1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
|
# -*- coding: utf-8 -*-
import base64
import hmac
import urllib
from hashlib import sha1 as sha
from urlparse import urlparse
from email.utils import formatdate
from requests.auth import AuthBase
class S3Auth(AuthBase):
"""Attaches AWS Authentication to the given Request object."""
service_base_url = 's3.amazonaws.com'
# List of Query String Arguments of Interest
special_params = [
'acl', 'location', 'logging', 'partNumber', 'policy', 'requestPayment',
'torrent', 'versioning', 'versionId', 'versions', 'website', 'uploads',
'uploadId', 'response-content-type', 'response-content-language',
'response-expires', 'response-cache-control', 'delete', 'lifecycle',
'response-content-disposition', 'response-content-encoding'
]
def __init__(self, access_key, secret_key, service_url=None):
if service_url:
self.service_base_url = service_url
self.access_key = str(access_key)
self.secret_key = str(secret_key)
def __call__(self, r):
# Create date header if it is not created yet.
if not 'date' in r.headers and not 'x-amz-date' in r.headers:
r.headers['date'] = formatdate(timeval=None, localtime=False, usegmt=True)
r.headers['Authorization'] = 'AWS %s:%s'%(self.access_key, self.get_signature(r))
return r
def get_signature(self, r):
canonical_string = self.get_canonical_string(r.url, r.headers, r.method)
h = hmac.new(self.secret_key, canonical_string, digestmod=sha)
return base64.encodestring(h.digest()).strip()
def get_canonical_string(self, url, headers, method):
parsedurl = urlparse(url)
objectkey = parsedurl.path[1:]
query_args = sorted(parsedurl.query.split('&'))
bucket = parsedurl.netloc[:-len(self.service_base_url)]
if len(bucket) > 1:
# remove last dot
bucket = bucket[:-1]
interesting_headers = {'content-md5':'', 'content-type':'', 'date':''}
for key in headers:
lk = key.lower()
if headers[key] and (lk in interesting_headers.keys() or lk.startswith('x-amz-')):
interesting_headers[lk] = headers[key].strip()
# If x-amz-date is used it supersedes the date header.
if interesting_headers.has_key('x-amz-date'):
interesting_headers['date'] = ''
buf = '%s\n' % method
for key in sorted(interesting_headers.keys()):
val = interesting_headers[key]
if key.startswith('x-amz-'):
buf += '%s:%s\n' % (key, val)
else:
buf += '%s\n' % val
# append the bucket if it exists
if bucket != '':
buf += '/%s' % bucket
# add the objectkey. even if it doesn't exist, add the slash
buf += '/%s' % urllib.unquote(objectkey)
params_found = False
# handle special query string arguments
for q in query_args:
k = q.split('=')[0]
if k in self.special_params:
if params_found:
buf += '&%s' % q
else:
buf += '?%s' % q
params_found = True
return buf
|