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
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
|
# Copyright 2013-2023 Lawrence Livermore National Security, LLC and other
# Spack Project Developers. See the top-level COPYRIGHT file for details.
#
# SPDX-License-Identifier: (Apache-2.0 OR MIT)
"""Test Spack's URL handling utility functions."""
import os
import os.path
import urllib.parse
import spack.util.path
import spack.util.url as url_util
def test_url_local_file_path(tmpdir):
# Create a file
path = str(tmpdir.join("hello.txt"))
with open(path, "wb") as f:
f.write(b"hello world")
# Go from path -> url -> path.
roundtrip = url_util.local_file_path(url_util.path_to_file_url(path))
# Verify it's the same file.
assert os.path.samefile(roundtrip, path)
# Test if it accepts urlparse objects
parsed = urllib.parse.urlparse(url_util.path_to_file_url(path))
assert os.path.samefile(url_util.local_file_path(parsed), path)
def test_url_local_file_path_no_file_scheme():
assert url_util.local_file_path("https://example.com/hello.txt") is None
assert url_util.local_file_path("C:\\Program Files\\hello.txt") is None
def test_relative_path_to_file_url(tmpdir):
# Create a file
path = str(tmpdir.join("hello.txt"))
with open(path, "wb") as f:
f.write(b"hello world")
with tmpdir.as_cwd():
roundtrip = url_util.local_file_path(url_util.path_to_file_url("hello.txt"))
assert os.path.samefile(roundtrip, path)
def test_url_join_local_paths():
# Resolve local link against page URL
# wrong:
assert (
url_util.join("s3://bucket/index.html", "../other-bucket/document.txt")
== "s3://bucket/other-bucket/document.txt"
)
# correct - need to specify resolve_href=True:
assert (
url_util.join("s3://bucket/index.html", "../other-bucket/document.txt", resolve_href=True)
== "s3://other-bucket/document.txt"
)
# same as above: make sure several components are joined together correctly
assert (
url_util.join(
# with resolve_href=True, first arg is the base url; can not be
# broken up
"s3://bucket/index.html",
# with resolve_href=True, remaining arguments are the components of
# the local href that needs to be resolved
"..",
"other-bucket",
"document.txt",
resolve_href=True,
)
== "s3://other-bucket/document.txt"
)
# Append local path components to prefix URL
# wrong:
assert (
url_util.join("https://mirror.spack.io/build_cache", "my-package", resolve_href=True)
== "https://mirror.spack.io/my-package"
)
# correct - Need to specify resolve_href=False:
assert (
url_util.join("https://mirror.spack.io/build_cache", "my-package", resolve_href=False)
== "https://mirror.spack.io/build_cache/my-package"
)
# same as above; make sure resolve_href=False is default
assert (
url_util.join("https://mirror.spack.io/build_cache", "my-package")
== "https://mirror.spack.io/build_cache/my-package"
)
# same as above: make sure several components are joined together correctly
assert (
url_util.join(
# with resolve_href=False, first arg is just a prefix. No
# resolution is done. So, there should be no difference between
# join('/a/b/c', 'd/e'),
# join('/a/b', 'c', 'd/e'),
# join('/a', 'b/c', 'd', 'e'), etc.
"https://mirror.spack.io",
"build_cache",
"my-package",
)
== "https://mirror.spack.io/build_cache/my-package"
)
# For s3:// URLs, the "netloc" (bucket) is considered part of the path.
# Make sure join() can cross bucket boundaries in this case.
args = ["s3://bucket/a/b", "new-bucket", "c"]
assert url_util.join(*args) == "s3://bucket/a/b/new-bucket/c"
args.insert(1, "..")
assert url_util.join(*args) == "s3://bucket/a/new-bucket/c"
args.insert(1, "..")
assert url_util.join(*args) == "s3://bucket/new-bucket/c"
# new-bucket is now the "netloc" (bucket name)
args.insert(1, "..")
assert url_util.join(*args) == "s3://new-bucket/c"
def test_url_join_absolute_paths():
# Handling absolute path components is a little tricky. To this end, we
# distinguish "absolute path components", from the more-familiar concept of
# "absolute paths" as they are understood for local filesystem paths.
#
# - All absolute paths are absolute path components. Joining a URL with
# these components has the effect of completely replacing the path of the
# URL with the absolute path. These components do not specify a URL
# scheme, so the scheme of the URL procuced when joining them depend on
# those provided by components that came before it (file:// assumed if no
# such scheme is provided).
# For eaxmple:
p = "/path/to/resource"
# ...is an absolute path
# http:// URL
assert url_util.join("http://example.com/a/b/c", p) == "http://example.com/path/to/resource"
# s3:// URL
# also notice how the netloc is treated as part of the path for s3:// URLs
assert url_util.join("s3://example.com/a/b/c", p) == "s3://path/to/resource"
# - URL components that specify a scheme are always absolute path
# components. Joining a base URL with these components effectively
# discards the base URL and "resets" the joining logic starting at the
# component in question and using it as the new base URL.
# For eaxmple:
p = "http://example.com/path/to"
# ...is an http:// URL
join_result = url_util.join(p, "resource")
assert join_result == "http://example.com/path/to/resource"
# works as if everything before the http:// URL was left out
assert url_util.join("literally", "does", "not", "matter", p, "resource") == join_result
assert url_util.join("file:///a/b/c", "./d") == "file:///a/b/c/d"
# Finally, resolve_href should have no effect for how absolute path
# components are handled because local hrefs can not be absolute path
# components.
args = [
"s3://does",
"not",
"matter",
"http://example.com",
"also",
"does",
"not",
"matter",
"/path",
]
expected = "http://example.com/path"
assert url_util.join(*args, resolve_href=True) == expected
assert url_util.join(*args, resolve_href=False) == expected
# resolve_href only matters for the local path components at the end of the
# argument list.
args[-1] = "/path/to/page"
args.extend(("..", "..", "resource"))
assert url_util.join(*args, resolve_href=True) == "http://example.com/resource"
assert url_util.join(*args, resolve_href=False) == "http://example.com/path/resource"
def test_default_download_name():
url = "https://example.com:1234/path/to/file.txt;params?abc=def#file=blob.tar"
filename = url_util.default_download_filename(url)
assert filename == spack.util.path.sanitize_filename(filename)
def test_default_download_name_dot_dot():
"""Avoid that downloaded files get names computed as ., .. or any hidden file."""
assert url_util.default_download_filename("https://example.com/.") == "_"
assert url_util.default_download_filename("https://example.com/..") == "_."
assert url_util.default_download_filename("https://example.com/.abcdef") == "_abcdef"
|
