Merge pull request OHDSI#321 from james-cockayne/feature/duckdb_configuration

Feature/Duckdb configuration

Author: schuemie

R/Connect.R

@@ -842,10 +842,24 @@ connectUsingDbi <- function(dbiConnectionDetails) {
 connectDuckdb <- function(connectionDetails) {
   inform("Connecting using DuckDB driver")
   ensure_installed("duckdb")
+  
+  # Extract config from extraSettings if provided
+  duckdbConfig <- NULL
+  if (!is.null(connectionDetails$extraSettings) && !is.null(connectionDetails$extraSettings$config)) {
+    duckdbConfig <- connectionDetails$extraSettings$config
+  }
+  
+  # Create DuckDB driver with config if provided
+  drv <- if (is.null(duckdbConfig)) {
+    duckdb::duckdb()
+  } else {
+    duckdb::duckdb(config = duckdbConfig)
+  }
+  
   connection <- connectUsingDbi(
     createDbiConnectionDetails(
       dbms = connectionDetails$dbms,
-      drv = duckdb::duckdb(),
+      drv = drv,
       dbdir = connectionDetails$server(),
       bigint = "integer64"
     )

man-roxygen/DefaultConnectionDetails.R

@@ -107,6 +107,14 @@
 #' 
 #'  - `server`. The path to the SQLIte file.
 #' 
+#' DuckDB:
+#' 
+#'  - `server`. The path to the DuckDB file.
+#'  - `extraSettings`. Additional settings for DuckDB. For DuckDB specifically, if 
+#'         `extraSettings$config` is provided, it will be passed to the 
+#'         `duckdb::duckdb()` constructor. For example: 
+#'         `extraSettings = list(config = list(memory_limit = "8GB", preserve_insertion_order = "false"))`
+#' 
 #' Spark / Databricks:
 #' 
 #' Currently both JDBC and ODBC connections are supported for Spark. Set the 

test_duckdb_config.R

@@ -0,0 +1,71 @@
+# Test script for DuckDB config functionality
+library(DatabaseConnector)
+library(duckdb)
+
+# Create a temporary DuckDB file
+temp_db_file <- tempfile(fileext = ".duckdb")
+
+# Test 1: Connection without config (existing behavior)
+cat("Test 1: Connection without config\n")
+connectionDetails1 <- createConnectionDetails(
+  dbms = "duckdb",
+  server = temp_db_file
+)
+
+tryCatch({
+  connection1 <- connect(connectionDetails1)
+  cat("✓ Connection without config successful\n")
+  disconnect(connection1)
+}, error = function(e) {
+  cat("✗ Connection without config failed:", e$message, "\n")
+})
+
+# Test 2: Connection with config (new functionality)
+cat("\nTest 2: Connection with config\n")
+connectionDetails2 <- createConnectionDetails(
+  dbms = "duckdb",
+  server = temp_db_file,
+  extraSettings = list(
+    config = list(
+      memory_limit = "8GB",
+      preserve_insertion_order = "false"
+    )
+  )
+)
+
+tryCatch({
+  connection2 <- connect(connectionDetails2)
+  cat("✓ Connection with config successful\n")
+  
+  # Verify that the config was applied by checking memory_limit setting
+  result <- querySql(connection2, "SELECT current_setting('memory_limit') as memory_limit")
+  cat("Memory limit setting:", result$MEMORY_LIMIT, "\n")
+  
+  disconnect(connection2)
+}, error = function(e) {
+  cat("✗ Connection with config failed:", e$message, "\n")
+})
+
+# Test 3: Connection with invalid config (error handling)
+cat("\nTest 3: Connection with invalid config\n")
+connectionDetails3 <- createConnectionDetails(
+  dbms = "duckdb",
+  server = temp_db_file,
+  extraSettings = list(
+    config = list(
+      invalid_setting = "invalid_value"
+    )
+  )
+)
+
+tryCatch({
+  connection3 <- connect(connectionDetails3)
+  cat("✗ Connection with invalid config should have failed\n")
+  disconnect(connection3)
+}, error = function(e) {
+  cat("✓ Connection with invalid config correctly failed:", e$message, "\n")
+})
+
+# Cleanup
+unlink(temp_db_file)
+cat("\nTest completed\n")