From 5ae6285c42b8ce4af9163e28325f1d47a67af2ce Mon Sep 17 00:00:00 2001
From: Piotr Findeisen <piotr.findeisen@gmail.com>
Date: Thu, 31 Oct 2024 15:29:07 +0100
Subject: [PATCH] Allow testing values with trailing whitespace in SLT tests

---
 datafusion/sqllogictest/bin/sqllogictests.rs  | 23 +++++++++++++++++--
 .../test_files/string/string_literal.slt      |  5 ++++
 2 files changed, 26 insertions(+), 2 deletions(-)

diff --git a/datafusion/sqllogictest/bin/sqllogictests.rs b/datafusion/sqllogictest/bin/sqllogictests.rs
index 2479252a7b5b0..c3e739d146c6c 100644
--- a/datafusion/sqllogictest/bin/sqllogictests.rs
+++ b/datafusion/sqllogictest/bin/sqllogictests.rs
@@ -22,6 +22,7 @@ use std::path::{Path, PathBuf};
 use clap::Parser;
 use datafusion_sqllogictest::{DataFusion, TestContext};
 use futures::stream::StreamExt;
+use itertools::Itertools;
 use log::info;
 use sqllogictest::strict_column_validator;
 
@@ -39,6 +40,23 @@ pub fn main() -> Result<()> {
         .block_on(run_tests())
 }
 
+fn value_validator(actual: &[Vec<String>], expected: &[String]) -> bool {
+    let expected = expected
+        .iter()
+        // Trailing whitespace from lines in SLT will typically be removed, but do not fail if it is not
+        // If particular test wants to cover trailing whitespace on a value,
+        // it should project additional non-whitespace column on the right.
+        .map(|s| s.trim_end().to_owned())
+        .collect::<Vec<_>>();
+    let actual = actual
+        .iter()
+        .map(|strs| strs.iter().join(" "))
+        // Editors do not preserve trailing whitespace, so expected may or may not lack it included
+        .map(|s| s.trim_end().to_owned())
+        .collect::<Vec<_>>();
+    actual == expected
+}
+
 /// Sets up an empty directory at test_files/scratch/<name>
 /// creating it if needed and clearing any file contents if it exists
 /// This allows tests for inserting to external tables or copy to
@@ -140,6 +158,7 @@ async fn run_test_file(test_file: TestFile) -> Result<()> {
         ))
     });
     runner.with_column_validator(strict_column_validator);
+    runner.with_validator(value_validator);
     runner
         .run_file_async(path)
         .await
@@ -158,6 +177,7 @@ async fn run_test_file_with_postgres(test_file: TestFile) -> Result<()> {
     let mut runner =
         sqllogictest::Runner::new(|| Postgres::connect(relative_path.clone()));
     runner.with_column_validator(strict_column_validator);
+    runner.with_validator(value_validator);
     runner
         .run_file_async(path)
         .await
@@ -176,7 +196,6 @@ async fn run_complete_file(test_file: TestFile) -> Result<()> {
         path,
         relative_path,
     } = test_file;
-    use sqllogictest::default_validator;
 
     info!("Using complete mode to complete: {}", path.display());
 
@@ -196,7 +215,7 @@ async fn run_complete_file(test_file: TestFile) -> Result<()> {
         .update_test_file(
             path,
             col_separator,
-            default_validator,
+            value_validator,
             strict_column_validator,
         )
         .await
diff --git a/datafusion/sqllogictest/test_files/string/string_literal.slt b/datafusion/sqllogictest/test_files/string/string_literal.slt
index 5d847747693d8..80bd7fc59c008 100644
--- a/datafusion/sqllogictest/test_files/string/string_literal.slt
+++ b/datafusion/sqllogictest/test_files/string/string_literal.slt
@@ -816,3 +816,8 @@ query B
 SELECT starts_with('foobar', 'bar')
 ----
 false
+
+query TT
+select '   ', '|'
+----
+    |