continuo-validation-trino 0.1.0__py3-none-any.whl

continuo_validation_trino/__init__.py

@@ -0,0 +1 @@
+"""Trino warehouse adapter for the continuo validation runner."""

continuo_validation_trino/adapter.py

@@ -0,0 +1,130 @@
+"""Trino implementation of the WarehouseAdapter port.
+
+Trino's namespace is catalog.schema: the catalog comes from TRINO_CATALOG and every
+statement is fully qualified, so the contract's `schema` argument stays an opaque bare
+name. Two Trino traits shape this adapter: there is no DROP SCHEMA ... CASCADE, so
+teardown drops the schema's tables first; and there are no advisory locks, so
+ensure_schema relies on IF NOT EXISTS plus tolerating a concurrent creation.
+"""
+import logging
+import os
+
+from continuo_validation_contract.port import WarehouseAdapter
+
+import trino
+
+from trino.auth import BasicAuthentication
+
+logger = logging.getLogger("continuo_validation_trino")
+
+
+def _quote(identifier: str) -> str:
+    """Double-quote a Trino identifier, escaping any embedded double quotes."""
+    escaped = identifier.replace('"', '""')
+    return f'"{escaped}"'
+
+
+class TrinoAdapter(WarehouseAdapter):
+    """WarehouseAdapter speaking Trino DDL over the trino DBAPI."""
+
+    def __init__(self, conn: "trino.dbapi.Connection", catalog: str) -> None:
+        self._conn = conn
+        self._catalog = catalog
+
+    @classmethod
+    def required_env(cls) -> list[str]:
+        """Vars that must be non-empty before connecting."""
+        return ["TRINO_HOST", "TRINO_CATALOG"]
+
+    @classmethod
+    def from_env(cls) -> "TrinoAdapter":
+        """Connect from TRINO_* env (port 8080, user continuo, http; password optional)."""
+        catalog = os.environ["TRINO_CATALOG"]
+        user = os.environ.get("TRINO_USER", "continuo")
+        http_scheme = os.environ.get("TRINO_HTTP_SCHEME", "http")
+        password = os.environ.get("TRINO_PASSWORD", "")
+        if password and http_scheme != "https":
+            raise ValueError(
+                "TRINO_PASSWORD is set but TRINO_HTTP_SCHEME is not 'https'; "
+                "Trino refuses basic auth over plaintext"
+            )
+        conn = trino.dbapi.connect(
+            host=os.environ["TRINO_HOST"],
+            port=int(os.environ.get("TRINO_PORT", "8080")),
+            user=user,
+            catalog=catalog,
+            http_scheme=http_scheme,
+            auth=BasicAuthentication(user, password) if password else None,
+        )
+        return cls(conn, catalog)
+
+    def _schema_ref(self, schema: str) -> str:
+        return f"{_quote(self._catalog)}.{_quote(schema)}"
+
+    def _table_ref(self, schema: str, table: str) -> str:
+        return f"{self._schema_ref(schema)}.{_quote(table)}"
+
+    def _execute(self, statement: str) -> list[tuple]:
+        """Run one statement to completion and return its rows.
+
+        The trino DBAPI is lazy: execute() only starts the query, so the results must
+        be consumed for DDL to actually take effect.
+        """
+        cur = self._conn.cursor()
+        try:
+            cur.execute(statement)
+            rows: list[tuple] = cur.fetchall()
+            return rows
+        finally:
+            cur.close()
+
+    def ensure_schema(self, schema: str) -> None:
+        """Idempotently create *schema*; safe under concurrent callers."""
+        logger.info("ensuring candidate schema %s.%s exists", self._catalog, schema)
+        try:
+            self._execute(f"CREATE SCHEMA IF NOT EXISTS {self._schema_ref(schema)}")
+        except trino.exceptions.TrinoUserError as exc:
+            # Trino has no advisory locks; a concurrent creator can still win the race
+            # between IF NOT EXISTS and the metastore write. The desired end state holds.
+            if "already exists" not in str(exc).lower():
+                raise
+            logger.info("schema %s already exists (concurrent create); continuing", schema)
+
+    def drop_schema(self, schema: str) -> None:
+        """Idempotently drop *schema* and everything in it; no-op if absent.
+
+        Trino has no DROP SCHEMA ... CASCADE, so the schema's tables are dropped first.
+        """
+        ref = self._schema_ref(schema)
+        try:
+            rows = self._execute(f"SHOW TABLES FROM {ref}")
+        except trino.exceptions.TrinoUserError as exc:
+            if "does not exist" not in str(exc).lower():
+                raise
+            logger.info("candidate schema %s absent; nothing to drop", schema)
+            return
+        logger.info("dropping candidate schema %s (%d tables)", schema, len(rows))
+        for row in rows:
+            self._execute(f"DROP TABLE IF EXISTS {self._table_ref(schema, row[0])}")
+        self._execute(f"DROP SCHEMA IF EXISTS {ref}")
+
+    def build_empty_from_sql(self, schema: str, table: str, compiled_sql: str) -> None:
+        """Create ``schema.table`` empty, shaped by the compiled SELECT."""
+        # Strip any trailing terminator so the SELECT nests cleanly inside AS ( ... ).
+        inner = compiled_sql.strip().rstrip(";").strip()
+        ref = self._table_ref(schema, table)
+        self._execute(f"DROP TABLE IF EXISTS {ref}")
+        self._execute(f"CREATE TABLE {ref} AS ({inner}) WITH NO DATA")
+
+    def clone_empty_from_prod(self, candidate_schema: str, prod_schema: str, table: str) -> None:
+        """Create ``candidate_schema.table`` empty, shaped like ``prod_schema.table``."""
+        ref = self._table_ref(candidate_schema, table)
+        self._execute(f"DROP TABLE IF EXISTS {ref}")
+        self._execute(
+            f"CREATE TABLE {ref} AS "
+            f"SELECT * FROM {self._table_ref(prod_schema, table)} WITH NO DATA"
+        )
+
+    def close(self) -> None:
+        """Release the underlying connection."""
+        self._conn.close()

continuo_validation_trino-0.1.0.dist-info/METADATA

@@ -0,0 +1,37 @@
+Metadata-Version: 2.4
+Name: continuo-validation-trino
+Version: 0.1.0
+Summary: Trino warehouse adapter for the continuo validation runner.
+Author: Simone Carolini
+Maintainer: Simone Carolini
+Classifier: Development Status :: 4 - Beta
+Classifier: Intended Audience :: Developers
+Classifier: Programming Language :: Python :: 3.14
+Requires-Python: >=3.14
+Requires-Dist: continuo-validation-contract==0.2.0
+Requires-Dist: trino==0.338.0
+Description-Content-Type: text/markdown
+
+# continuo-validation-trino
+
+Trino warehouse adapter for the [continuo](https://github.com/carolsimone/continuo)
+validation runner. Implements `continuo_validation_contract.port.WarehouseAdapter`;
+continuo's harness discovers it through the `continuo_validation.adapters` entry point.
+
+Trino's namespace is `catalog.schema`. The catalog comes from `TRINO_CATALOG` and every
+statement is fully qualified, so continuo passes a bare candidate schema name and never
+needs to know about catalogs.
+
+## Connection env
+
+| Env | Required | Default | Purpose |
+|---|---|---|---|
+| `TRINO_HOST` | yes | — | Coordinator host |
+| `TRINO_CATALOG` | yes | — | Catalog holding candidate schemas (e.g. `iceberg`) |
+| `TRINO_PORT` | no | `8080` | Coordinator port |
+| `TRINO_USER` | no | `continuo` | Trino user |
+| `TRINO_HTTP_SCHEME` | no | `http` | `http` or `https` |
+| `TRINO_PASSWORD` | no | unset | Basic auth; requires `TRINO_HTTP_SCHEME=https` |
+
+The catalog must be backed by a connector that supports schema and table creation
+(Iceberg and Hive do). The library is tested live against Iceberg (REST catalog + S3).

continuo_validation_trino-0.1.0.dist-info/RECORD

@@ -0,0 +1,6 @@
+continuo_validation_trino/__init__.py,sha256=xN9flSm3q9z3FPg4ol2bqhsPF3CvoHAaw6rkE8fCXa0,66
+continuo_validation_trino/adapter.py,sha256=2zNv2xLf8lbU-ccoUMpueG5gM8Z5w0qDTUlaTcwRjIY,5478
+continuo_validation_trino-0.1.0.dist-info/METADATA,sha256=ySIHFNs5s5uIbFUxnJcYnJ2MSBxO5aGfVV0TKvid2hk,1585
+continuo_validation_trino-0.1.0.dist-info/WHEEL,sha256=lCkmxWfQsSc9CfIClYeavTdQeEX2toPqufh9gI35EQA,87
+continuo_validation_trino-0.1.0.dist-info/entry_points.txt,sha256=2FGDkX5J9yrnyYlmoetrgm59lgMk7JYEq1CJ1jklH4k,86
+continuo_validation_trino-0.1.0.dist-info/RECORD,,

continuo_validation_trino-0.1.0.dist-info/WHEEL

@@ -0,0 +1,4 @@
+Wheel-Version: 1.0
+Generator: hatchling 1.31.0
+Root-Is-Purelib: true
+Tag: py3-none-any

continuo_validation_trino-0.1.0.dist-info/entry_points.txt

@@ -0,0 +1,2 @@
+[continuo_validation.adapters]
+trino = continuo_validation_trino.adapter:TrinoAdapter