docs/document_loader.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Google Cloud SQL for SQL server\n",
    "\n",
    "> [Cloud SQL](https://cloud.google.com/sql) is a fully managed relational database service that offers high performance, seamless integration, and impressive scalability. It offers [MySQL](https://cloud.google.com/sql/mysql), [PostgreSQL](https://cloud.google.com/sql/postgres), and [SQL Server](https://cloud.google.com/sql/sqlserver) database engines. Extend your database application to build AI-powered experiences leveraging Cloud SQL's Langchain integrations.\n",
    "\n",
    "This notebook goes over how to use [Cloud SQL for SQL server](https://cloud.google.com/sql/sqlserver) to [save, load and delete langchain documents](https://python.langchain.com/docs/modules/data_connection/document_loaders/) with `MSSQLLoader` and `MSSQLDocumentSaver`.\n",
    "\n",
    "Learn more about the package on [GitHub](https://github.com/googleapis/langchain-google-cloud-sql-mssql-python/).\n",
    "\n",
    "[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/googleapis/langchain-google-cloud-sql-mssql-python/blob/main/docs/document_loader.ipynb)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Before You Begin\n",
    "\n",
    "To run this notebook, you will need to do the following:\n",
    "\n",
    "* [Create a Google Cloud Project](https://developers.google.com/workspace/guides/create-project)\n",
    "* [Enable the Cloud SQL Admin API.](https://console.cloud.google.com/marketplace/product/google/sqladmin.googleapis.com)\n",
    "* [Create a Cloud SQL for SQL server instance](https://cloud.google.com/sql/docs/sqlserver/create-instance)\n",
    "* [Create a Cloud SQL database](https://cloud.google.com/sql/docs/sqlserver/create-manage-databases)\n",
    "* [Add an IAM database user to the database](https://cloud.google.com/sql/docs/sqlserver/create-manage-users) (Optional)\n",
    "\n",
    "After confirmed access to database in the runtime environment of this notebook, filling the following values and run the cell before running example scripts."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# @markdown Please fill in the both the Google Cloud region and name of your Cloud SQL instance.\n",
    "REGION = \"us-central1\"  # @param {type:\"string\"}\n",
    "INSTANCE = \"test-instance\"  # @param {type:\"string\"}\n",
    "\n",
    "# @markdown Please fill in user name and password of your Cloud SQL instance.\n",
    "DB_USER = \"sqlserver\"  # @param {type:\"string\"}\n",
    "DB_PASS = \"password\"  # @param {type:\"string\"}\n",
    "\n",
    "# @markdown Please specify a database and a table for demo purpose.\n",
    "DATABASE = \"test\"  # @param {type:\"string\"}\n",
    "TABLE_NAME = \"test-default\"  # @param {type:\"string\"}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 🦜🔗 Library Installation\n",
    "\n",
    "The integration lives in its own `langchain-google-cloud-sql-mssql` package, so we need to install it."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "%pip install --upgrade --quiet langchain-google-cloud-sql-mssql"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Colab only**: Uncomment the following cell to restart the kernel or use the button to restart the kernel. For Vertex AI Workbench you can restart the terminal using the button on top."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# # Automatically restart kernel after installs so that your environment can access the new packages\n",
    "# import IPython\n",
    "\n",
    "# app = IPython.Application.instance()\n",
    "# app.kernel.do_shutdown(True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 🔐 Authentication\n",
    "\n",
    "Authenticate to Google Cloud as the IAM user logged into this notebook in order to access your Google Cloud Project.\n",
    "\n",
    "- If you are using Colab to run this notebook, use the cell below and continue.\n",
    "- If you are using Vertex AI Workbench, check out the setup instructions [here](https://github.com/GoogleCloudPlatform/generative-ai/tree/main/setup-env)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from google.colab import auth\n",
    "\n",
    "auth.authenticate_user()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### ☁ Set Your Google Cloud Project\n",
    "Set your Google Cloud project so that you can leverage Google Cloud resources within this notebook.\n",
    "\n",
    "If you don't know your project ID, try the following:\n",
    "\n",
    "* Run `gcloud config list`.\n",
    "* Run `gcloud projects list`.\n",
    "* See the support page: [Locate the project ID](https://support.google.com/googleapi/answer/7014113)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# @markdown Please fill in the value below with your Google Cloud project ID and then run the cell.\n",
    "\n",
    "PROJECT_ID = \"my-project-id\"  # @param {type:\"string\"}\n",
    "\n",
    "# Set the project id\n",
    "!gcloud config set project {PROJECT_ID}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 💡 API Enablement\n",
    "The `langchain-google-cloud-sql-mssql` package requires that you [enable the Cloud SQL Admin API](https://console.cloud.google.com/flows/enableapi?apiid=sqladmin.googleapis.com) in your Google Cloud Project."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# enable Cloud SQL Admin API\n",
    "!gcloud services enable sqladmin.googleapis.com"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Basic Usage"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### MSSQLEngine Connection Pool\n",
    "\n",
    "Before saving or loading documents from MSSQL table, we need first configures a connection pool to Cloud SQL database. The `MSSQLEngine` configures a [SQLAlchemy connection pool](https://docs.sqlalchemy.org/en/20/core/pooling.html#module-sqlalchemy.pool) to your Cloud SQL database, enabling successful connections from your application and following industry best practices.\n",
    "\n",
    "To create a `MSSQLEngine` using `MSSQLEngine.from_instance()` you need to provide only 4 things:\n",
    "\n",
    "1. `project_id` : Project ID of the Google Cloud Project where the Cloud SQL instance is located.\n",
    "1. `region` : Region where the Cloud SQL instance is located.\n",
    "1. `instance` : The name of the Cloud SQL instance.\n",
    "1. `database` : The name of the database to connect to on the Cloud SQL instance.\n",
    "1. `user` : Database user to use for built-in database authentication and login.\n",
    "1. `password` : Database password to use for built-in database authentication and login."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from langchain_google_cloud_sql_mssql import MSSQLEngine\n",
    "\n",
    "engine = MSSQLEngine.from_instance(\n",
    "    project_id=PROJECT_ID,\n",
    "    region=REGION,\n",
    "    instance=INSTANCE,\n",
    "    database=DATABASE,\n",
    "    user=DB_USER,\n",
    "    password=DB_PASS,\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Initialize a table\n",
    "\n",
    "Initialize a table of default schema via `MSSQLEngine.init_document_table(<table_name>)`. Table Columns:\n",
    "\n",
    "- page_content (type: text)\n",
    "- langchain_metadata (type: JSON)\n",
    "\n",
    "`overwrite_existing=True` flag means the newly initialized table will replace any existing table of the same name."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "engine.init_document_table(TABLE_NAME, overwrite_existing=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Save documents\n",
    "\n",
    "Save langchain documents with `MSSQLDocumentSaver.add_documents(<documents>)`. To initialize `MSSQLDocumentSaver` class you need to provide 2 things:\n",
    "\n",
    "1. `engine` - An instance of a `MSSQLEngine` engine.\n",
    "2. `table_name` - The name of the table within the Cloud SQL database to store langchain documents."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from langchain_core.documents import Document\n",
    "from langchain_google_cloud_sql_mssql import MSSQLDocumentSaver\n",
    "\n",
    "test_docs = [\n",
    "    Document(\n",
    "        page_content=\"Apple Granny Smith 150 0.99 1\",\n",
    "        metadata={\"fruit_id\": 1},\n",
    "    ),\n",
    "    Document(\n",
    "        page_content=\"Banana Cavendish 200 0.59 0\",\n",
    "        metadata={\"fruit_id\": 2},\n",
    "    ),\n",
    "    Document(\n",
    "        page_content=\"Orange Navel 80 1.29 1\",\n",
    "        metadata={\"fruit_id\": 3},\n",
    "    ),\n",
    "]\n",
    "saver = MSSQLDocumentSaver(engine=engine, table_name=TABLE_NAME)\n",
    "saver.add_documents(test_docs)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Load documents"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Load langchain documents with `MSSQLLoader.load()` or `MSSQLLoader.lazy_load()`. `lazy_load` returns a generator that only queries database during the iteration. To initialize `MSSQLDocumentSaver` class you need to provide:\n",
    "\n",
    "1. `engine` - An instance of a `MSSQLEngine` engine.\n",
    "2. `table_name` - The name of the table within the Cloud SQL database to store langchain documents."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from langchain_google_cloud_sql_mssql import MSSQLLoader\n",
    "\n",
    "loader = MSSQLLoader(engine=engine, table_name=TABLE_NAME)\n",
    "docs = loader.lazy_load()\n",
    "for doc in docs:\n",
    "    print(\"Loaded documents:\", doc)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Load documents via query"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Other than loading documents from a table, we can also choose to load documents from a view generated from a SQL query. For example:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from langchain_google_cloud_sql_mssql import MSSQLLoader\n",
    "\n",
    "loader = MSSQLLoader(\n",
    "    engine=engine,\n",
    "    query=f\"select * from \\\"{TABLE_NAME}\\\" where JSON_VALUE(langchain_metadata, '$.fruit_id') = 1;\",\n",
    ")\n",
    "onedoc = loader.load()\n",
    "onedoc"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The view generated from SQL query can have different schema than default table. In such cases, the behavior of MSSQLLoader is the same as loading from table with non-default schema. Please refer to section [Load documents with customized document page content & metadata](#Load-documents-with-customized-document-page-content-&-metadata)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Delete documents"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Delete a list of langchain documents from MSSQL table with `MSSQLDocumentSaver.delete(<documents>)`.\n",
    "\n",
    "For table with default schema (page_content, langchain_metadata), the deletion criteria is:\n",
    "\n",
    "A `row` should be deleted if there exists a `document` in the list, such that\n",
    "\n",
    "- `document.page_content` equals `row[page_content]`\n",
    "- `document.metadata` equals `row[langchain_metadata]`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from langchain_google_cloud_sql_mssql import MSSQLLoader\n",
    "\n",
    "loader = MSSQLLoader(engine=engine, table_name=TABLE_NAME)\n",
    "docs = loader.load()\n",
    "print(\"Documents before delete:\", docs)\n",
    "saver.delete(onedoc)\n",
    "print(\"Documents after delete:\", loader.load())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Advanced Usage"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Load documents with customized document page content & metadata"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First we prepare an example table with non-default schema, and populate it with some arbitary data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import sqlalchemy\n",
    "\n",
    "with engine.connect() as conn:\n",
    "    conn.execute(sqlalchemy.text(f'DROP TABLE IF EXISTS \"{TABLE_NAME}\"'))\n",
    "    conn.commit()\n",
    "    conn.execute(\n",
    "        sqlalchemy.text(\n",
    "            f\"\"\"\n",
    "            IF NOT EXISTS (SELECT * FROM sys.objects WHERE object_id = OBJECT_ID(N'[dbo].[{TABLE_NAME}]') AND type in (N'U'))\n",
    "                BEGIN\n",
    "                    CREATE TABLE [dbo].[{TABLE_NAME}](\n",
    "                        fruit_id INT IDENTITY(1,1) PRIMARY KEY,\n",
    "                        fruit_name VARCHAR(100) NOT NULL,\n",
    "                        variety VARCHAR(50),\n",
    "                        quantity_in_stock INT NOT NULL,\n",
    "                        price_per_unit DECIMAL(6,2) NOT NULL,\n",
    "                        organic BIT NOT NULL\n",
    "                    )\n",
    "                END\n",
    "            \"\"\"\n",
    "        )\n",
    "    )\n",
    "    conn.execute(\n",
    "        sqlalchemy.text(\n",
    "            f\"\"\"\n",
    "            INSERT INTO \"{TABLE_NAME}\" (fruit_name, variety, quantity_in_stock, price_per_unit, organic)\n",
    "            VALUES\n",
    "                ('Apple', 'Granny Smith', 150, 0.99, 1),\n",
    "                ('Banana', 'Cavendish', 200, 0.59, 0),\n",
    "                ('Orange', 'Navel', 80, 1.29, 1);\n",
    "            \"\"\"\n",
    "        )\n",
    "    )\n",
    "    conn.commit()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "If we still load langchain documents with default parameters of `MSSQLLoader` from this example table, the `page_content` of loaded documents will be the first column of the table, and `metadata` will be consisting of key-value pairs of all the other columns."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "loader = MSSQLLoader(\n",
    "    engine=engine,\n",
    "    table_name=TABLE_NAME,\n",
    ")\n",
    "loader.load()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can specify the content and metadata we want to load by setting the `content_columns` and `metadata_columns` when initializing the `MSSQLLoader`.\n",
    "\n",
    "1. `content_columns`: The columns to write into the `page_content` of the document.\n",
    "2. `metadata_columns`: The columns to write into the `metadata` of the document.\n",
    "\n",
    "For example here, the values of columns in `content_columns` will be joined together into a space-separated string, as `page_content` of loaded documents, and `metadata` of loaded documents will only contain key-value pairs of columns specified in `metadata_columns`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "loader = MSSQLLoader(\n",
    "    engine=engine,\n",
    "    table_name=TABLE_NAME,\n",
    "    content_columns=[\n",
    "        \"variety\",\n",
    "        \"quantity_in_stock\",\n",
    "        \"price_per_unit\",\n",
    "        \"organic\",\n",
    "    ],\n",
    "    metadata_columns=[\"fruit_id\", \"fruit_name\"],\n",
    ")\n",
    "loader.load()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Save document with customized page content & metadata"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In order to save langchain document into table with customized metadata fields. We need first create such a table via `MSSQLEngine.init_document_table()`, and specify the list of `metadata_columns` we want it to have. In this example, the created table will have table columns:\n",
    "\n",
    "- description (type: text): for storing fruit description.\n",
    "- fruit_name (type text): for storing fruit name.\n",
    "- organic (type tinyint(1)): to tell if the fruit is organic.\n",
    "- other_metadata (type: JSON): for storing other metadata information of the fruit.\n",
    "\n",
    "We can use the following parameters with `MSSQLEngine.init_document_table()` to create the table:\n",
    "\n",
    "1. `table_name`: The name of the table within the Cloud SQL database to store langchain documents.\n",
    "2. `metadata_columns`: A list of `sqlalchemy.Column` indicating the list of metadata columns we need.\n",
    "3. `content_column`: The name of column to store `page_content` of langchain document. Default: `page_content`.\n",
    "4. `metadata_json_column`: The name of JSON column to store extra `metadata` of langchain document. Default: `langchain_metadata`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "engine.init_document_table(\n",
    "    TABLE_NAME,\n",
    "    metadata_columns=[\n",
    "        sqlalchemy.Column(\n",
    "            \"fruit_name\",\n",
    "            sqlalchemy.UnicodeText,\n",
    "            primary_key=False,\n",
    "            nullable=True,\n",
    "        ),\n",
    "        sqlalchemy.Column(\n",
    "            \"organic\",\n",
    "            sqlalchemy.Boolean,\n",
    "            primary_key=False,\n",
    "            nullable=True,\n",
    "        ),\n",
    "    ],\n",
    "    content_column=\"description\",\n",
    "    metadata_json_column=\"other_metadata\",\n",
    "    overwrite_existing=True,\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Save documents with `MSSQLDocumentSaver.add_documents(<documents>)`. As you can see in this example, \n",
    "\n",
    "- `document.page_content` will be saved into `description` column.\n",
    "- `document.metadata.fruit_name` will be saved into `fruit_name` column.\n",
    "- `document.metadata.organic` will be saved into `organic` column.\n",
    "- `document.metadata.fruit_id` will be saved into `other_metadata` column in JSON format."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_docs = [\n",
    "    Document(\n",
    "        page_content=\"Granny Smith 150 0.99\",\n",
    "        metadata={\"fruit_id\": 1, \"fruit_name\": \"Apple\", \"organic\": 1},\n",
    "    ),\n",
    "]\n",
    "saver = MSSQLDocumentSaver(\n",
    "    engine=engine,\n",
    "    table_name=TABLE_NAME,\n",
    "    content_column=\"description\",\n",
    "    metadata_json_column=\"other_metadata\",\n",
    ")\n",
    "saver.add_documents(test_docs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "with engine.connect() as conn:\n",
    "    result = conn.execute(sqlalchemy.text(f'select * from \"{TABLE_NAME}\";'))\n",
    "    print(result.keys())\n",
    "    print(result.fetchall())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Delete documents with customized page content & metadata"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can also delete documents from table with customized metadata columns via `MSSQLDocumentSaver.delete(<documents>)`. The deletion criteria is:\n",
    "\n",
    "A `row` should be deleted if there exists a `document` in the list, such that\n",
    "\n",
    "- `document.page_content` equals `row[page_content]`\n",
    "- For every metadata field `k` in `document.metadata`\n",
    "    - `document.metadata[k]` equals `row[k]` or `document.metadata[k]` equals `row[langchain_metadata][k]`\n",
    "- There no extra metadata field presents in `row` but not in `document.metadata`.\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "loader = MSSQLLoader(engine=engine, table_name=TABLE_NAME)\n",
    "docs = loader.load()\n",
    "print(\"Documents before delete:\", docs)\n",
    "saver.delete(docs)\n",
    "print(\"Documents after delete:\", loader.load())"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}