Files
awx/awx/lib/site-packages/azure/http/batchclient.py
2015-01-29 12:13:53 -05:00

340 lines
13 KiB
Python

#-------------------------------------------------------------------------
# Copyright (c) Microsoft. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
#--------------------------------------------------------------------------
import sys
import uuid
from azure import (
_update_request_uri_query,
WindowsAzureError,
WindowsAzureBatchOperationError,
_get_children_from_path,
url_unquote,
_ERROR_CANNOT_FIND_PARTITION_KEY,
_ERROR_CANNOT_FIND_ROW_KEY,
_ERROR_INCORRECT_TABLE_IN_BATCH,
_ERROR_INCORRECT_PARTITION_KEY_IN_BATCH,
_ERROR_DUPLICATE_ROW_KEY_IN_BATCH,
_ERROR_BATCH_COMMIT_FAIL,
)
from azure.http import HTTPError, HTTPRequest, HTTPResponse
from azure.http.httpclient import _HTTPClient
from azure.storage import (
_update_storage_table_header,
METADATA_NS,
_sign_storage_table_request,
)
from xml.dom import minidom
_DATASERVICES_NS = 'http://schemas.microsoft.com/ado/2007/08/dataservices'
if sys.version_info < (3,):
def _new_boundary():
return str(uuid.uuid1())
else:
def _new_boundary():
return str(uuid.uuid1()).encode('utf-8')
class _BatchClient(_HTTPClient):
'''
This is the class that is used for batch operation for storage table
service. It only supports one changeset.
'''
def __init__(self, service_instance, account_key, account_name,
protocol='http'):
_HTTPClient.__init__(self, service_instance, account_name=account_name,
account_key=account_key, protocol=protocol)
self.is_batch = False
self.batch_requests = []
self.batch_table = ''
self.batch_partition_key = ''
self.batch_row_keys = []
def get_request_table(self, request):
'''
Extracts table name from request.uri. The request.uri has either
"/mytable(...)" or "/mytable" format.
request: the request to insert, update or delete entity
'''
if '(' in request.path:
pos = request.path.find('(')
return request.path[1:pos]
else:
return request.path[1:]
def get_request_partition_key(self, request):
'''
Extracts PartitionKey from request.body if it is a POST request or from
request.path if it is not a POST request. Only insert operation request
is a POST request and the PartitionKey is in the request body.
request: the request to insert, update or delete entity
'''
if request.method == 'POST':
doc = minidom.parseString(request.body)
part_key = _get_children_from_path(
doc, 'entry', 'content', (METADATA_NS, 'properties'),
(_DATASERVICES_NS, 'PartitionKey'))
if not part_key:
raise WindowsAzureError(_ERROR_CANNOT_FIND_PARTITION_KEY)
return part_key[0].firstChild.nodeValue
else:
uri = url_unquote(request.path)
pos1 = uri.find('PartitionKey=\'')
pos2 = uri.find('\',', pos1)
if pos1 == -1 or pos2 == -1:
raise WindowsAzureError(_ERROR_CANNOT_FIND_PARTITION_KEY)
return uri[pos1 + len('PartitionKey=\''):pos2]
def get_request_row_key(self, request):
'''
Extracts RowKey from request.body if it is a POST request or from
request.path if it is not a POST request. Only insert operation request
is a POST request and the Rowkey is in the request body.
request: the request to insert, update or delete entity
'''
if request.method == 'POST':
doc = minidom.parseString(request.body)
row_key = _get_children_from_path(
doc, 'entry', 'content', (METADATA_NS, 'properties'),
(_DATASERVICES_NS, 'RowKey'))
if not row_key:
raise WindowsAzureError(_ERROR_CANNOT_FIND_ROW_KEY)
return row_key[0].firstChild.nodeValue
else:
uri = url_unquote(request.path)
pos1 = uri.find('RowKey=\'')
pos2 = uri.find('\')', pos1)
if pos1 == -1 or pos2 == -1:
raise WindowsAzureError(_ERROR_CANNOT_FIND_ROW_KEY)
row_key = uri[pos1 + len('RowKey=\''):pos2]
return row_key
def validate_request_table(self, request):
'''
Validates that all requests have the same table name. Set the table
name if it is the first request for the batch operation.
request: the request to insert, update or delete entity
'''
if self.batch_table:
if self.get_request_table(request) != self.batch_table:
raise WindowsAzureError(_ERROR_INCORRECT_TABLE_IN_BATCH)
else:
self.batch_table = self.get_request_table(request)
def validate_request_partition_key(self, request):
'''
Validates that all requests have the same PartitiionKey. Set the
PartitionKey if it is the first request for the batch operation.
request: the request to insert, update or delete entity
'''
if self.batch_partition_key:
if self.get_request_partition_key(request) != \
self.batch_partition_key:
raise WindowsAzureError(_ERROR_INCORRECT_PARTITION_KEY_IN_BATCH)
else:
self.batch_partition_key = self.get_request_partition_key(request)
def validate_request_row_key(self, request):
'''
Validates that all requests have the different RowKey and adds RowKey
to existing RowKey list.
request: the request to insert, update or delete entity
'''
if self.batch_row_keys:
if self.get_request_row_key(request) in self.batch_row_keys:
raise WindowsAzureError(_ERROR_DUPLICATE_ROW_KEY_IN_BATCH)
else:
self.batch_row_keys.append(self.get_request_row_key(request))
def begin_batch(self):
'''
Starts the batch operation. Intializes the batch variables
is_batch: batch operation flag.
batch_table: the table name of the batch operation
batch_partition_key: the PartitionKey of the batch requests.
batch_row_keys: the RowKey list of adding requests.
batch_requests: the list of the requests.
'''
self.is_batch = True
self.batch_table = ''
self.batch_partition_key = ''
self.batch_row_keys = []
self.batch_requests = []
def insert_request_to_batch(self, request):
'''
Adds request to batch operation.
request: the request to insert, update or delete entity
'''
self.validate_request_table(request)
self.validate_request_partition_key(request)
self.validate_request_row_key(request)
self.batch_requests.append(request)
def commit_batch(self):
''' Resets batch flag and commits the batch requests. '''
if self.is_batch:
self.is_batch = False
self.commit_batch_requests()
def commit_batch_requests(self):
''' Commits the batch requests. '''
batch_boundary = b'batch_' + _new_boundary()
changeset_boundary = b'changeset_' + _new_boundary()
# Commits batch only the requests list is not empty.
if self.batch_requests:
request = HTTPRequest()
request.method = 'POST'
request.host = self.batch_requests[0].host
request.path = '/$batch'
request.headers = [
('Content-Type', 'multipart/mixed; boundary=' + \
batch_boundary.decode('utf-8')),
('Accept', 'application/atom+xml,application/xml'),
('Accept-Charset', 'UTF-8')]
request.body = b'--' + batch_boundary + b'\n'
request.body += b'Content-Type: multipart/mixed; boundary='
request.body += changeset_boundary + b'\n\n'
content_id = 1
# Adds each request body to the POST data.
for batch_request in self.batch_requests:
request.body += b'--' + changeset_boundary + b'\n'
request.body += b'Content-Type: application/http\n'
request.body += b'Content-Transfer-Encoding: binary\n\n'
request.body += batch_request.method.encode('utf-8')
request.body += b' http://'
request.body += batch_request.host.encode('utf-8')
request.body += batch_request.path.encode('utf-8')
request.body += b' HTTP/1.1\n'
request.body += b'Content-ID: '
request.body += str(content_id).encode('utf-8') + b'\n'
content_id += 1
# Add different headers for different type requests.
if not batch_request.method == 'DELETE':
request.body += \
b'Content-Type: application/atom+xml;type=entry\n'
for name, value in batch_request.headers:
if name == 'If-Match':
request.body += name.encode('utf-8') + b': '
request.body += value.encode('utf-8') + b'\n'
break
request.body += b'Content-Length: '
request.body += str(len(batch_request.body)).encode('utf-8')
request.body += b'\n\n'
request.body += batch_request.body + b'\n'
else:
for name, value in batch_request.headers:
# If-Match should be already included in
# batch_request.headers, but in case it is missing,
# just add it.
if name == 'If-Match':
request.body += name.encode('utf-8') + b': '
request.body += value.encode('utf-8') + b'\n\n'
break
else:
request.body += b'If-Match: *\n\n'
request.body += b'--' + changeset_boundary + b'--' + b'\n'
request.body += b'--' + batch_boundary + b'--'
request.path, request.query = _update_request_uri_query(request)
request.headers = _update_storage_table_header(request)
auth = _sign_storage_table_request(request,
self.account_name,
self.account_key)
request.headers.append(('Authorization', auth))
# Submit the whole request as batch request.
response = self.perform_request(request)
if response.status >= 300:
raise HTTPError(response.status,
_ERROR_BATCH_COMMIT_FAIL,
self.respheader,
response.body)
# http://www.odata.org/documentation/odata-version-2-0/batch-processing/
# The body of a ChangeSet response is either a response for all the
# successfully processed change request within the ChangeSet,
# formatted exactly as it would have appeared outside of a batch,
# or a single response indicating a failure of the entire ChangeSet.
responses = self._parse_batch_response(response.body)
if responses and responses[0].status >= 300:
self._report_batch_error(responses[0])
def cancel_batch(self):
''' Resets the batch flag. '''
self.is_batch = False
def _parse_batch_response(self, body):
parts = body.split(b'--changesetresponse_')
responses = []
for part in parts:
httpLocation = part.find(b'HTTP/')
if httpLocation > 0:
response = self._parse_batch_response_part(part[httpLocation:])
responses.append(response)
return responses
def _parse_batch_response_part(self, part):
lines = part.splitlines();
# First line is the HTTP status/reason
status, _, reason = lines[0].partition(b' ')[2].partition(b' ')
# Followed by headers and body
headers = []
body = b''
isBody = False
for line in lines[1:]:
if line == b'' and not isBody:
isBody = True
elif isBody:
body += line
else:
headerName, _, headerVal = line.partition(b':')
headers.append((headerName.lower(), headerVal))
return HTTPResponse(int(status), reason.strip(), headers, body)
def _report_batch_error(self, response):
xml = response.body.decode('utf-8')
doc = minidom.parseString(xml)
n = _get_children_from_path(doc, (METADATA_NS, 'error'), 'code')
code = n[0].firstChild.nodeValue if n and n[0].firstChild else ''
n = _get_children_from_path(doc, (METADATA_NS, 'error'), 'message')
message = n[0].firstChild.nodeValue if n and n[0].firstChild else xml
raise WindowsAzureBatchOperationError(message, code)