chore: import upstream snapshot with attribution

scripts/singlestore-test-helpers/docker-compose.yml

@@ -0,0 +1,21 @@
+services:
+  singlestore:
+    container_name: "singlestore"
+    image: ghcr.io/singlestore-labs/singlestoredb-dev:latest
+    platform: linux/amd64
+    ports:
+      - 3306:3306
+      - 8080:8080
+      - 9000:9000
+    environment:
+      - ROOT_PASSWORD=password
+    volumes:
+      - ./schema.sql:/init.sql
+
+  # Allow docker compose up --wait to exit only when singlestore is healthy
+  wait:
+    image: hello-world:latest
+    container_name: singlestore-waiter
+    depends_on:
+      singlestore:
+        condition: service_healthy

scripts/singlestore-test-helpers/schema.sql

@@ -0,0 +1,48 @@
+CREATE DATABASE ingest_test;
+USE ingest_test;
+
+CREATE TABLE elements (
+    id INT PRIMARY KEY NOT NULL AUTO_INCREMENT,
+    element_id TEXT,
+    text TEXT,
+    embeddings Vector(384),
+    type TEXT,
+    url TEXT,
+    version TEXT,
+    data_source_date_created TIMESTAMP,
+    data_source_date_modified TIMESTAMP,
+    data_source_date_processed TIMESTAMP,
+    data_source_permissions_data TEXT,
+    data_source_url TEXT,
+    data_source_version TEXT,
+    data_source_record_locator JSON,
+    category_depth INTEGER,
+    parent_id TEXT,
+    attached_filename TEXT,
+    filetype TEXT,
+    last_modified TIMESTAMP,
+    file_directory TEXT,
+    filename TEXT,
+    languages TEXT,
+    page_number TEXT,
+    links TEXT,
+    page_name TEXT,
+    link_urls TEXT,
+    link_texts TEXT,
+    sent_from TEXT,
+    sent_to TEXT,
+    subject TEXT,
+    section TEXT,
+    header_footer_type TEXT,
+    emphasized_text_contents TEXT,
+    emphasized_text_tags TEXT,
+    text_as_html TEXT,
+    detection_class_prob DECIMAL,
+    is_continuation BOOLEAN,
+    orig_elements TEXT,
+    coordinates_points TEXT,
+    coordinates_system TEXT,
+    coordinates_layout_width DECIMAL,
+    coordinates_layout_height DECIMAL
+);
+

scripts/singlestore-test-helpers/test_outputs.py

@@ -0,0 +1,56 @@
+#!/usr/bin/env python3
+
+import click
+import singlestoredb as s2
+from singlestoredb.connection import Connection
+
+
+def get_connection(
+    host: str = None, port: int = None, database: str = None, user: str = None, password: str = None
+) -> Connection:
+    conn = s2.connect(
+        host=host,
+        port=port,
+        database=database,
+        user=user,
+        password=password,
+    )
+    return conn
+
+
+def validate(table_name: str, conn: Connection, num_elements: int):
+    with conn.cursor() as cur:
+        stmt = f"select * from {table_name}"
+        count = cur.execute(stmt)
+        assert count == num_elements, (
+            f"found count ({count}) doesn't match expected value: {num_elements}"
+        )
+    print("validation successful")
+
+
+@click.command()
+@click.option("--host", type=str, default="localhost", show_default=True)
+@click.option("--port", type=int, default=3306, show_default=True)
+@click.option("--user", type=str, default="root", show_default=True)
+@click.option("--password", type=str, default="password")
+@click.option("--database", type=str, required=True)
+@click.option("--table-name", type=str, required=True)
+@click.option(
+    "--num-elements", type=int, required=True, help="The expected number of elements to exist"
+)
+def run_validation(
+    host: str,
+    port: int,
+    user: str,
+    database: str,
+    password: str,
+    table_name: str,
+    num_elements: int,
+):
+    print(f"Validating that table {table_name} in database {database} has {num_elements} entries")
+    conn = get_connection(host=host, port=port, database=database, user=user, password=password)
+    validate(table_name=table_name, conn=conn, num_elements=num_elements)
+
+
+if __name__ == "__main__":
+    run_validation()