Added Redshift support (#2)

DESCRIPTION

@@ -1,6 +1,6 @@
 Package: snowquery
 Title: Query 'Snowflake' Databases with 'SQL'
-Version: 1.0.0
+Version: 1.1.0
 Authors@R: 
     person("Dani", "Mermelstein", , "dmermelstein@hey.com", role = c("aut", "cre", "cph"))
 Maintainer: Dani Mermelstein <dmermelstein@hey.com>

NEWS.md

@@ -1,3 +1,7 @@
+# snowquery 1.1.0
+
+* Added Redshift support
+
 # snowquery 1.0.0
 
 * Formalized structure for database credential file

R/queryDB.R

@@ -1,9 +1,9 @@
-#' @title Query a Snowflake or Postgres database
-#' @description Run a SQL query on a Snowflake or Postgres database (requires a `~/snowquery_creds.yaml` file)
+#' @title Query a database
+#' @description Run a SQL query on a Snowflake, Redshift or Postgres database and return the results as a data frame. See the [snowquery README](https://github.com/mermelstein/snowquery#credentials) for more information on how to pass in your credentials.
 #'
 #' @param query A string of the SQL query to execute
 #' @param conn_name The name of the connection to use in snowquery_creds.yaml (e.g. "my_snowflake_dwh")
-#' @param db_type The type of database to connect to (e.g. "snowflake" or "postgres")
+#' @param db_type The type of database to connect to (e.g. "snowflake", "redshift" or "postgres")
 #' @param username The username to use for authentication
 #' @param password The password to use for authentication
 #' @param host The hostname or IP address of the database server
@@ -53,6 +53,7 @@ queryDB <- function(
   warehouse = NULL,
   account = NULL,
   role = NULL,
+  sslmode = NULL,
   timeout = 15)
 {
 
@@ -64,6 +65,9 @@ queryDB <- function(
     }
   }
 
+  db_type_error_message <- paste0("Invalid db_type '", db_type, "'. \n",
+  "Snowquery currently only supports the following database types: 'snowflake', 'redshift' and 'postgres'")
+
   # pull in the credential file
   snowquery_creds_filepath <- '~/snowquery_creds.yaml'
   snowquery_creds <- yaml::read_yaml(snowquery_creds_filepath,  fileEncoding = "UTF-8")
@@ -76,10 +80,10 @@ queryDB <- function(
 
   # Check if db_type is provided
   if (missing(db_type) || is.null(db_type)) {
-    stop(paste0("db_type is missing.\n",
-    "Please provide a database type to queryDB(). Expected values are 'snowflake' or 'postgres'.\n",
+    stop(paste0("db_type is missing for the '", conn_name, "' connection.\n",
+    "Please provide a database type to queryDB(). Expected values are 'snowflake', 'redshift' or 'postgres'.\n",
     "You can add a db_type variable to the '", conn_name, "' connection in the snowquery_creds.yaml file or pass it in manually:\n",
-    "For example: queryDB('SELECT * FROM my_table', conn_name = 'snowflake', db_type = 'snowflake')"))
+    "For example: queryDB('SELECT * FROM my_table', conn_name = '", conn_name, "', db_type = 'snowflake')"))
   }
 
   if (tolower(db_type) == "snowflake") {
@@ -98,13 +102,9 @@ queryDB <- function(
       # Import the snowflake.connector module from the snowflake-connector-python package
       snowflake <- import("snowflake.connector")
     }, error = function(e) {
-      # stop(paste0("Failed to import the snowflake.connector module. Please make sure it is installed and accessible from your environment. \n",
-      # "Try running the following command from your terminal or command line:\n\n",
-      # "pip install 'snowflake-connector-python[pandas]'\n\n",
-      # "Error message: ", e$message))
-      stop(paste0("Failed to find the python executable. Please make sure python 3 is installed and accessible from your environment.\n",
-      "You can download Python 3 from https://www.python.org/downloads/ or via Homebrew if on MacOS. \n",
-      "After installing Python 3, make sure it is added to your system PATH. \n",
+      stop(paste0("Failed to import the snowflake.connector module. Please make sure it is installed and accessible from your environment. \n",
+      "Try running the following command from your terminal or command line:\n\n",
+      "pip install 'snowflake-connector-python[pandas]'\n\n",
       "Error message: ", e$message))
     })
     username_ <- check_null(username, check_null(conn_details$user, NULL))
@@ -129,7 +129,7 @@ queryDB <- function(
       "The following credential variable(s) are missing: ", paste(missing_vars, collapse = ", "), ".\n",
       "Please pass in credentials to queryDB() or add them to the snowquery_creds.yaml file."))
     } else {
-      # Use available credentials to build connection string
+      # Use credentials to build connection string
       con <- snowflake$connect(
         user = username_,
         password = password_,
@@ -151,13 +151,14 @@ queryDB <- function(
     # Return the query results
     return(df)
 
-  } else if (tolower(db_type) == "postgres") {
+  } else if (tolower(db_type) %in% c("postgres", "redshift")) {
     # Check if credentials are provided manually by user
     database_ <- check_null(database, check_null(conn_details$database, NULL))
     username_ <- check_null(username, check_null(conn_details$username, NULL))
     password_ <- check_null(password, check_null(conn_details$password, NULL))
     port_ <- check_null(port, check_null(conn_details$port, NULL))
     host_ <- check_null(host, check_null(conn_details$host, NULL))
+    sslmode_ <- check_null(sslmode, check_null(conn_details$sslmode, NULL))
     # Check if any credentials are missing
     if (is.null(username_) || is.null(password_) || is.null(host_) || is.null(database_) || is.null(port_)) {
       # Get the names of the missing credential variables
@@ -168,17 +169,25 @@ queryDB <- function(
       if (is.null(port_)) missing_vars <- c(missing_vars, "port")
       if (is.null(host_)) missing_vars <- c(missing_vars, "host")
       # Error message if credentials are missing
-      stop(paste0("Missing credentials for the postgres connection. \n",
+      stop(paste0("Missing credentials for the ", db_type, " connection. \n",
       "The following credential variable(s) are missing: ", paste(missing_vars, collapse = ", "), ".\n",
       "Please pass in credentials to queryDB() or add them to the snowquery_creds.yaml file."))
     } else {
-      # Use available credentials to build connection string
-      con <- DBI::dbConnect(RPostgres::Postgres(),
+      if (tolower(db_type) == "postgres") {
+        driver_type <- RPostgres::Postgres()
+      } else if (tolower(db_type) == "redshift") {
+        driver_type <- RPostgres::Redshift()
+      } else {
+        stop(db_type_error_message)
+      }
+      # Use credentials to build connection string
+      con <- DBI::dbConnect(driver_type,
           dbname = database_,
           host = host_,
           port = port_,
           user = username_,
           password = password_,
+          sslmode = sslmode_,
           connect_timeout = timeout # seconds
         )
     }
@@ -193,7 +202,6 @@ queryDB <- function(
     return(df)
 
   } else {
-    stop(paste0("Invalid db_type '", db_type, "'. \n",
-    "Snowquery currently only supports 'snowflake' and 'postgres' database types."))
+    stop(db_type_error_message)
   }
 }

R/snowquery-package.R

@@ -1,15 +1,15 @@
 #' @keywords internal
 "_PACKAGE"
 
-#' This package provides functions for querying Snowflake databases using R.
-#' To use this package, you will need to provide your Snowflake credentials in a
+#' This package provides functions for querying Snowflake, Redshift and Postgres databases using R.
+#' To use this package, you will need to provide your database credentials in a
 #' YAML file called `snowquery_creds.yaml`. The file should be located in the
 #' root directory of your R project and should have the following format:
 #'
 #' ```yaml
 #' ---
 #' my_snowflake_dwh:
-#'    db_type: 'snowflake'
+#'    db_type: 'snowflake' # or 'redshift' or 'postgres'
 #'    account: 'your_account_name'
 #'    warehouse: 'your_warehouse_name'
 #'    database: 'your_database_name'
@@ -22,6 +22,6 @@
 #' This follows a named connection format, where you can have multiple named connections in the same file. 
 #' For example you might have a `my_snowflake_dwh` connection and a `my_snowflake_admin` connection, each with their own credentials.
 #' 
-#' Replace the values in the YAML file with your own Snowflake credentials.
+#' Replace the values in the YAML file with your own credentials.
 #' Once you have created the `snowquery_creds.yaml` file, you can use the
-#' `queryDB()` function to query your Snowflake database.
+#' `queryDB()` function to query your database.

README.md

@@ -6,14 +6,9 @@
 
 ## Overview
 
-Run SQL queries on a Snowflake instance from an R script. This will be similar to how you might be using DBI or odbc to query a postgres or Redshift database, but because Snowflake's driver requires a ton of fiddling in order to make it work for R this is an alternate solution.
+Run SQL queries on Snowflake, Redshift, or a postgres database from an R script. 
 
-This sums up the current experience of running SQL against Snowflake from:
-
-  - python: good ✅
-  - R: bad ❌
-
-That's why the `snowquery` package takes the [Snowflake python connector](https://docs.snowflake.com/en/developer-guide/python-connector/python-connector-pandas) and leverages it in the background to run queries from R.
+This package is designed to make it easy to run SQL queries from R. It is designed to work with Snowflake, Redshift, or a postgres database. It is not designed to work with other databases, but it could be extended to do so.
 
 ### Installation
 
@@ -26,6 +21,23 @@ install.packages("snowquery")
 devtools::install_github("mermelstein/snowquery")
 ```
 
+### Redshift notes
+
+Redshift is currently only available on the development version of this package. See [installation instructions](#installation) above.
+
+When connecting to a Redshift DWH you might need to specify an SSL connection. You can do this with a `sslmode='require'` connection variable or by passing that to the `queryDB()` function directly.
+
+### Snowflake notes
+
+Because Snowflake's driver requires a ton of fiddling in order to make it work for R. It sucks. A lot.
+
+To sum up the current experience of running SQL against Snowflake from:
+
+  - python: good ✅
+  - R: bad ❌
+
+That's why the `snowquery` package takes the [Snowflake python connector](https://docs.snowflake.com/en/developer-guide/python-connector/python-connector-pandas) and leverages it in the background to run queries from R.
+
 ### Documentation
 
 For more information on using `snowquery`, please see the [package website](https://snowquery.org).
@@ -44,7 +56,9 @@ If you need to install the Snowflake python connector, you can do that with the
 pip install "snowflake-connector-python[pandas]"
 ```
 
-You will also need to have your Snowflake credentials in a YAML file called `snowquery_creds.yaml`. The file should be located in the
+### Credentials
+
+You will need to have your database credentials in a YAML file called `snowquery_creds.yaml`. The file should be located in the
 root directory of your machine and should have the following format:
 
 ```yaml
@@ -62,15 +76,15 @@ my_snowflake_dwh:
 
 This follows a named connection format, where you can have multiple named connections in the same file. For example you might have a `my_snowflake_dwh` connection and a `my_snowflake_admin` connection, each with their own credentials.
 
-The main function of this package looks for that file at this location: `~/snowquery_creds.yaml`. **If it is in any other location it will not work.** If the package cannot locate the file you will receive an error like: `cannot open file '/expected/path/to/file/snowquery_creds.yaml': No such file or directory`.
+This package looks for the credential file at this location: `~/snowquery_creds.yaml`. **If it is in any other location it will not work.** If the package cannot locate the file you will receive an error like: `cannot open file '/expected/path/to/file/snowquery_creds.yaml': No such file or directory`. You can manually pass credentials to the `queryDB()` function but it is recommended to use the YAML file.
 
 You are now ready to query away!
 
 ### Usage
 
 Load this library in your R environment with `library(snowquery)`.
 
-There is one function you need: `queryDB()`. It will take a string parameter and run that as a SQL query.
+There is one function you need: `queryDB()`. It will take a SQL query as a string parameter and run it on the db.
 
 For example: