etna
|
|
36
36
|
documents = Documents.new({})
|
37
37
|
last_page = nil
|
38
38
|
while last_page.nil? || last_page.models.model_keys.map { |k| last_page.models.model(k).documents.raw.length }.sum > 0
|
39
|
+
attempts = 0
|
39
40
|
begin
|
41
|
+
attempts += 1
|
40
42
|
last_page = magma_client.retrieve(request)
|
43
|
+
# Unfortunately, paging in magma is not great and times out from time to time.
|
44
|
+
rescue Net::ReadTimeout => e
|
45
|
+
if attempts > 5
|
46
|
+
raise e
|
47
|
+
end
|
48
|
+
|
49
|
+
retry
|
41
50
|
rescue Etna::Error => e
|
42
51
|
raise e unless e.message.include?('not found')
|
43
52
|
break
|
@@ -1,5 +1,4 @@
|
|
1
1
|
require 'ostruct'
|
2
|
-
require 'digest'
|
3
2
|
require 'fileutils'
|
4
3
|
require 'tempfile'
|
5
4
|
|
@@ -9,11 +8,11 @@ module Etna
|
|
9
8
|
class MaterializeDataWorkflow < Struct.new(
|
10
9
|
:metis_client, :magma_client, :project_name,
|
11
10
|
:model_name, :model_filters, :model_attributes_mask,
|
12
|
-
:filesystem, :logger, :stub_files,
|
13
|
-
:
|
11
|
+
:filesystem, :logger, :stub_files, :concurrency,
|
12
|
+
:record_names, keyword_init: true)
|
14
13
|
|
15
14
|
def initialize(**kwds)
|
16
|
-
super(**({filesystem: Etna::Filesystem.new}.update(kwds)))
|
15
|
+
super(**({filesystem: Etna::Filesystem.new, concurrency: 10, record_names: "all"}.update(kwds)))
|
17
16
|
end
|
18
17
|
|
19
18
|
def magma_crud
|
@@ -25,31 +24,47 @@ module Etna
|
|
25
24
|
end
|
26
25
|
|
27
26
|
def materialize_all(dest)
|
28
|
-
|
27
|
+
templates = {}
|
28
|
+
|
29
|
+
semaphore = Concurrent::Semaphore.new(concurrency)
|
30
|
+
errors = Queue.new
|
31
|
+
|
32
|
+
model_walker.walk_from(
|
33
|
+
model_name,
|
34
|
+
record_names,
|
35
|
+
model_attributes_mask: model_attributes_mask,
|
36
|
+
model_filters: model_filters,
|
37
|
+
page_size: 500,
|
38
|
+
) do |template, document|
|
39
|
+
logger&.info("Materializing #{template.name}##{document[template.identifier]}")
|
40
|
+
templates[template.name] = template
|
41
|
+
|
42
|
+
begin
|
43
|
+
if (error = errors.pop(true))
|
44
|
+
raise error
|
45
|
+
end
|
46
|
+
rescue ThreadError
|
47
|
+
end
|
29
48
|
|
30
|
-
|
31
|
-
|
32
|
-
|
33
|
-
|
34
|
-
|
35
|
-
|
36
|
-
|
37
|
-
|
49
|
+
semaphore.acquire
|
50
|
+
Thread.new do
|
51
|
+
begin
|
52
|
+
materialize_record(dest, template, document)
|
53
|
+
rescue => e
|
54
|
+
errors << e
|
55
|
+
ensure
|
56
|
+
semaphore.release
|
57
|
+
end
|
38
58
|
end
|
39
|
-
ensure
|
40
|
-
filesystem.rm_rf(tmpdir) unless skip_tmpdir
|
41
59
|
end
|
42
|
-
end
|
43
60
|
|
44
|
-
|
45
|
-
|
46
|
-
|
47
|
-
|
48
|
-
|
49
|
-
template = model.template
|
50
|
-
model.documents.document_keys.each do |key|
|
51
|
-
yield template, model.documents.document(key)
|
61
|
+
semaphore.acquire(concurrency)
|
62
|
+
|
63
|
+
begin
|
64
|
+
if (error = errors.pop(true))
|
65
|
+
raise error
|
52
66
|
end
|
67
|
+
rescue ThreadError
|
53
68
|
end
|
54
69
|
end
|
55
70
|
|
@@ -78,11 +93,10 @@ module Etna
|
|
78
93
|
@sync_metis_data_workflow ||= Etna::Clients::Metis::SyncMetisDataWorkflow.new(
|
79
94
|
metis_client: metis_client,
|
80
95
|
logger: logger,
|
81
|
-
skip_tmpdir: skip_tmpdir,
|
82
96
|
filesystem: filesystem)
|
83
97
|
end
|
84
98
|
|
85
|
-
def materialize_record(dest_dir,
|
99
|
+
def materialize_record(dest_dir, template, record)
|
86
100
|
record_to_serialize = record.dup
|
87
101
|
|
88
102
|
each_file(template, record) do |attr_name, url, filename, idx|
|
@@ -91,13 +105,14 @@ module Etna
|
|
91
105
|
end
|
92
106
|
|
93
107
|
dest_file = File.join(dest_dir, metadata_file_name(record_name: record[template.identifier], record_model_name: template.name, ext: "_#{attr_name}_#{idx}#{File.extname(filename)}"))
|
94
|
-
sync_metis_data_workflow.copy_file(
|
95
|
-
record_to_serialize[attr_name] << {
|
108
|
+
sync_metis_data_workflow.copy_file(dest: dest_file, url: url, stub: stub_files)
|
109
|
+
record_to_serialize[attr_name] << {file: dest_file, original_filename: filename}
|
96
110
|
end
|
97
111
|
|
98
112
|
dest_file = File.join(dest_dir, metadata_file_name(record_name: record[template.identifier], record_model_name: template.name, ext: '.json'))
|
99
113
|
filesystem.mkdir_p(File.dirname(dest_file))
|
100
114
|
json = record_to_serialize.to_json
|
115
|
+
|
101
116
|
filesystem.with_writeable(dest_file, "w", size_hint: json.bytes.length) do |io|
|
102
117
|
io.write(json)
|
103
118
|
end
|
@@ -7,6 +7,10 @@ module Etna
|
|
7
7
|
module Clients
|
8
8
|
class Magma
|
9
9
|
class WalkModelTreeWorkflow < Struct.new(:magma_crud, :logger, keyword_init: true)
|
10
|
+
def initialize(**args)
|
11
|
+
super(**({}.update(args)))
|
12
|
+
end
|
13
|
+
|
10
14
|
def walk_from(
|
11
15
|
model_name,
|
12
16
|
record_names = 'all',
|
@@ -43,7 +47,8 @@ module Etna
|
|
43
47
|
|
44
48
|
template.attributes.attribute_keys.each do |attr_name|
|
45
49
|
attributes_mask = model_attributes_mask[model_name]
|
46
|
-
|
50
|
+
black_listed = !attributes_mask.nil? && !attributes_mask.include?(attr_name)
|
51
|
+
next if black_listed && attr_name != template.identifier && attr_name != 'parent'
|
47
52
|
attributes << attr_name
|
48
53
|
|
49
54
|
attr = template.attributes.attribute(attr_name)
|
@@ -58,7 +63,7 @@ module Etna
|
|
58
63
|
elsif attr.attribute_type == AttributeType::CHILD
|
59
64
|
related_models[attr.link_model_name] ||= Set.new
|
60
65
|
links << attr_name
|
61
|
-
elsif attr.attribute_type == AttributeType::PARENT
|
66
|
+
elsif attr.attribute_type == AttributeType::PARENT && !black_listed
|
62
67
|
related_models[attr.link_model_name] ||= Set.new
|
63
68
|
links << attr_name
|
64
69
|
end
|
@@ -13,23 +13,27 @@ module Etna
|
|
13
13
|
end
|
14
14
|
|
15
15
|
# TODO: Might be possible to use range headers to select and resume downloads on failure in the future.
|
16
|
-
def do_download(
|
16
|
+
def do_download(dest_file_or_io, metis_file, &block)
|
17
17
|
size = metis_file.size
|
18
18
|
completed = 0.0
|
19
19
|
start = Time.now
|
20
20
|
|
21
|
-
|
22
|
-
|
23
|
-
|
24
|
-
completed += chunk.size
|
25
|
-
|
26
|
-
block.call([
|
27
|
-
:progress,
|
28
|
-
size == 0 ? 1 : completed / size,
|
29
|
-
(completed / (Time.now - start)).round(2),
|
30
|
-
]) unless block.nil?
|
21
|
+
unless dest_file_or_io.is_a?(IO)
|
22
|
+
::File.open(dest_file_or_io, 'w') do |io|
|
23
|
+
return do_download(dest_file_or_io, metis_file, &block)
|
31
24
|
end
|
32
25
|
end
|
26
|
+
|
27
|
+
metis_client.download_file(metis_file) do |chunk|
|
28
|
+
dest_file_or_io.write chunk
|
29
|
+
completed += chunk.size
|
30
|
+
|
31
|
+
block.call([
|
32
|
+
:progress,
|
33
|
+
size == 0 ? 1 : completed / size,
|
34
|
+
(completed / (Time.now - start)).round(2),
|
35
|
+
]) unless block.nil?
|
36
|
+
end
|
33
37
|
end
|
34
38
|
end
|
35
39
|
end
|
@@ -4,17 +4,26 @@ require 'fileutils'
|
|
4
4
|
require 'tempfile'
|
5
5
|
require 'securerandom'
|
6
6
|
|
7
|
+
$digest_mutx = Mutex.new
|
8
|
+
|
7
9
|
module Etna
|
8
10
|
module Clients
|
9
11
|
class Metis
|
10
12
|
class MetisUploadWorkflow < Struct.new(:metis_client, :metis_uid, :project_name, :bucket_name, :max_attempts, keyword_init: true)
|
13
|
+
class StreamingUploadError < StandardError
|
14
|
+
end
|
15
|
+
|
11
16
|
|
12
17
|
def initialize(args)
|
13
18
|
super({max_attempts: 3, metis_uid: SecureRandom.hex}.update(args))
|
14
19
|
end
|
15
20
|
|
16
|
-
def do_upload(
|
17
|
-
|
21
|
+
def do_upload(source_file_or_upload, dest_path, &block)
|
22
|
+
unless source_file_or_upload.is_a?(Upload)
|
23
|
+
upload = Upload.new(source_file: source_file_or_upload)
|
24
|
+
else
|
25
|
+
upload = source_file_or_upload
|
26
|
+
end
|
18
27
|
|
19
28
|
dir = ::File.dirname(dest_path)
|
20
29
|
metis_client.create_folder(CreateFolderRequest.new(
|
@@ -68,9 +77,11 @@ module Etna
|
|
68
77
|
|
69
78
|
unsent_zero_byte_file = false
|
70
79
|
rescue Etna::Error => e
|
71
|
-
|
72
|
-
|
73
|
-
|
80
|
+
unless block.nil?
|
81
|
+
m = yield [:error, e]
|
82
|
+
if m == false
|
83
|
+
raise e
|
84
|
+
end
|
74
85
|
end
|
75
86
|
|
76
87
|
if e.status == 422
|
@@ -90,13 +101,17 @@ module Etna
|
|
90
101
|
end
|
91
102
|
end
|
92
103
|
|
93
|
-
class Upload
|
104
|
+
class Upload
|
94
105
|
INITIAL_BLOB_SIZE = 2 ** 10
|
95
106
|
MAX_BLOB_SIZE = 2 ** 22
|
96
107
|
ZERO_HASH = 'd41d8cd98f00b204e9800998ecf8427e'
|
97
108
|
|
98
|
-
|
99
|
-
|
109
|
+
attr_accessor :source_file, :next_blob_size, :current_byte_position
|
110
|
+
|
111
|
+
def initialize(source_file: nil, next_blob_size: nil, current_byte_position: nil)
|
112
|
+
self.source_file = source_file
|
113
|
+
self.next_blob_size = next_blob_size
|
114
|
+
self.current_byte_position = current_byte_position
|
100
115
|
self.next_blob_size = [file_size, INITIAL_BLOB_SIZE].min
|
101
116
|
self.current_byte_position = 0
|
102
117
|
end
|
@@ -108,10 +123,10 @@ module Etna
|
|
108
123
|
def advance_position!
|
109
124
|
self.current_byte_position = self.current_byte_position + self.next_blob_size
|
110
125
|
self.next_blob_size = [
|
111
|
-
|
112
|
-
|
113
|
-
|
114
|
-
|
126
|
+
MAX_BLOB_SIZE,
|
127
|
+
# in fact we should stop when we hit the end of the file
|
128
|
+
file_size - current_byte_position
|
129
|
+
].min
|
115
130
|
end
|
116
131
|
|
117
132
|
def complete?
|
@@ -119,7 +134,14 @@ module Etna
|
|
119
134
|
end
|
120
135
|
|
121
136
|
def next_blob_hash
|
122
|
-
|
137
|
+
bytes = next_blob_bytes
|
138
|
+
if bytes.empty?
|
139
|
+
return ZERO_HASH
|
140
|
+
end
|
141
|
+
|
142
|
+
$digest_mutx.synchronize do
|
143
|
+
return Digest::MD5.hexdigest(bytes)
|
144
|
+
end
|
123
145
|
end
|
124
146
|
|
125
147
|
def next_blob_bytes
|
@@ -131,6 +153,54 @@ module Etna
|
|
131
153
|
self.next_blob_size = upload_response.next_blob_size
|
132
154
|
end
|
133
155
|
end
|
156
|
+
|
157
|
+
class StreamingIOUpload < Upload
|
158
|
+
def initialize(readable_io:, size_hint: 0, **args)
|
159
|
+
@readable_io = readable_io
|
160
|
+
@size_hint = size_hint
|
161
|
+
@read_position = 0
|
162
|
+
@last_bytes = ""
|
163
|
+
super(**args)
|
164
|
+
end
|
165
|
+
|
166
|
+
def file_size
|
167
|
+
@size_hint
|
168
|
+
end
|
169
|
+
|
170
|
+
def next_blob_bytes
|
171
|
+
next_left = current_byte_position
|
172
|
+
next_right = current_byte_position + next_blob_size
|
173
|
+
|
174
|
+
if next_right < @read_position
|
175
|
+
raise StreamingUploadError.new("Upload needs restart, but source is streaming and ephemeral. #{next_right} #{@read_position} You need to restart the source stream and create a new upload.")
|
176
|
+
elsif @read_position < next_left
|
177
|
+
# read from the stream and discard until we are positioned for the next read.
|
178
|
+
data = @readable_io.read(next_left - @read_position)
|
179
|
+
raise StreamingUploadError.new("Unexpected EOF in read stream") if data.nil?
|
180
|
+
|
181
|
+
@read_position += data.bytes.length
|
182
|
+
end
|
183
|
+
|
184
|
+
# If we have consumed all requested data, return what we have consumed.
|
185
|
+
# If we have requested no data, make sure to provide "" as the result.
|
186
|
+
if next_right == @read_position
|
187
|
+
return @last_bytes
|
188
|
+
end
|
189
|
+
|
190
|
+
if @read_position != next_left
|
191
|
+
raise StreamingUploadError.new("Alignment error, source data does not match expected upload resume. #{@read_position} #{next_left} Restart the upload to address.")
|
192
|
+
end
|
193
|
+
|
194
|
+
@last_bytes = "".tap do |bytes|
|
195
|
+
while @read_position < next_right
|
196
|
+
bytes << @readable_io.read(next_right - @read_position).tap do |data|
|
197
|
+
raise StreamingUploadError.new("Unexpected EOF in read stream") if data.nil?
|
198
|
+
@read_position += data.bytes.length
|
199
|
+
end
|
200
|
+
end
|
201
|
+
end
|
202
|
+
end
|
203
|
+
end
|
134
204
|
end
|
135
205
|
end
|
136
206
|
end
|
@@ -1,107 +1,71 @@
|
|
1
1
|
require 'ostruct'
|
2
|
-
require 'digest'
|
3
2
|
require 'fileutils'
|
4
3
|
require 'tempfile'
|
5
4
|
|
6
5
|
module Etna
|
7
6
|
module Clients
|
8
7
|
class Metis
|
9
|
-
class SyncMetisDataWorkflow < Struct.new(:metis_client, :filesystem, :project_name, :bucket_name,
|
10
|
-
|
11
|
-
|
12
|
-
own_tmpdir = tmpdir.nil? && !skip_tmpdir
|
13
|
-
if own_tmpdir
|
14
|
-
tmpdir = filesystem.tmpdir
|
15
|
-
end
|
16
|
-
|
17
|
-
begin
|
18
|
-
response = metis_client.list_folder(ListFolderRequest.new(project_name: project_name, bucket_name: bucket_name, folder_path: src))
|
8
|
+
class SyncMetisDataWorkflow < Struct.new(:metis_client, :filesystem, :project_name, :bucket_name, :logger, keyword_init: true)
|
9
|
+
def copy_directory(src, dest, root = dest)
|
10
|
+
response = metis_client.list_folder(ListFolderRequest.new(project_name: project_name, bucket_name: bucket_name, folder_path: src))
|
19
11
|
|
20
|
-
|
21
|
-
|
22
|
-
|
23
|
-
end
|
24
|
-
|
25
|
-
response.folders.all.each do |folder|
|
26
|
-
copy_directory(::File.join(src, folder.folder_name), ::File.join(dest, folder.folder_name), root, tmpdir)
|
27
|
-
end
|
28
|
-
ensure
|
29
|
-
filesystem.rm_rf(tmpdir) if own_tmpdir
|
12
|
+
response.files.all.each do |file|
|
13
|
+
logger&.info("Copying file #{file.file_path} (#{Etna::Formatting.as_size(file.size)})")
|
14
|
+
copy_file(dest: ::File.join(dest, file.file_name), url: file.download_url)
|
30
15
|
end
|
31
|
-
end
|
32
|
-
|
33
|
-
def bin_file_name(etag:)
|
34
|
-
"bin/#{etag}"
|
35
|
-
end
|
36
16
|
|
37
|
-
|
38
|
-
|
39
|
-
if tmpdir.nil?
|
40
|
-
yield dest_file_name
|
41
|
-
else
|
42
|
-
tmp_file = ::File.join(tmpdir, ::File.basename(bin_file_name))
|
43
|
-
yield tmp_file
|
44
|
-
filesystem.mv(tmp_file, dest_file_name)
|
17
|
+
response.folders.all.each do |folder|
|
18
|
+
copy_directory(::File.join(src, folder.folder_name), ::File.join(dest, folder.folder_name), root)
|
45
19
|
end
|
46
20
|
end
|
47
21
|
|
48
|
-
def copy_file(
|
22
|
+
def copy_file(dest:, url:, stub: false)
|
49
23
|
metadata = metis_client.file_metadata(url)
|
50
|
-
etag = metadata[:etag]
|
51
24
|
size = metadata[:size]
|
52
25
|
|
53
|
-
|
54
|
-
|
55
|
-
|
56
|
-
|
57
|
-
|
58
|
-
|
59
|
-
|
60
|
-
|
61
|
-
|
62
|
-
|
63
|
-
|
64
|
-
|
65
|
-
|
66
|
-
|
67
|
-
|
68
|
-
|
69
|
-
|
70
|
-
|
71
|
-
|
72
|
-
upload_amount
|
73
|
-
|
74
|
-
if upload_timings.length > 150
|
75
|
-
s, _ = upload_timings.shift
|
76
|
-
upload_amount -= s
|
77
|
-
end
|
26
|
+
tmp_file = dest
|
27
|
+
upload_timings = []
|
28
|
+
upload_amount = 0
|
29
|
+
last_rate = 0.00001
|
30
|
+
remaining = size
|
31
|
+
|
32
|
+
filesystem.with_writeable(tmp_file, "w", size_hint: size) do |io|
|
33
|
+
if stub
|
34
|
+
io.write("(stub) #{size} bytes")
|
35
|
+
else
|
36
|
+
metis_client.download_file(url) do |chunk|
|
37
|
+
io.write(chunk)
|
38
|
+
|
39
|
+
upload_timings << [chunk.length, Time.now.to_f]
|
40
|
+
upload_amount += chunk.length
|
41
|
+
remaining -= chunk.length
|
42
|
+
|
43
|
+
if upload_timings.length > 150
|
44
|
+
s, _ = upload_timings.shift
|
45
|
+
upload_amount -= s
|
46
|
+
end
|
78
47
|
|
79
|
-
|
80
|
-
|
48
|
+
_, start_time = upload_timings.first
|
49
|
+
_, end_time = upload_timings.last
|
81
50
|
|
82
|
-
|
83
|
-
|
84
|
-
|
51
|
+
if start_time == end_time
|
52
|
+
next
|
53
|
+
end
|
85
54
|
|
86
|
-
|
55
|
+
rate = upload_amount / (end_time - start_time)
|
87
56
|
|
88
|
-
|
89
|
-
|
57
|
+
if rate / last_rate > 1.3 || rate / last_rate < 0.7
|
58
|
+
logger&.info("Uploading #{Etna::Formatting.as_size(rate)} per second, #{Etna::Formatting.as_size(remaining)} remaining")
|
90
59
|
|
91
|
-
|
92
|
-
|
93
|
-
|
94
|
-
|
95
|
-
end
|
60
|
+
if rate == 0
|
61
|
+
last_rate = 0.0001
|
62
|
+
else
|
63
|
+
last_rate = rate
|
96
64
|
end
|
97
65
|
end
|
98
66
|
end
|
99
|
-
end
|
100
|
-
end
|
101
67
|
|
102
|
-
|
103
|
-
filesystem.with_writeable(dest_bin_file, 'w', size_hint: 0) do |io|
|
104
|
-
# empty file marking that this etag has been moved, to save a future write.
|
68
|
+
end
|
105
69
|
end
|
106
70
|
end
|
107
71
|
end
|
data/lib/etna/cwl.rb
CHANGED
data/lib/etna/directed_graph.rb
CHANGED
@@ -23,7 +23,7 @@ class DirectedGraph
|
|
23
23
|
parents_of_map = descendants(root)
|
24
24
|
seen = Set.new
|
25
25
|
|
26
|
-
parents_of_map.to_a.sort_by { |k, parents| [-parents.length, k] }.each do |k, parents|
|
26
|
+
parents_of_map.to_a.sort_by { |k, parents| [-parents.length, k.inspect] }.each do |k, parents|
|
27
27
|
unless seen.include?(k)
|
28
28
|
if @children[k].keys.empty?
|
29
29
|
result << parents + [k]
|
data/lib/etna/filesystem.rb
CHANGED
@@ -1,36 +1,52 @@
|
|
1
1
|
require 'yaml'
|
2
2
|
require 'fileutils'
|
3
3
|
require 'open3'
|
4
|
+
require 'securerandom'
|
5
|
+
require 'concurrent-ruby'
|
4
6
|
|
5
7
|
module Etna
|
6
8
|
# A class that encapsulates opening / reading file system entries that abstracts normal file access in order
|
7
9
|
# to make stubbing, substituting, and testing easier.
|
8
10
|
class Filesystem
|
9
11
|
def with_writeable(dest, opts = 'w', size_hint: nil, &block)
|
12
|
+
raise "with_writeable not supported by #{self.class.name}" unless self.class == Filesystem
|
10
13
|
::File.open(dest, opts, &block)
|
11
14
|
end
|
12
15
|
|
16
|
+
def ls(dir)
|
17
|
+
raise "ls not supported by #{self.class.name}" unless self.class == Filesystem
|
18
|
+
::Dir.entries(dir).select { |p| !p.start_with?('.') }.map do |path|
|
19
|
+
::File.file?(::File.join(dir, path)) ? [:file, path] : [:dir, path]
|
20
|
+
end
|
21
|
+
end
|
22
|
+
|
13
23
|
def with_readable(src, opts = 'r', &block)
|
24
|
+
raise "with_readable not supported by #{self.class.name}" unless self.class == Filesystem
|
14
25
|
::File.open(src, opts, &block)
|
15
26
|
end
|
16
27
|
|
17
28
|
def mkdir_p(dir)
|
29
|
+
raise "mkdir_p not supported by #{self.class.name}" unless self.class == Filesystem
|
18
30
|
::FileUtils.mkdir_p(dir)
|
19
31
|
end
|
20
32
|
|
21
33
|
def rm_rf(dir)
|
34
|
+
raise "rm_rf not supported by #{self.class.name}" unless self.class == Filesystem
|
22
35
|
::FileUtils.rm_rf(dir)
|
23
36
|
end
|
24
37
|
|
25
38
|
def tmpdir
|
39
|
+
raise "tmpdir not supported by #{self.class.name}" unless self.class == Filesystem
|
26
40
|
::Dir.mktmpdir
|
27
41
|
end
|
28
42
|
|
29
43
|
def exist?(src)
|
44
|
+
raise "exist? not supported by #{self.class.name}" unless self.class == Filesystem
|
30
45
|
::File.exist?(src)
|
31
46
|
end
|
32
47
|
|
33
48
|
def mv(src, dest)
|
49
|
+
raise "mv not supported by #{self.class.name}" unless self.class == Filesystem
|
34
50
|
::FileUtils.mv(src, dest)
|
35
51
|
end
|
36
52
|
|
@@ -179,7 +195,7 @@ module Etna
|
|
179
195
|
cmd << remote_path
|
180
196
|
cmd << local_path
|
181
197
|
|
182
|
-
cmd << {
|
198
|
+
cmd << {out: wd}
|
183
199
|
elsif opts.include?('w')
|
184
200
|
cmd << '--mode=send'
|
185
201
|
cmd << "--host=#{@host}"
|
@@ -187,7 +203,7 @@ module Etna
|
|
187
203
|
cmd << local_path
|
188
204
|
cmd << remote_path
|
189
205
|
|
190
|
-
cmd << {
|
206
|
+
cmd << {in: rd}
|
191
207
|
end
|
192
208
|
|
193
209
|
cmd
|
@@ -196,23 +212,10 @@ module Etna
|
|
196
212
|
|
197
213
|
# Genentech's aspera deployment doesn't support modern commands, unfortunately...
|
198
214
|
class GeneAsperaCliFilesystem < AsperaCliFilesystem
|
199
|
-
def tmpdir
|
200
|
-
raise "tmpdir is not supported"
|
201
|
-
end
|
202
|
-
|
203
|
-
def rm_rf
|
204
|
-
raise "rm_rf is not supported"
|
205
|
-
end
|
206
|
-
|
207
215
|
def mkdir_p(dest)
|
208
216
|
# Pass through -- this file system creates containing directories by default, womp womp.
|
209
217
|
end
|
210
218
|
|
211
|
-
def mv
|
212
|
-
raise "mv is not supported"
|
213
|
-
end
|
214
|
-
|
215
|
-
|
216
219
|
def mkcommand(rd, wd, file, opts, size_hint: nil)
|
217
220
|
if opts.include?('w')
|
218
221
|
super.map do |e|
|
@@ -241,6 +244,131 @@ module Etna
|
|
241
244
|
end
|
242
245
|
end
|
243
246
|
|
247
|
+
class Metis < Filesystem
|
248
|
+
def initialize(metis_client:, project_name:, bucket_name:, root: '/', uuid: SecureRandom.uuid)
|
249
|
+
@metis_client = metis_client
|
250
|
+
@project_name = project_name
|
251
|
+
@bucket_name = bucket_name
|
252
|
+
@root = root
|
253
|
+
@metis_uid = uuid
|
254
|
+
end
|
255
|
+
|
256
|
+
def metis_path_of(path)
|
257
|
+
joined = ::File.join(@root, path)
|
258
|
+
joined[0] == "/" ? joined.slice(1..-1) : joined
|
259
|
+
end
|
260
|
+
|
261
|
+
def create_upload_workflow
|
262
|
+
Etna::Clients::Metis::MetisUploadWorkflow.new(metis_client: @metis_client, metis_uid: @metis_uid, project_name: @project_name, bucket_name: @bucket_name, max_attempts: 3)
|
263
|
+
end
|
264
|
+
|
265
|
+
def with_hot_pipe(opts, receiver, *args, &block)
|
266
|
+
rp, wp = IO.pipe
|
267
|
+
begin
|
268
|
+
executor = Concurrent::SingleThreadExecutor.new(fallback_policy: :abort)
|
269
|
+
begin
|
270
|
+
if opts.include?('w')
|
271
|
+
future = Concurrent::Promises.future_on(executor) do
|
272
|
+
self.send(receiver, rp, *args)
|
273
|
+
rescue => e
|
274
|
+
Etna::Application.instance.logger.log_error(e)
|
275
|
+
raise e
|
276
|
+
ensure
|
277
|
+
rp.close
|
278
|
+
end
|
279
|
+
|
280
|
+
yield wp
|
281
|
+
else
|
282
|
+
future = Concurrent::Promises.future_on(executor) do
|
283
|
+
self.send(receiver, wp, *args)
|
284
|
+
rescue => e
|
285
|
+
Etna::Application.instance.logger.log_error(e)
|
286
|
+
raise e
|
287
|
+
ensure
|
288
|
+
wp.close
|
289
|
+
end
|
290
|
+
|
291
|
+
yield rp
|
292
|
+
end
|
293
|
+
|
294
|
+
future.wait!
|
295
|
+
ensure
|
296
|
+
executor.shutdown
|
297
|
+
executor.kill unless executor.wait_for_termination(5)
|
298
|
+
end
|
299
|
+
ensure
|
300
|
+
rp.close
|
301
|
+
wp.close
|
302
|
+
end
|
303
|
+
end
|
304
|
+
|
305
|
+
def do_streaming_upload(rp, dest, size_hint)
|
306
|
+
streaming_upload = Etna::Clients::Metis::MetisUploadWorkflow::StreamingIOUpload.new(readable_io: rp, size_hint: size_hint)
|
307
|
+
create_upload_workflow.do_upload(
|
308
|
+
streaming_upload,
|
309
|
+
metis_path_of(dest)
|
310
|
+
)
|
311
|
+
end
|
312
|
+
|
313
|
+
def with_writeable(dest, opts = 'w', size_hint: nil, &block)
|
314
|
+
self.with_hot_pipe(opts, :do_streaming_upload, dest, size_hint) do |wp|
|
315
|
+
yield wp
|
316
|
+
end
|
317
|
+
end
|
318
|
+
|
319
|
+
def create_download_workflow
|
320
|
+
Etna::Clients::Metis::MetisDownloadWorkflow.new(metis_client: @metis_client, project_name: @project_name, bucket_name: @bucket_name, max_attempts: 3)
|
321
|
+
end
|
322
|
+
|
323
|
+
def do_streaming_download(wp, metis_file)
|
324
|
+
create_download_workflow.do_download(wp, metis_file)
|
325
|
+
end
|
326
|
+
|
327
|
+
def with_readable(src, opts = 'r', &block)
|
328
|
+
metis_file = list_metis_directory(::File.dirname(src)).files.all.find { |f| f.file_name == ::File.basename(src) }
|
329
|
+
raise "Metis file at #{@project_name}/#{@bucket_name}/#{@root}/#{src} not found. No such file" if metis_file.nil?
|
330
|
+
|
331
|
+
self.with_hot_pipe(opts, :do_streaming_download, metis_file) do |rp|
|
332
|
+
yield rp
|
333
|
+
end
|
334
|
+
end
|
335
|
+
|
336
|
+
def list_metis_directory(path)
|
337
|
+
@metis_client.list_folder(Etna::Clients::Metis::ListFolderRequest.new(project_name: @project_name, bucket_name: @bucket_name, folder_path: metis_path_of(path)))
|
338
|
+
end
|
339
|
+
|
340
|
+
def mkdir_p(dir)
|
341
|
+
create_folder_request = Etna::Clients::Metis::CreateFolderRequest.new(
|
342
|
+
project_name: @project_name,
|
343
|
+
bucket_name: @bucket_name,
|
344
|
+
folder_path: metis_path_of(dir),
|
345
|
+
)
|
346
|
+
@metis_client.create_folder(create_folder_request)
|
347
|
+
end
|
348
|
+
|
349
|
+
def ls(dir)
|
350
|
+
response = list_metis_directory(::File.dirname(dir))
|
351
|
+
response.files.map { |f| [:file, f.file_name] } + response.folders.map { |f| [:folder, f.folder_name] }
|
352
|
+
end
|
353
|
+
|
354
|
+
def exist?(src)
|
355
|
+
begin
|
356
|
+
response = list_metis_directory(::File.dirname(src))
|
357
|
+
rescue Etna::Error => e
|
358
|
+
if e.status == 404
|
359
|
+
return false
|
360
|
+
elsif e.message =~ /Invalid folder/
|
361
|
+
return false
|
362
|
+
end
|
363
|
+
|
364
|
+
raise e
|
365
|
+
end
|
366
|
+
|
367
|
+
response.files.all.any? { |f| f.file_name == ::File.basename(src) } ||
|
368
|
+
response.folders.all.any? { |f| f.folder_name == ::File.basename(src) }
|
369
|
+
end
|
370
|
+
end
|
371
|
+
|
244
372
|
class Mock < Filesystem
|
245
373
|
def initialize(&new_io)
|
246
374
|
@files = {}
|
data/lib/etna/spec/auth.rb
CHANGED
@@ -2,22 +2,22 @@ module Etna::Spec
|
|
2
2
|
module Auth
|
3
3
|
AUTH_USERS = {
|
4
4
|
superuser: {
|
5
|
-
email: 'zeus@olympus.org',
|
5
|
+
email: 'zeus@olympus.org', name: 'Zeus', perm: 'A:administration'
|
6
6
|
},
|
7
7
|
admin: {
|
8
|
-
email: 'hera@olympus.org',
|
8
|
+
email: 'hera@olympus.org', name: 'Hera', perm: 'a:labors'
|
9
9
|
},
|
10
10
|
editor: {
|
11
|
-
email: 'eurystheus@twelve-labors.org',
|
11
|
+
email: 'eurystheus@twelve-labors.org', name: 'Eurystheus', perm: 'E:labors'
|
12
12
|
},
|
13
13
|
restricted_editor: {
|
14
|
-
email: 'copreus@twelve-labors.org',
|
14
|
+
email: 'copreus@twelve-labors.org', name: 'Copreus', perm: 'e:labors'
|
15
15
|
},
|
16
16
|
viewer: {
|
17
|
-
email: 'hercules@twelve-labors.org',
|
17
|
+
email: 'hercules@twelve-labors.org', name: 'Hercules', perm: 'v:labors'
|
18
18
|
},
|
19
19
|
non_user: {
|
20
|
-
email: 'nessus@centaurs.org',
|
20
|
+
email: 'nessus@centaurs.org', name: 'Nessus', perm: ''
|
21
21
|
}
|
22
22
|
}
|
23
23
|
|
data/lib/etna/user.rb
CHANGED
@@ -7,14 +7,14 @@ module Etna
|
|
7
7
|
}
|
8
8
|
|
9
9
|
def initialize params, token=nil
|
10
|
-
@
|
10
|
+
@name, @email, @encoded_permissions, encoded_flags = params.values_at(:name, :email, :perm, :flags)
|
11
11
|
|
12
12
|
@flags = encoded_flags&.split(/;/) || []
|
13
13
|
@token = token unless !token
|
14
14
|
raise ArgumentError, "No email given!" unless @email
|
15
15
|
end
|
16
16
|
|
17
|
-
attr_reader :
|
17
|
+
attr_reader :name, :email, :token
|
18
18
|
|
19
19
|
def permissions
|
20
20
|
@permissions ||= @encoded_permissions.split(/\;/).map do |roles|
|
@@ -36,10 +36,6 @@ module Etna
|
|
36
36
|
@flags.include?(flag)
|
37
37
|
end
|
38
38
|
|
39
|
-
def name
|
40
|
-
"#{first} #{last}"
|
41
|
-
end
|
42
|
-
|
43
39
|
def projects
|
44
40
|
permissions.keys
|
45
41
|
end
|
metadata
CHANGED
@@ -1,14 +1,14 @@
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
2
2
|
name: etna
|
3
3
|
version: !ruby/object:Gem::Version
|
4
|
-
version: 0.1.
|
4
|
+
version: 0.1.29
|
5
5
|
platform: ruby
|
6
6
|
authors:
|
7
7
|
- Saurabh Asthana
|
8
8
|
autorequire:
|
9
9
|
bindir: bin
|
10
10
|
cert_chain: []
|
11
|
-
date: 2021-02-
|
11
|
+
date: 2021-02-23 00:00:00.000000000 Z
|
12
12
|
dependencies:
|
13
13
|
- !ruby/object:Gem::Dependency
|
14
14
|
name: rack
|
@@ -80,6 +80,34 @@ dependencies:
|
|
80
80
|
- - ">="
|
81
81
|
- !ruby/object:Gem::Version
|
82
82
|
version: '0'
|
83
|
+
- !ruby/object:Gem::Dependency
|
84
|
+
name: concurrent-ruby
|
85
|
+
requirement: !ruby/object:Gem::Requirement
|
86
|
+
requirements:
|
87
|
+
- - ">="
|
88
|
+
- !ruby/object:Gem::Version
|
89
|
+
version: '0'
|
90
|
+
type: :runtime
|
91
|
+
prerelease: false
|
92
|
+
version_requirements: !ruby/object:Gem::Requirement
|
93
|
+
requirements:
|
94
|
+
- - ">="
|
95
|
+
- !ruby/object:Gem::Version
|
96
|
+
version: '0'
|
97
|
+
- !ruby/object:Gem::Dependency
|
98
|
+
name: concurrent-ruby-ext
|
99
|
+
requirement: !ruby/object:Gem::Requirement
|
100
|
+
requirements:
|
101
|
+
- - ">="
|
102
|
+
- !ruby/object:Gem::Version
|
103
|
+
version: '0'
|
104
|
+
type: :runtime
|
105
|
+
prerelease: false
|
106
|
+
version_requirements: !ruby/object:Gem::Requirement
|
107
|
+
requirements:
|
108
|
+
- - ">="
|
109
|
+
- !ruby/object:Gem::Version
|
110
|
+
version: '0'
|
83
111
|
description: See summary
|
84
112
|
email: Saurabh.Asthana@ucsf.edu
|
85
113
|
executables:
|