Adding unit tests for BQ

OHDSI · Oct 22, 2023 · baa049f · baa049f
1 parent bd94697
commit baa049f
Show file tree

Hide file tree

Showing 3 changed files with 38 additions and 9 deletions.
diff --git a/.github/workflows/R_CMD_check_Hades.yaml b/.github/workflows/R_CMD_check_Hades.yaml
@@ -58,6 +58,10 @@ jobs:
       CDM5_SPARK_CONNECTION_STRING: ${{ secrets.CDM5_SPARK_CONNECTION_STRING }}
       CDM5_SPARK_CDM_SCHEMA: ${{ secrets.CDM5_SPARK_CDM_SCHEMA }}
       CDM5_SPARK_OHDSI_SCHEMA: ${{ secrets.CDM5_SPARK_OHDSI_SCHEMA }}
+      CDM_BIG_QUERY_CONNECTION_STRING: ${{ secrets.CDM_BIG_QUERY_CONNECTION_STRING }}
+      CDM_BIG_QUERY_KEY_FILE: ${{ secrets.CDM_BIG_QUERY_KEY_FILE }}
+      CDM_BIG_QUERY_CDM_SCHEMA: ${{ secrets.CDM_BIG_QUERY_CDM_SCHEMA }}
+      CDM_BIG_QUERY_OHDSI_SCHEMA: ${{ secrets.CDM_BIG_QUERY_OHDSI_SCHEMA }}
 
     steps:
       - uses: actions/checkout@v3

diff --git a/tests/testthat/setup.R b/tests/testthat/setup.R
@@ -1,4 +1,5 @@
 library(DatabaseConnector)
+
 # Download the JDBC drivers used in the tests ----------------------------------
 if (Sys.getenv("DONT_DOWNLOAD_JDBC_DRIVERS", "") != "TRUE") {
   oldJarFolder <- Sys.getenv("DATABASECONNECTOR_JAR_FOLDER")
@@ -10,6 +11,7 @@ if (Sys.getenv("DONT_DOWNLOAD_JDBC_DRIVERS", "") != "TRUE") {
   downloadJdbcDrivers("redshift")
   downloadJdbcDrivers("spark")
   downloadJdbcDrivers("snowflake")
+  downloadJdbcDrivers("bigquery")
 
   if (testthat::is_testing()) {
     withr::defer({
@@ -21,6 +23,13 @@ if (Sys.getenv("DONT_DOWNLOAD_JDBC_DRIVERS", "") != "TRUE") {
   }
 }
 
+# Helper functions -------------------------------------------------------------
+addDbmsToLabel <- function(label, testServer) {
+  # Test sections are not shown in R check, so also printing them here:
+  writeLines(sprintf("Test: %s (%s)", label, testServer$connectionDetails$dbms))
+  return(sprintf("%s (%s)", label, testServer$connectionDetails$dbms))
+}
+
 # Create a list with testing server details ------------------------------
 testServers <- list()
 
@@ -138,6 +147,27 @@ testServers[[length(testServers) + 1]] <- list(
   tempEmulationSchema = Sys.getenv("CDM5_SPARK_OHDSI_SCHEMA")
 )
 
+# BigQuery
+bqKeyFile <- tempfile(fileext = ".json")
+writeLines(Sys.getenv("CDM_BIG_QUERY_KEY_FILE"), bqKeyFile)
+if (testthat::is_testing()) {
+  withr::defer(unlink(bqKeyFile, force = TRUE), testthat::teardown_env())
+}
+bqConnectionString <- gsub("<keyfile path>",
+                           normalizePath(bqKeyFile, winslash = "/"),
+                           Sys.getenv("CDM_BIG_QUERY_CONNECTION_STRING"))
+testServers[[length(testServers) + 1]] <- list(
+  connectionDetails = details <- createConnectionDetails(
+    dbms = "bigquery",
+    user = "",
+    password = "",
+    connectionString = !!bqConnectionString
+  ),
+  NULL,
+  cdmDatabaseSchema = Sys.getenv("CDM_BIG_QUERY_CDM_SCHEMA"),
+  tempEmulationSchema = Sys.getenv("CDM_BIG_QUERY_OHDSI_SCHEMA")
+)
+
 # SQLite
 sqliteFile <- tempfile(fileext = ".sqlite")
 if (testthat::is_testing()) {
@@ -233,10 +263,3 @@ testServers[[length(testServers) + 1]] <- list(
   cdmDatabaseSchema = cdmDatabaseSchema,
   tempEmulationSchema = NULL
 )
-
-addDbmsToLabel <- function(label, testServer) {
-  # Test sections are not shown in R check, so also printing them here:
-  writeLines(sprintf("Test: %s (%s)", label, testServer$connectionDetails$dbms))
-  return(sprintf("%s (%s)", label, testServer$connectionDetails$dbms))
-}
-
diff --git a/tests/testthat/test-insertTable.R b/tests/testthat/test-insertTable.R
@@ -42,8 +42,8 @@ data$big_ints[8] <- 3.3043e+10
 
 for (testServer in testServers) {
   test_that(addDbmsToLabel("Insert data", testServer), {
-    if (testServer$connectionDetails$dbms == "redshift") {
-      # Inserting on RedShift is slow (Without bulk upload), so 
+    if (testServer$connectionDetails$dbms %in% c("redshift", "bigquery")) {
+      # Inserting on RedShift or BigQuery is slow (Without bulk upload), so 
       # taking subset:
       dataCopy1 <- data[1:10, ]
     } else {
@@ -94,6 +94,8 @@ for (testServer in testServers) {
       expect_equal(as.character(columnInfo$field.type), c("DATE", "TIMESTAMPNTZ", "NUMBER", "DOUBLE", "VARCHAR", "NUMBER"))
     } else if (dbms == "spark") {
       expect_equal(as.character(columnInfo$field.type), c("DATE", "TIMESTAMP", "INT", "FLOAT", "STRING", "BIGINT"))
+    } else if (dbms == "bigquery") {
+      expect_equal(as.character(columnInfo$field.type), c("DATE", "DATETIME", "INT64", "FLOAT64", "STRING", "INT64"))
     } else {
       warning("Unable to check column types for ", dbms)
     }