feat: Use duckdb or sqlite based on first available

Author: gadenbuie

pkg-r/R/DataSource.R

@@ -98,24 +98,30 @@ DataSource <- R6::R6Class(
 #' database table. See [DBISource] for the full description of available
 #' methods.
 #'
-#' Use the `engine` parameter to choose between "duckdb" (default) or "sqlite",
-#' or set the global option `querychat.DataFrameSource.engine to choose the
-#' default engine for all DataFrameSource instances.
+#' By default, DataFrameSource uses the first available engine from duckdb
+#' (checked first) or RSQLite. You can explicitly set the `engine` parameter to
+#' choose between "duckdb" or "sqlite", or set the global option
+#' `querychat.DataFrameSource.engine` to choose the default engine for all
+#' DataFrameSource instances. At least one of these packages must be installed.
 #'
 #' @export
 #' @examples
 #' \dontrun{
-#' # Create a data frame source
+#' # Create a data frame source (uses first available: duckdb or sqlite)
 #' df_source <- DataFrameSource$new(mtcars, "mtcars")
 #'
 #' # Get database type
-#' df_source$get_db_type()  # Returns "DuckDB"
+#' df_source$get_db_type()  # Returns "DuckDB" or "SQLite"
 #'
 #' # Execute a query
 #' result <- df_source$execute_query("SELECT * FROM mtcars WHERE mpg > 25")
 #'
+#' # Explicitly choose an engine
+#' df_sqlite <- DataFrameSource$new(mtcars, "mtcars", engine = "sqlite")
+#'
 #' # Clean up when done
 #' df_source$cleanup()
+#' df_sqlite$cleanup()
 #' }
 DataFrameSource <- R6::R6Class(
   "DataFrameSource",
@@ -131,6 +137,10 @@ DataFrameSource <- R6::R6Class(
     #' @param table_name Name to use for the table in SQL queries. Must be a
     #'   valid table name (start with letter, contain only letters, numbers,
     #'   and underscores)
+    #' @param engine Database engine to use: "duckdb" or "sqlite". Set the
+    #'   global option `querychat.DataFrameSource.engine` to specify the default
+    #'   engine for all instances. If NULL (default), uses the first available
+    #'   engine from duckdb or RSQLite (in that order).
     #' @return A new DataFrameSource object
     #' @examples
     #' \dontrun{
@@ -139,11 +149,12 @@ DataFrameSource <- R6::R6Class(
     initialize = function(
       df,
       table_name,
-      engine = getOption("querychat.DataFrameSource.engine", "duckdb")
+      engine = getOption("querychat.DataFrameSource.engine", NULL)
     ) {
       check_data_frame(df)
       check_sql_table_name(table_name)
 
+      engine <- engine %||% get_default_dataframe_engine()
       engine <- tolower(engine)
       arg_match(engine, c("duckdb", "sqlite"))
 
@@ -341,6 +352,22 @@ is_data_source <- function(x) {
 }
 
 
+get_default_dataframe_engine <- function() {
+  if (is_installed("duckdb")) {
+    return("duckdb")
+  }
+  if (is_installed("RSQLite")) {
+    return("sqlite")
+  }
+  cli::cli_abort(c(
+    "No compatible database engine installed for DataFrameSource",
+    "i" = "Install either {.pkg duckdb} or {.pkg RSQLite}:",
+    " " = "{.run install.packages(\"duckdb\")}",
+    " " = "{.run install.packages(\"RSQLite\")}"
+  ))
+}
+
+
 get_schema_impl <- function(conn, table_name, categorical_threshold = 20) {
   # Get column information
   columns <- DBI::dbListFields(conn, table_name)

pkg-r/R/querychat-package.R

@@ -84,6 +84,5 @@ release_bullets <- function() {
 }
 
 suppress_rcmdcheck <- function() {
-  duckdb::duckdb
   S7::S7_class
 }

pkg-r/man/DataFrameSource.Rd

@@ -71,6 +71,13 @@ local_sqlite_connection <- function(
   list(conn = conn, path = temp_db)
 }
 
+# Skip test if no DataFrameSource engine is available
+skip_if_no_dataframe_engine <- function() {
+  if (!rlang::is_installed("duckdb") && !rlang::is_installed("RSQLite")) {
+    skip("Neither duckdb nor RSQLite is installed")
+  }
+}
+
 # Create a DataFrameSource with automatic cleanup
 local_data_frame_source <- function(
   data,

pkg-r/tests/testthat/helper-fixtures.R

@@ -30,6 +30,8 @@ describe("DataSource base class", {
 })
 
 describe("DataFrameSource$new()", {
+  skip_if_no_dataframe_engine()
+
   it("creates proper R6 object for DataFrameSource", {
     test_df <- new_test_df()
 
@@ -268,6 +270,8 @@ describe("DBISource$new()", {
 
 describe("DataSource$get_schema()", {
   it("returns proper schema for DataFrameSource", {
+    skip_if_no_dataframe_engine()
+
     df_source <- local_data_frame_source(new_mixed_types_df())
 
     schema <- df_source$get_schema()
@@ -295,6 +299,8 @@ describe("DataSource$get_schema()", {
   })
 
   it("correctly reports min/max values for numeric columns", {
+    skip_if_no_dataframe_engine()
+
     df_source <- local_data_frame_source(new_metrics_df())
 
     schema <- df_source$get_schema()
@@ -306,9 +312,12 @@ describe("DataSource$get_schema()", {
 })
 
 describe("DataSource$get_db_type()", {
-  it("returns DuckDB for DataFrameSource", {
+  it("returns correct database type for DataFrameSource", {
+    skip_if_no_dataframe_engine()
+
     df_source <- local_data_frame_source(new_test_df())
-    expect_equal(df_source$get_db_type(), "DuckDB")
+    db_type <- df_source$get_db_type()
+    expect_true(db_type %in% c("DuckDB", "SQLite"))
   })
 
   it("returns correct type for SQLite connections", {
@@ -325,6 +334,9 @@ describe("DataSource$get_data()", {
   test_df <- new_test_df()
 
   it("returns all data for both DataFrameSource and DBISource", {
+    skip_if_no_dataframe_engine()
+    skip_if_not_installed("RSQLite")
+
     df_source <- local_data_frame_source(test_df)
 
     result <- df_source$get_data()
@@ -343,6 +355,9 @@ describe("DataSource$get_data()", {
 })
 
 describe("DataSource$execute_query()", {
+  skip_if_no_dataframe_engine()
+  skip_if_not_installed("RSQLite")
+
   test_df <- new_test_df(rows = 4)
   df_source <- local_data_frame_source(test_df)
   db <- local_sqlite_connection(test_df)
@@ -589,6 +604,8 @@ describe("DBISource$test_query()", {
 })
 
 describe("DataFrameSource$test_query()", {
+  skip_if_no_dataframe_engine()
+
   test_df <- new_users_df()
   df_source <- local_data_frame_source(test_df, "test_table")
 

pkg-r/tests/testthat/test-DataSource.R

@@ -1,4 +1,6 @@
 describe("QueryChat$new()", {
+  skip_if_no_dataframe_engine()
+
   it("automatically converts data.frame to DataFrameSource", {
     qc <- QueryChat$new(
       data_source = new_test_df(),
@@ -231,6 +233,8 @@ describe("QueryChat$system_prompt", {
 })
 
 describe("QueryChat$data_source", {
+  skip_if_no_dataframe_engine()
+
   it("returns the data source object", {
     test_df <- new_test_df()
     qc <- QueryChat$new(test_df, greeting = "Test")
@@ -470,6 +474,7 @@ test_that("QueryChat$server() errors when called outside Shiny context", {
 })
 
 describe("querychat()", {
+  skip_if_no_dataframe_engine()
   withr::local_envvar(OPENAI_API_KEY = "boop")
 
   it("creates a QueryChat object", {
@@ -597,6 +602,8 @@ describe("QueryChat$console()", {
 })
 
 describe("normalize_data_source()", {
+  skip_if_no_dataframe_engine()
+
   it("returns DataSource objects unchanged", {
     test_df <- new_test_df()
     df_source <- DataFrameSource$new(test_df, "test_df")

pkg-r/tests/testthat/test-QueryChat.R

@@ -1,4 +1,6 @@
 describe("QueryChatSystemPrompt$new()", {
+  skip_if_no_dataframe_engine()
+
   it("initializes with string template", {
     df <- new_test_df()
     ds <- DataFrameSource$new(df, "test_table")

pkg-r/tests/testthat/test-QueryChatSystemPrompt.R

@@ -1,4 +1,6 @@
 test_that("tool_update_dashboard() checks inputs", {
+  skip_if_no_dataframe_engine()
+
   expect_snapshot(error = TRUE, tool_update_dashboard("foo"))
 
   df_source <- local_data_frame_source(new_test_df())
@@ -88,6 +90,8 @@ describe("querychat_tool_details_option()", {
 })
 
 describe("querychat_tool_result()", {
+  skip_if_no_dataframe_engine()
+
   it("returns successful result for valid query action", {
     df_source <- local_data_frame_source(new_test_df())
 
@@ -283,6 +287,8 @@ describe("querychat_tool_result()", {
 })
 
 describe("tool_query()", {
+  skip_if_no_dataframe_engine()
+
   it("returns an ellmer tool object", {
     df_source <- local_data_frame_source(new_test_df())
     tool <- tool_query(df_source)
@@ -312,6 +318,8 @@ describe("tool_query()", {
 })
 
 describe("tool_update_dashboard()", {
+  skip_if_no_dataframe_engine()
+
   it("returns an ellmer tool object", {
     df_source <- local_data_frame_source(new_test_df())
 
@@ -370,6 +378,8 @@ describe("tool_reset_dashboard()", {
 })
 
 describe("tool_update_dashboard_impl()", {
+  skip_if_no_dataframe_engine()
+
   it("returns a function", {
     df_source <- local_data_frame_source(new_test_df())
     current_query <- shiny::reactiveVal("SELECT * FROM test_table")