collate-data-diff 0.11.3__tar.gz → 0.11.6__tar.gz

{collate_data_diff-0.11.3 → collate_data_diff-0.11.6}/PKG-INFO

@@ -1,8 +1,7 @@
-Metadata-Version: 2.1
+Metadata-Version: 2.3
 Name: collate-data-diff
-Version: 0.11.3
+Version: 0.11.6
 Summary: Command-line tool and Python library to efficiently diff rows across two different databases.
-Home-page: https://github.com/datafold/data-diff
 License: MIT
 Author: Collate
 Author-email: info@getcollate.io
@@ -19,6 +18,7 @@ Classifier: Programming Language :: Python :: 3.9
 Classifier: Programming Language :: Python :: 3.10
 Classifier: Programming Language :: Python :: 3.11
 Classifier: Programming Language :: Python :: 3.12
+Classifier: Programming Language :: Python :: 3.13
 Classifier: Topic :: Database :: Database Engines/Servers
 Classifier: Typing :: Typed
 Provides-Extra: all-dbs
@@ -43,7 +43,8 @@ Requires-Dist: dsnparse (<0.2.0)
 Requires-Dist: duckdb ; extra == "duckdb" or extra == "all-dbs"
 Requires-Dist: keyring
 Requires-Dist: mashumaro[msgpack] (>=2.9,<3.11.0)
-Requires-Dist: mysql-connector-python (>=8.0.29) ; extra == "mysql" or extra == "all-dbs"
+Requires-Dist: mysql-connector-python (>=8.0.29) ; (python_version < "3.9") and (extra == "mysql" or extra == "all-dbs")
+Requires-Dist: mysql-connector-python (>=9.1) ; (python_version >= "3.9") and (extra == "mysql" or extra == "all-dbs")
 Requires-Dist: oracledb ; extra == "oracle" or extra == "all-dbs"
 Requires-Dist: preql (>=0.2.19) ; extra == "preql" or extra == "all-dbs"
 Requires-Dist: presto-python-client ; extra == "presto" or extra == "all-dbs"
@@ -51,7 +52,7 @@ Requires-Dist: psycopg2 ; extra == "postgresql" or extra == "redshift" or extra
 Requires-Dist: pydantic (>=1.10.12)
 Requires-Dist: pyodbc (>=4.0.39) ; extra == "mssql" or extra == "all-dbs"
 Requires-Dist: rich
-Requires-Dist: snowflake-connector-python (>=3.0.2,<4.0.0) ; extra == "snowflake" or extra == "all-dbs"
+Requires-Dist: snowflake-connector-python (>=3.13.1,<4.0.0) ; extra == "snowflake" or extra == "all-dbs"
 Requires-Dist: tabulate (>=0.9.0)
 Requires-Dist: toml (>=0.10.2)
 Requires-Dist: trino (>=0.314.0) ; extra == "trino" or extra == "all-dbs"
@@ -66,6 +67,17 @@ Forked from the archived repository [datafold/data-diff](https://github.com/data
 
 # data-diff: Compare datasets fast, within or across SQL databases
 
+## Publishing
+
+To publish a new version of the package, follow these steps:
+
+1. Bump the version number in the [`pyproject.toml`](https://github.com/open-metadata/collate-data-diff/blob/main/pyproject.toml#L3) file.
+2. Trigger the [publish action](https://github.com/open-metadata/collate-data-diff/actions/workflows/publish.yml).
+
+To publish to Test PyPI, set the "test" input to "true" when triggering the action.
+
+To publish a dev release, set the "dev" input to "true" when triggering the action.
+
 ## Contributors
 
 <a href="https://github.com/datafold/data-diff/graphs/contributors">
@@ -76,5 +88,3 @@ Forked from the archived repository [datafold/data-diff](https://github.com/data
 
 This project is licensed under the terms of the [MIT License](https://github.com/datafold/data-diff/blob/master/LICENSE).
 
-## Publishing
-

{collate_data_diff-0.11.3 → collate_data_diff-0.11.6}/README.md

@@ -3,6 +3,17 @@ Forked from the archived repository [datafold/data-diff](https://github.com/data
 
 # data-diff: Compare datasets fast, within or across SQL databases
 
+## Publishing
+
+To publish a new version of the package, follow these steps:
+
+1. Bump the version number in the [`pyproject.toml`](https://github.com/open-metadata/collate-data-diff/blob/main/pyproject.toml#L3) file.
+2. Trigger the [publish action](https://github.com/open-metadata/collate-data-diff/actions/workflows/publish.yml).
+
+To publish to Test PyPI, set the "test" input to "true" when triggering the action.
+
+To publish a dev release, set the "dev" input to "true" when triggering the action.
+
 ## Contributors
 
 <a href="https://github.com/datafold/data-diff/graphs/contributors">
@@ -12,5 +23,3 @@ Forked from the archived repository [datafold/data-diff](https://github.com/data
 ## License
 
 This project is licensed under the terms of the [MIT License](https://github.com/datafold/data-diff/blob/master/LICENSE).
-
-## Publishing

{collate_data_diff-0.11.3 → collate_data_diff-0.11.6}/data_diff/__init__.py

@@ -32,12 +32,35 @@ def connect_to_table(
     if isinstance(key_columns, str):
         key_columns = (key_columns,)
 
-    db: Database = connect(db_info, thread_count=thread_count)
+    kw, kwargs = _get_table_segment_args(**kwargs)
+
+    db: Database = connect(db_info, thread_count=thread_count, **kwargs)
 
     if isinstance(table_name, str):
         table_name = db.dialect.parse_table_name(table_name)
 
-    return TableSegment(db, table_name, key_columns, **kwargs)
+    return TableSegment(db, table_name, key_columns, **kw)
+
+def _get_table_segment_args(**kwargs):
+    accepted_kwargs = {
+        "update_column",
+        "extra_columns",
+        "min_key",
+        "max_key",
+        "min_update",
+        "max_update",
+        "where",
+        "case_sensitive",
+    }
+    kw = {}
+    copy_kwargs = dict(kwargs)
+    for k, v in copy_kwargs.items():
+        if k in accepted_kwargs:
+            kw[k] = kwargs.pop(k)
+        if v is None and k not in accepted_kwargs:
+            kwargs.pop(k)
+
+    return kw, kwargs
 
 
 def diff_tables(

{collate_data_diff-0.11.3 → collate_data_diff-0.11.6}/data_diff/databases/_connect.py

@@ -1,4 +1,5 @@
 import logging
+import json
 from typing import Hashable, MutableMapping, Type, Optional, Union, Dict
 from itertools import zip_longest
 from contextlib import suppress
@@ -291,7 +292,7 @@ class Connect:
 
     def __make_cache_key(self, db_conf: Union[str, dict]) -> Hashable:
         if isinstance(db_conf, dict):
-            return tuple(db_conf.items())
+            return json.dumps(db_conf)
         return db_conf
 
 

{collate_data_diff-0.11.3 → collate_data_diff-0.11.6}/data_diff/databases/presto.py

@@ -154,6 +154,9 @@ class Presto(Database):
 
     def __init__(self, **kw) -> None:
         super().__init__()
+        if not len(kw) or kw.get("driver") == "trino":
+            # this is a trino connection
+            return
         self.default_schema = "public"
         prestodb = import_presto()
 

{collate_data_diff-0.11.3 → collate_data_diff-0.11.6}/data_diff/databases/snowflake.py

@@ -139,11 +139,11 @@ class Snowflake(Database):
             with open(key, "rb") as f:
                 key_bytes = f.read()
         if key_content:
-            key_bytes = base64.b64decode(key_content)
+            key_bytes = bytes(key_content, "utf-8")
 
         # If a private key is used, read it from the specified path and pass it as "private_key" to the connector.
         if key_bytes:
-            if "password" in kw:
+            if "password" in kw and (kw.get("password") != "" or kw.get("password") is not None):
                 raise ConnectError("Cannot use password and key at the same time")
             if kw.get("private_key_passphrase"):
                 encoded_passphrase = kw.get("private_key_passphrase").encode()

{collate_data_diff-0.11.3 → collate_data_diff-0.11.6}/data_diff/databases/trino.py

@@ -1,6 +1,7 @@
 from typing import Any, ClassVar, Type
 
 import attrs
+from requests import Session
 
 from data_diff.abcs.database_types import TemporalType, ColType_UUID
 from data_diff.databases import presto
@@ -47,4 +48,30 @@ class Trino(presto.Presto):
         if kw.get("schema"):
             self.default_schema = kw.get("schema")
 
+        if kw.get("http_session"):
+            session = Session()
+            session.proxies = kw.get("http_session", {}).get("proxies")
+
+            kw["http_session"] = session
+
+        auth = kw.get("auth")
+
+        if auth:
+            if auth.get("authType") == "basic":
+                kw["auth"] = trino.auth.BasicAuthentication(
+                    auth.get("username"),
+                    auth.get("password")
+                )
+                kw["http_scheme"] = "https"
+
+            elif auth.get("authType") == "jwt":
+                kw["auth"] = trino.auth.JWTAuthentication(
+                    auth.get("jwt")
+                )
+                kw["http_scheme"] = "https"
+
+            elif auth.get("authType") == "oauth2":
+                kw["auth"] = trino.auth.OAuth2Authentication()
+                kw["http_scheme"] = "https"
+                
         self._conn = trino.dbapi.connect(**kw)

{collate_data_diff-0.11.3 → collate_data_diff-0.11.6}/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "collate-data-diff"
-version = "0.11.3"
+version = "0.11.6"
 description = "Command-line tool and Python library to efficiently diff rows across two different databases."
 authors = ["Collate <info@getcollate.io>", "Datafold <data-diff@datafold.com>"]
 license = "MIT"
@@ -27,9 +27,12 @@ dsnparse = "<0.2.0"
 click = ">=8.1"
 rich = "*"
 toml = ">=0.10.2"
-mysql-connector-python = {version=">=8.0.29", optional=true}
+mysql-connector-python = [
+    {version=">=8.0.29", python = "<3.9", optional=true},
+    {version=">=9.1", python = ">=3.9", optional=true}
+]
 psycopg2 = {version="*", optional=true}
-snowflake-connector-python = {version = ">=3.0.2,<4.0.0", optional=true}
+snowflake-connector-python = {version = ">=3.13.1,<4.0.0", optional=true}
 cryptography = {version="*", optional=true}
 trino = {version=">=0.314.0", optional=true}
 presto-python-client = {version="*", optional=true}
@@ -51,9 +54,12 @@ mashumaro = {version = ">=2.9,<3.11.0", extras = ["msgpack"]}
 parameterized = "*"
 unittest-parallel = "*"
 preql = ">=0.2.19"
-mysql-connector-python = "*"
+mysql-connector-python = [
+    {version=">=8.0.29", python = "<3.9"},
+    {version=">=9.1", python = ">=3.9"}
+]
 psycopg2 = "*"
-snowflake-connector-python = ">=3.0.2,<4.0.0"
+snowflake-connector-python = ">=3.13.1,<4.0.0"
 cryptography = "*"
 trino = ">=0.314.0"
 presto-python-client = ">=0.8.3"