collate-data-diff 0.11.2__tar.gz → 0.11.5__tar.gz

{collate_data_diff-0.11.2 → collate_data_diff-0.11.5}/PKG-INFO

@@ -1,10 +1,10 @@
-Metadata-Version: 2.1
+Metadata-Version: 2.3
 Name: collate-data-diff
-Version: 0.11.2
+Version: 0.11.5
 Summary: Command-line tool and Python library to efficiently diff rows across two different databases.
-Home-page: https://github.com/open-metadata/collate-data-diff
 License: MIT
 Author: Collate
+Author-email: info@getcollate.io
 Requires-Python: >=3.8.0,<4.0
 Classifier: Development Status :: 2 - Pre-Alpha
 Classifier: Environment :: Console
@@ -18,6 +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
@@ -42,7 +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"
@@ -50,14 +52,14 @@ 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"
 Requires-Dist: typing-extensions (>=4.0.1)
 Requires-Dist: urllib3 (<2)
 Requires-Dist: vertica-python ; extra == "vertica" or extra == "all-dbs"
-Project-URL: Repository, https://github.com/open-metadata/collate-data-diff
+Project-URL: Repository, https://github.com/datafold/data-diff
 Description-Content-Type: text/markdown
 
 Forked from the archived repository [datafold/data-diff](https://github.com/datafold/data-diff)
@@ -65,6 +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">

collate_data_diff-0.11.5/README.md

@@ -0,0 +1,25 @@
+Forked from the archived repository [datafold/data-diff](https://github.com/datafold/data-diff)
+---
+
+# 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">
+  <img src="https://contributors-img.web.app/image?repo=datafold/data-diff" />
+</a>
+
+## License
+
+This project is licensed under the terms of the [MIT License](https://github.com/datafold/data-diff/blob/master/LICENSE).

{collate_data_diff-0.11.2 → collate_data_diff-0.11.5}/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-0.11.2 → collate_data_diff-0.11.5}/data_diff/databases/_connect.py

@@ -3,6 +3,7 @@ from typing import Hashable, MutableMapping, Type, Optional, Union, Dict
 from itertools import zip_longest
 from contextlib import suppress
 import weakref
+from urllib.parse import unquote
 
 import attrs
 import dsnparse
@@ -26,6 +27,15 @@ from data_diff.databases.duckdb import DuckDB
 from data_diff.databases.mssql import MsSQL
 
 
+class CustomParseResult(dsnparse.ParseResult):
+    @classmethod
+    def parse_credentials(cls, dsn):
+        username, password, dsn = super().parse_credentials(dsn)
+        if password:
+            password = unquote(password)
+        return username, password, dsn
+
+
 @attrs.frozen
 class MatchUriPath:
     database_cls: Type[Database]
@@ -136,7 +146,7 @@ class Connect:
         - duckdb
         """
 
-        dsn = dsnparse.parse(db_uri)
+        dsn = dsnparse.parse(db_uri, parse_class=CustomParseResult)
         if len(dsn.schemes) > 1:
             raise NotImplementedError("No support for multiple schemes")
         (scheme,) = dsn.schemes

{collate_data_diff-0.11.2 → collate_data_diff-0.11.5}/data_diff/databases/presto.py

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

{collate_data_diff-0.11.2 → collate_data_diff-0.11.5}/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-0.11.2 → collate_data_diff-0.11.5}/pyproject.toml

@@ -1,11 +1,11 @@
 [tool.poetry]
 name = "collate-data-diff"
-version = "0.11.2"
+version = "0.11.5"
 description = "Command-line tool and Python library to efficiently diff rows across two different databases."
-authors = ["Collate"]
+authors = ["Collate <info@getcollate.io>", "Datafold <data-diff@datafold.com>"]
 license = "MIT"
 readme = "README.md"
-repository = "https://github.com/open-metadata/collate-data-diff"
+repository = "https://github.com/datafold/data-diff"
 classifiers = [
     "Intended Audience :: Developers",
     "Intended Audience :: Information Technology",
@@ -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"

collate_data_diff-0.11.2/README.md

@@ -1,14 +0,0 @@
-Forked from the archived repository [datafold/data-diff](https://github.com/datafold/data-diff)
----
-
-# data-diff: Compare datasets fast, within or across SQL databases
-
-## Contributors
-
-<a href="https://github.com/datafold/data-diff/graphs/contributors">
-  <img src="https://contributors-img.web.app/image?repo=datafold/data-diff" />
-</a>
-
-## License
-
-This project is licensed under the terms of the [MIT License](https://github.com/datafold/data-diff/blob/master/LICENSE).