fix: Support fetching multiple databases and schemas and their associated tables

duckdb_engine/__init__.py

@@ -1,3 +1,4 @@
+import re
 import warnings
 from typing import (
     TYPE_CHECKING,
@@ -23,6 +24,7 @@
 from sqlalchemy.dialects.postgresql.base import PGDialect, PGInspector, PGTypeCompiler
 from sqlalchemy.dialects.postgresql.psycopg2 import PGDialect_psycopg2
 from sqlalchemy.engine.default import DefaultDialect
+from sqlalchemy.engine.reflection import cache
 from sqlalchemy.engine.url import URL
 from sqlalchemy.ext.compiler import compiles
 
@@ -31,6 +33,8 @@
 
 __version__ = "0.9.3"
 sqlalchemy_version = Version(sqlalchemy.__version__)
+duckdb_version: str = duckdb.__version__  # type: ignore[attr-defined]
+supports_attach: bool = Version(duckdb_version) >= Version("0.7.0")
 
 if TYPE_CHECKING:
     from sqlalchemy.base import Connection
@@ -265,6 +269,72 @@ def get_view_names(
 
         return [row[0] for row in rs]
 
+    @cache  # type: ignore[call-arg]
+    def get_schema_names(self, connection: "Connection", **kw: "Any"):  # type: ignore[no-untyped-def]
+        """
+        Return unquoted database_name.schema_name unless either contains spaces or double quotes.
+        In that case, escape double quotes and then wrap in double quotes.
+        SQLAlchemy definition of a schema includes database name for databases like SQL Server (Ex: databasename.dbo)
+        (see https://docs.sqlalchemy.org/en/20/dialects/mssql.html#multipart-schema-names)
+        """
+
+        if not supports_attach:
+            return super().get_schema_names(connection, **kw)
+
+        s = """
+            SELECT database_name, schema_name AS npspname
+            FROM duckdb_schemas()
+            WHERE schema_name NOT LIKE 'pg\\_%' ESCAPE '\\'
+            ORDER BY database_name, npspname
+            """
+        rs = connection.execute(text(s))
+
+        qs = self.identifier_preparer.quote_schema
+        return [f"{qs(db)}.{qs(schema)}" for (db, schema) in rs]
+
+    @cache  # type: ignore[call-arg]
+    def get_table_names(self, connection: "Connection", schema=None, **kw: "Any"):  # type: ignore[no-untyped-def]
+        """
+        Return unquoted database_name.schema_name unless either contains spaces or double quotes.
+        In that case, escape double quotes and then wrap in double quotes.
+        SQLAlchemy definition of a schema includes database name for databases like SQL Server (Ex: databasename.dbo)
+        (see https://docs.sqlalchemy.org/en/20/dialects/mssql.html#multipart-schema-names)
+        """
+
+        if not supports_attach:
+            return super().get_table_names(connection, schema, **kw)
+
+        s = """
+            SELECT database_name, schema_name, table_name
+            FROM duckdb_tables()
+            WHERE schema_name NOT LIKE 'pg\\_%' ESCAPE '\\'
+            """
+        params = {}
+        if schema is not None:
+            params = {"schema_name": schema}
+            if "." in schema:
+                # Get database name and schema name from schema if it contains a database name
+                # Format:
+                #   <db_name>.<schema_name>
+                #   db_name and schema_name are double quoted if contains spaces or double quotes
+                database_name, schema_name = (
+                    max(s) for s in re.findall(r'"([^.]+)"|([^.]+)', schema)
+                )
+                params = {"database_name": database_name, "schema_name": schema_name}
+                s += "AND database_name = :database_name\n"
+            s += "AND schema_name = :schema_name"
+
+        rs = connection.execute(text(s), params)
+
+        return [
+            table
+            for (
+                db,
+                sc,
+                table,
+            ) in rs
+        ]
+
     def get_indexes(
         self,
         connection: "Connection",

duckdb_engine/tests/test_basic.py

@@ -35,7 +35,7 @@
 from sqlalchemy.ext.declarative import declarative_base
 from sqlalchemy.orm import Session, relationship, sessionmaker
 
-from .. import DBAPI, Dialect
+from .. import DBAPI, Dialect, supports_attach
 
 try:
     # sqlalchemy 2
@@ -166,6 +166,51 @@ def test_get_tables(inspector: Inspector) -> None:
     assert inspector.get_view_names() == []
 
 
+@mark.skipif(
+    supports_attach is False,
+    reason="ATTACH is not supported for DuckDB version < 0.7.0",
+)
+def test_get_schema_names(inspector: Inspector, engine: Engine) -> None:
+    # Run each command in a separate transaction
+    for cmd in [
+        """CREATE SCHEMA "quack quack" """,
+        """ATTACH ':memory:' AS "daffy duck" """,
+        """CREATE SCHEMA "daffy duck"."quack quack" """,
+        """CREATE TABLE "daffy duck"."quack quack"."t1" (i INTEGER, j INTEGER);""",
+        """CREATE TABLE "daffy duck"."quack quack"."t2" (i INTEGER, j INTEGER);""",
+        """CREATE SCHEMA "daffy duck"."you're "" despicable" """,
+    ]:
+        with engine.connect() as conn:
+            # Using multi-line strings because of all the single and double quotes flying around...
+            conn.execute(text(cmd))
+
+    # Deliberately excluding pg_catalog schema (to align with Postgres)
+    names = inspector.get_schema_names()
+    if supports_attach:
+        assert set(names) == {
+            'memory."quack quack"',
+            "memory.information_schema",
+            "memory.main",
+            '"daffy duck"."you\'re "" despicable"',
+            '"daffy duck"."quack quack"',
+            '"daffy duck".information_schema',
+            '"daffy duck".main',
+            "system.information_schema",
+            "system.main",
+            "temp.information_schema",
+            "temp.main",
+        }
+    else:
+        assert names == ["quack quack", "information_schema", "main", "temp"]
+
+    table_names = inspector.get_table_names(schema='"daffy duck"."quack quack"')
+    assert set(table_names) == {"t1", "t2"}
+
+    table_names_all = inspector.get_table_names()
+    assert "t1" in table_names_all
+    assert "t2" in table_names_all
+
+
 def test_get_views(engine: Engine) -> None:
     con = engine.connect()
     views = engine.dialect.get_view_names(con)