Refactor common database interface into Sqeleton (databases, queries)

data_diff/__init__.py

@@ -1,8 +1,8 @@
 from typing import Sequence, Tuple, Iterator, Optional, Union
 
 from .tracking import disable_tracking
-from .databases.connect import connect
-from .databases.database_types import DbKey, DbTime, DbPath
+from .databases import connect
+from .sqeleton.databases import DbKey, DbTime, DbPath
 from .diff_tables import Algorithm
 from .hashdiff_tables import HashDiffer, DEFAULT_BISECTION_THRESHOLD, DEFAULT_BISECTION_FACTOR
 from .joindiff_tables import JoinDiffer

data_diff/__main__.py

@@ -14,8 +14,8 @@
 from .hashdiff_tables import HashDiffer, DEFAULT_BISECTION_THRESHOLD, DEFAULT_BISECTION_FACTOR
 from .joindiff_tables import TABLE_WRITE_LIMIT, JoinDiffer
 from .table_segment import TableSegment
-from .databases.database_types import create_schema
-from .databases.connect import connect
+from .sqeleton.databases import create_schema
+from .databases import connect
 from .parse_time import parse_time_before_now, UNITS_STR, ParseError
 from .config import apply_config_from_file
 from .tracking import disable_tracking

data_diff/databases/__init__.py

@@ -1,4 +1,4 @@
-from .base import MD5_HEXDIGITS, CHECKSUM_HEXDIGITS, QueryError, ConnectError
+from data_diff.sqeleton.databases import MD5_HEXDIGITS, CHECKSUM_HEXDIGITS, QueryError, ConnectError
 
 from .postgresql import PostgreSQL
 from .mysql import MySQL
@@ -13,4 +13,4 @@
 from .vertica import Vertica
 from .duckdb import DuckDB
 
-from .connect import connect_to_uri
+from ._connect import connect

data_diff/databases/_connect.py

@@ -0,0 +1,42 @@
+from data_diff.sqeleton.databases.connect import MatchUriPath, Connect
+
+from .postgresql import PostgreSQL
+from .mysql import MySQL
+from .oracle import Oracle
+from .snowflake import Snowflake
+from .bigquery import BigQuery
+from .redshift import Redshift
+from .presto import Presto
+from .databricks import Databricks
+from .trino import Trino
+from .clickhouse import Clickhouse
+from .vertica import Vertica
+from .duckdb import DuckDB
+
+
+MATCH_URI_PATH = {
+    "postgresql": MatchUriPath(PostgreSQL, ["database?"], help_str="postgresql://<user>:<pass>@<host>/<database>"),
+    "mysql": MatchUriPath(MySQL, ["database?"], help_str="mysql://<user>:<pass>@<host>/<database>"),
+    "oracle": MatchUriPath(Oracle, ["database?"], help_str="oracle://<user>:<pass>@<host>/<database>"),
+    # "mssql": MatchUriPath(MsSQL, ["database?"], help_str="mssql://<user>:<pass>@<host>/<database>"),
+    "redshift": MatchUriPath(Redshift, ["database?"], help_str="redshift://<user>:<pass>@<host>/<database>"),
+    "snowflake": MatchUriPath(
+        Snowflake,
+        ["database", "schema"],
+        ["warehouse"],
+        help_str="snowflake://<user>:<pass>@<account>/<database>/<SCHEMA>?warehouse=<WAREHOUSE>",
+    ),
+    "presto": MatchUriPath(Presto, ["catalog", "schema"], help_str="presto://<user>@<host>/<catalog>/<schema>"),
+    "bigquery": MatchUriPath(BigQuery, ["dataset"], help_str="bigquery://<project>/<dataset>"),
+    "databricks": MatchUriPath(
+        Databricks,
+        ["catalog", "schema"],
+        help_str="databricks://:<access_token>@<server_name>/<http_path>",
+    ),
+    "trino": MatchUriPath(Trino, ["catalog", "schema"], help_str="trino://<user>@<host>/<catalog>/<schema>"),
+    "clickhouse": MatchUriPath(Clickhouse, ["database?"], help_str="clickhouse://<user>:<pass>@<host>/<database>"),
+    "vertica": MatchUriPath(Vertica, ["database?"], help_str="vertica://<user>:<pass>@<host>/<database>"),
+    "duckdb": MatchUriPath(DuckDB, ["database", "dbpath"], help_str="duckdb://<database>@<dbpath>"),
+}
+
+connect = Connect(MATCH_URI_PATH)