Skip to content

Commit ff83437

Browse files
committed
Escape url parameters in sqlalchemy connection strings
1 parent cd614ff commit ff83437

File tree

5 files changed

+220
-29
lines changed

5 files changed

+220
-29
lines changed

README.md

Lines changed: 15 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -72,8 +72,13 @@ NOTE: `password` and `schema` are optional
7272
from sqlalchemy import create_engine
7373
from sqlalchemy.schema import Table, MetaData
7474
from sqlalchemy.sql.expression import select, text
75+
from trino.sqlalchemy import URL
7576

76-
engine = create_engine('trino://user@localhost:8080/system')
77+
engine = create_engine(URL(
78+
host="localhost",
79+
port=8080,
80+
catalog="system"
81+
))
7782
connection = engine.connect()
7883

7984
rows = connection.execute(text("SELECT * FROM runtime.nodes")).fetchall()
@@ -93,6 +98,7 @@ Attributes can also be passed in the connection string.
9398

9499
```python
95100
from sqlalchemy import create_engine
101+
from trino.sqlalchemy import URL
96102

97103
engine = create_engine(
98104
'trino://user@localhost:8080/system',
@@ -110,6 +116,14 @@ engine = create_engine(
110116
'&client_tags=["tag1", "tag2"]'
111117
'&experimental_python_types=true',
112118
)
119+
120+
# or using the URL factory method
121+
engine = create_engine(URL(
122+
host="localhost",
123+
port=8080,
124+
client_tags=["tag1", "tag2"],
125+
experimental_python_types=True
126+
))
113127
```
114128

115129
## Authentication mechanisms

tests/unit/sqlalchemy/test_dialect.py

Lines changed: 93 additions & 15 deletions
Original file line numberDiff line numberDiff line change
@@ -9,27 +9,43 @@
99
from trino.dbapi import Connection
1010
from trino.sqlalchemy.dialect import CertificateAuthentication, JWTAuthentication, TrinoDialect
1111
from trino.transaction import IsolationLevel
12+
from trino.sqlalchemy import URL as trino_url
1213

1314

1415
class TestTrinoDialect:
1516
def setup(self):
1617
self.dialect = TrinoDialect()
1718

1819
@pytest.mark.parametrize(
19-
"url, expected_args, expected_kwargs",
20+
"url, generated_url, expected_args, expected_kwargs",
2021
[
2122
(
22-
make_url("trino://user@localhost"),
23+
make_url(trino_url(
24+
user="user",
25+
host="localhost",
26+
)),
27+
'trino://user@localhost:8080?source=trino-sqlalchemy',
2328
list(),
24-
dict(host="localhost", catalog="system", user="user", source="trino-sqlalchemy"),
29+
dict(host="localhost", catalog="system", user="user", port=8080, source="trino-sqlalchemy"),
2530
),
2631
(
27-
make_url("trino://user@localhost:8080"),
32+
make_url(trino_url(
33+
user="user",
34+
host="localhost",
35+
port=443,
36+
)),
37+
'trino://user@localhost:443?source=trino-sqlalchemy',
2838
list(),
29-
dict(host="localhost", port=8080, catalog="system", user="user", source="trino-sqlalchemy"),
39+
dict(host="localhost", port=443, catalog="system", user="user", source="trino-sqlalchemy"),
3040
),
3141
(
32-
make_url("trino://user:pass@localhost:8080?source=trino-rulez"),
42+
make_url(trino_url(
43+
user="user",
44+
password="pass",
45+
host="localhost",
46+
source="trino-rulez",
47+
)),
48+
'trino://user:***@localhost:8080?source=trino-rulez',
3349
list(),
3450
dict(
3551
host="localhost",
@@ -42,13 +58,22 @@ def setup(self):
4258
),
4359
),
4460
(
45-
make_url(
46-
'trino://user@localhost:8080?'
47-
'session_properties={"query_max_run_time": "1d"}'
48-
'&http_headers={"trino": 1}'
49-
'&extra_credential=[("a", "b"), ("c", "d")]'
50-
'&client_tags=[1, "sql"]'
51-
'&experimental_python_types=true'),
61+
make_url(trino_url(
62+
user="user",
63+
host="localhost",
64+
session_properties={"query_max_run_time": "1d"},
65+
http_headers={"trino": 1},
66+
extra_credential=[("a", "b"), ("c", "d")],
67+
client_tags=["1", "sql"],
68+
experimental_python_types=True,
69+
)),
70+
'trino://user@localhost:8080'
71+
'?client_tags=%5B%221%22%2C+%22sql%22%5D'
72+
'&experimental_python_types=true'
73+
'&extra_credential=%5B%28%27a%27%2C+%27b%27%29%2C+%28%27c%27%2C+%27d%27%29%5D'
74+
'&http_headers=%7B%22trino%22%3A+1%7D'
75+
'&session_properties=%7B%22query_max_run_time%22%3A+%221d%22%7D'
76+
'&source=trino-sqlalchemy',
5277
list(),
5378
dict(
5479
host="localhost",
@@ -59,13 +84,66 @@ def setup(self):
5984
session_properties={"query_max_run_time": "1d"},
6085
http_headers={"trino": 1},
6186
extra_credential=[("a", "b"), ("c", "d")],
62-
client_tags=[1, "sql"],
87+
client_tags=["1", "sql"],
88+
experimental_python_types=True,
89+
),
90+
),
91+
# url encoding
92+
(
93+
make_url(trino_url(
94+
user="[email protected]/my_role",
95+
password="pass /*&",
96+
host="localhost",
97+
session_properties={"query_max_run_time": "1d"},
98+
http_headers={"trino": 1},
99+
extra_credential=[
100+
("[email protected]/my_role", "[email protected]/my_role"),
101+
("[email protected]/my_role", "[email protected]/my_role")],
63102
experimental_python_types=True,
103+
client_tags=["1 @& /\"", "sql"],
104+
verify=False,
105+
)),
106+
'trino://user%40test.org%2Fmy_role:***@localhost:8080'
107+
'?client_tags=%5B%221+%40%26+%2F%5C%22%22%2C+%22sql%22%5D'
108+
'&experimental_python_types=true'
109+
'&extra_credential=%5B%28%27user1%40test.org%2Fmy_role%27%2C'
110+
'+%27user2%40test.org%2Fmy_role%27%29%2C'
111+
'+%28%27user3%40test.org%2Fmy_role%27%2C'
112+
'+%27user36%40test.org%2Fmy_role%27%29%5D'
113+
'&http_headers=%7B%22trino%22%3A+1%7D'
114+
'&session_properties=%7B%22query_max_run_time%22%3A+%221d%22%7D'
115+
'&source=trino-sqlalchemy'
116+
'&verify=false',
117+
list(),
118+
dict(
119+
host="localhost",
120+
port=8080,
121+
catalog="system",
122+
user="[email protected]/my_role",
123+
auth=BasicAuthentication("[email protected]/my_role", "pass /*&"),
124+
http_scheme="https",
125+
source="trino-sqlalchemy",
126+
session_properties={"query_max_run_time": "1d"},
127+
http_headers={"trino": 1},
128+
extra_credential=[
129+
("[email protected]/my_role", "[email protected]/my_role"),
130+
("[email protected]/my_role", "[email protected]/my_role")],
131+
experimental_python_types=True,
132+
client_tags=["1 @& /\"", "sql"],
133+
verify=False,
64134
),
65135
),
66136
],
67137
)
68-
def test_create_connect_args(self, url: URL, expected_args: List[Any], expected_kwargs: Dict[str, Any]):
138+
def test_create_connect_args(
139+
self,
140+
url: URL,
141+
generated_url: str,
142+
expected_args: List[Any],
143+
expected_kwargs: Dict[str, Any]
144+
):
145+
assert repr(url) == generated_url
146+
69147
actual_args, actual_kwargs = self.dialect.create_connect_args(url)
70148

71149
assert actual_args == expected_args

trino/sqlalchemy/__init__.py

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -10,5 +10,6 @@
1010
# See the License for the specific language governing permissions and
1111
# limitations under the License.
1212
from sqlalchemy.dialects import registry
13+
from .util import _url as URL # noqa
1314

1415
registry.register("trino", "trino.sqlalchemy.dialect", "TrinoDialect")

trino/sqlalchemy/dialect.py

Lines changed: 17 additions & 13 deletions
Original file line numberDiff line numberDiff line change
@@ -13,6 +13,7 @@
1313
from ast import literal_eval
1414
from textwrap import dedent
1515
from typing import Any, Dict, List, Mapping, Optional, Sequence, Tuple
16+
from urllib.parse import unquote_plus, unquote
1617

1718
from sqlalchemy import exc, sql
1819
from sqlalchemy.engine.base import Connection
@@ -80,49 +81,52 @@ def create_connect_args(self, url: URL) -> Tuple[Sequence[Any], Mapping[str, Any
8081

8182
db_parts = (url.database or "system").split("/")
8283
if len(db_parts) == 1:
83-
kwargs["catalog"] = db_parts[0]
84+
kwargs["catalog"] = unquote_plus(db_parts[0])
8485
elif len(db_parts) == 2:
85-
kwargs["catalog"] = db_parts[0]
86-
kwargs["schema"] = db_parts[1]
86+
kwargs["catalog"] = unquote_plus(db_parts[0])
87+
kwargs["schema"] = unquote_plus(db_parts[1])
8788
else:
8889
raise ValueError(f"Unexpected database format {url.database}")
8990

9091
if url.username:
91-
kwargs["user"] = url.username
92+
kwargs["user"] = unquote(url.username)
9293

9394
if url.password:
9495
if not url.username:
9596
raise ValueError("Username is required when specify password in connection URL")
9697
kwargs["http_scheme"] = "https"
97-
kwargs["auth"] = BasicAuthentication(url.username, url.password)
98+
kwargs["auth"] = BasicAuthentication(unquote(url.username), unquote(url.password))
9899

99100
if "access_token" in url.query:
100101
kwargs["http_scheme"] = "https"
101-
kwargs["auth"] = JWTAuthentication(url.query["access_token"])
102+
kwargs["auth"] = JWTAuthentication(unquote(url.query["access_token"]))
102103

103104
if "cert" and "key" in url.query:
104105
kwargs["http_scheme"] = "https"
105-
kwargs["auth"] = CertificateAuthentication(url.query['cert'], url.query['key'])
106+
kwargs["auth"] = CertificateAuthentication(unquote(url.query['cert']), unquote(url.query['key']))
106107

107108
if "source" in url.query:
108-
kwargs["source"] = url.query["source"]
109+
kwargs["source"] = unquote(url.query["source"])
109110
else:
110111
kwargs["source"] = "trino-sqlalchemy"
111112

112113
if "session_properties" in url.query:
113-
kwargs["session_properties"] = json.loads(url.query["session_properties"])
114+
kwargs["session_properties"] = json.loads(unquote(url.query["session_properties"]))
114115

115116
if "http_headers" in url.query:
116-
kwargs["http_headers"] = json.loads(url.query["http_headers"])
117+
kwargs["http_headers"] = json.loads(unquote(url.query["http_headers"]))
117118

118119
if "extra_credential" in url.query:
119-
kwargs["extra_credential"] = literal_eval(url.query["extra_credential"])
120+
kwargs["extra_credential"] = literal_eval(unquote(url.query["extra_credential"]))
120121

121122
if "client_tags" in url.query:
122-
kwargs["client_tags"] = json.loads(url.query["client_tags"])
123+
kwargs["client_tags"] = json.loads(unquote(url.query["client_tags"]))
123124

124125
if "experimental_python_types" in url.query:
125-
kwargs["experimental_python_types"] = json.loads(url.query["experimental_python_types"])
126+
kwargs["experimental_python_types"] = json.loads(unquote(url.query["experimental_python_types"]))
127+
128+
if "verify" in url.query:
129+
kwargs["verify"] = json.loads(unquote(url.query["verify"]))
126130

127131
return args, kwargs
128132

trino/sqlalchemy/util.py

Lines changed: 94 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,94 @@
1+
import json
2+
from urllib.parse import quote_plus
3+
4+
from typing import Optional, Dict, List, Union, Tuple
5+
from sqlalchemy import exc
6+
from sqlalchemy.engine.url import _rfc_1738_quote # noqa
7+
8+
9+
def _url(
10+
host: str,
11+
port: Optional[int] = 8080,
12+
user: Optional[str] = None,
13+
password: Optional[str] = None,
14+
catalog: Optional[str] = None,
15+
schema: Optional[str] = None,
16+
source: Optional[str] = "trino-sqlalchemy",
17+
session_properties: Dict[str, str] = None,
18+
http_headers: Dict[str, Union[str, int]] = None,
19+
extra_credential: Optional[List[Tuple[str, str]]] = None,
20+
client_tags: Optional[List[str]] = None,
21+
experimental_python_types: Optional[bool] = None,
22+
access_token: Optional[str] = None,
23+
cert: Optional[str] = None,
24+
key: Optional[str] = None,
25+
verify: Optional[bool] = None,
26+
) -> str:
27+
"""
28+
Composes a SQLAlchemy connection string from the given database connection
29+
parameters.
30+
Parameters containing special characters (e.g., '@', '%') need to be encoded to be parsed correctly.
31+
"""
32+
33+
trino_url = "trino://"
34+
35+
if user is not None:
36+
trino_url += _rfc_1738_quote(user)
37+
38+
if password is not None:
39+
if user is None:
40+
raise exc.ArgumentError("user must be specified when specifying a password.")
41+
trino_url += f":{_rfc_1738_quote(password)}"
42+
43+
if user is not None:
44+
trino_url += "@"
45+
46+
if not host:
47+
raise exc.ArgumentError("host must be specified.")
48+
49+
trino_url += host
50+
51+
if not port:
52+
raise exc.ArgumentError("port must be specified.")
53+
54+
trino_url += f":{port}"
55+
56+
if catalog is not None:
57+
trino_url += f"/{quote_plus(catalog)}"
58+
59+
if schema is not None:
60+
if catalog is None:
61+
raise exc.ArgumentError("catalog must be specified when specifying a default schema.")
62+
trino_url += f"/{quote_plus(schema)}"
63+
64+
assert source
65+
trino_url += f"?source={quote_plus(source)}"
66+
67+
if session_properties is not None:
68+
trino_url += f"&session_properties={quote_plus(json.dumps(session_properties))}"
69+
70+
if http_headers is not None:
71+
trino_url += f"&http_headers={quote_plus(json.dumps(http_headers))}"
72+
73+
if extra_credential is not None:
74+
trino_url += f"&extra_credential={quote_plus(repr(extra_credential))}"
75+
76+
if client_tags is not None:
77+
trino_url += f"&client_tags={quote_plus(json.dumps(client_tags))}"
78+
79+
if experimental_python_types is not None:
80+
trino_url += f"&experimental_python_types={json.dumps(experimental_python_types)}"
81+
82+
if access_token is not None:
83+
trino_url += f"&access_token={quote_plus(access_token)}"
84+
85+
if cert is not None:
86+
trino_url += f"&cert={quote_plus(cert)}"
87+
88+
if key is not None:
89+
trino_url += f"&key={quote_plus(key)}"
90+
91+
if verify is not None:
92+
trino_url += f"&verify={json.dumps(verify)}"
93+
94+
return trino_url

0 commit comments

Comments
 (0)