chore: more cleanup of error messages

Author: wiedld

datafusion/optimizer/src/analyzer/mod.rs

@@ -145,7 +145,7 @@ impl Analyzer {
     {
         // verify the logical plan required invariants at the start, before analyzer
         plan.check_invariants(InvariantLevel::Always)
-            .map_err(|e| e.context("assert_lp_invariants_before_analyzers"))?;
+            .map_err(|e| e.context("Invalid input plan passed to Analyzer"))?;
 
         let start_time = Instant::now();
         let mut new_plan = plan;
@@ -178,7 +178,7 @@ impl Analyzer {
         // verify at the end, after the last LP analyzer pass, that the plan is executable.
         new_plan
             .check_invariants(InvariantLevel::Executable)
-            .map_err(|e| e.context("Invalid plan after Analyzer"))?;
+            .map_err(|e| e.context("Invalid (non-executable) plan after Analyzer"))?;
 
         log_plan("Final analyzed plan", &new_plan);
         debug!("Analyzer took {} ms", start_time.elapsed().as_millis());

datafusion/optimizer/src/decorrelate_predicate_subquery.rs

@@ -835,7 +835,7 @@ mod tests {
             .build()?;
 
         // Maybe okay if the table only has a single column?
-        let expected = "Invalid plan after Analyzer\
+        let expected = "Invalid (non-executable) plan after Analyzer\
         \ncaused by\
         \nError during planning: InSubquery should only return one column, but found 4";
         assert_analyzer_check_err(vec![], plan, expected);
@@ -930,7 +930,7 @@ mod tests {
             .project(vec![col("customer.c_custkey")])?
             .build()?;
 
-        let expected = "Invalid plan after Analyzer\
+        let expected = "Invalid (non-executable) plan after Analyzer\
         \ncaused by\
         \nError during planning: InSubquery should only return one column";
         assert_analyzer_check_err(vec![], plan, expected);

datafusion/optimizer/src/optimizer.rs

@@ -358,7 +358,7 @@ impl Optimizer {
     {
         // verify LP is valid, before the first LP optimizer pass.
         plan.check_invariants(InvariantLevel::Executable)
-            .map_err(|e| e.context("Invalid plan before LP Optimizers"))?;
+            .map_err(|e| e.context("Invalid input plan before LP Optimizers"))?;
 
         let start_time = Instant::now();
         let options = config.options();
@@ -394,12 +394,12 @@ impl Optimizer {
                 .and_then(|tnr| {
                     // run checks optimizer invariant checks, per optimizer rule applied
                     assert_valid_optimization(&tnr.data, &starting_schema)
-                        .map_err(|e| e.context(format!("check_optimizer_specific_invariants after optimizer rule: {}", rule.name())))?;
+                        .map_err(|e| e.context(format!("Check optimizer-specific invariants after optimizer rule: {}", rule.name())))?;
 
                     // run LP invariant checks only in debug mode for performance reasons
                     #[cfg(debug_assertions)]
                     tnr.data.check_invariants(InvariantLevel::Executable)
-                        .map_err(|e| e.context(format!("check_plan_is_executable after optimizer rule: {}", rule.name())))?;
+                        .map_err(|e| e.context(format!("Invalid (non-executable) plan after Optimizer rule: {}", rule.name())))?;
 
                     Ok(tnr)
                 });
@@ -462,13 +462,15 @@ impl Optimizer {
 
         // verify that the optimizer passes only mutated what was permitted.
         assert_valid_optimization(&new_plan, &starting_schema).map_err(|e| {
-            e.context("check_optimizer_specific_invariants after all passes")
+            e.context("Check optimizer-specific invariants after all passes")
         })?;
 
         // verify LP is valid, after the last optimizer pass.
         new_plan
             .check_invariants(InvariantLevel::Executable)
-            .map_err(|e| e.context("Invalid plan after LP Optimizers"))?;
+            .map_err(|e| {
+                e.context("Invalid (non-executable) plan after LP Optimizers")
+            })?;
 
         log_plan("Final optimized plan", &new_plan);
         debug!("Optimizer took {} ms", start_time.elapsed().as_millis());
@@ -545,7 +547,7 @@ mod tests {
         assert_eq!(
             "Optimizer rule 'get table_scan rule' failed\n\
             caused by\n\
-            check_optimizer_specific_invariants after optimizer rule: get table_scan rule\n\
+            Check optimizer-specific invariants after optimizer rule: get table_scan rule\n\
             caused by\n\
             Internal error: Failed due to a difference in schemas, \
             original schema: DFSchema { inner: Schema { \

datafusion/optimizer/src/scalar_subquery_to_join.rs

@@ -731,7 +731,7 @@ mod tests {
             .project(vec![col("customer.c_custkey")])?
             .build()?;
 
-        let expected = "Invalid plan after Analyzer\
+        let expected = "Invalid (non-executable) plan after Analyzer\
         \ncaused by\
         \nError during planning: Scalar subquery should only return one column";
         assert_analyzer_check_err(vec![], plan, expected);
@@ -793,7 +793,7 @@ mod tests {
             .project(vec![col("customer.c_custkey")])?
             .build()?;
 
-        let expected = "Invalid plan after Analyzer\
+        let expected = "Invalid (non-executable) plan after Analyzer\
         \ncaused by\
         \nError during planning: Scalar subquery should only return one column";
         assert_analyzer_check_err(vec![], plan, expected);

datafusion/sqllogictest/test_files/subquery.slt

@@ -433,17 +433,32 @@ logical_plan
 08)----------TableScan: t1 projection=[t1_int]
 
 #invalid_scalar_subquery
-statement error DataFusion error: Invalid plan after Analyzer\ncaused by\nError during planning: Scalar subquery should only return one column, but found 2: t2.t2_id, t2.t2_name
+statement error
 SELECT t1_id, t1_name, t1_int, (select t2_id, t2_name FROM t2 WHERE t2.t2_id = t1.t1_int) FROM t1
+----
+DataFusion error: Invalid (non-executable) plan after Analyzer
+caused by
+Error during planning: Scalar subquery should only return one column, but found 2: t2.t2_id, t2.t2_name
+
 
 #subquery_not_allowed
 #In/Exist Subquery is not allowed in ORDER BY clause.
-statement error DataFusion error: Invalid plan after Analyzer\ncaused by\nError during planning: In/Exist subquery can only be used in Projection, Filter, TableScan, Window functions, Aggregate and Join plan nodes, but was used in \[Sort: t1.t1_int IN \(<subquery>\) ASC NULLS LAST\]
+statement error
 SELECT t1_id, t1_name, t1_int FROM t1 order by t1_int in (SELECT t2_int FROM t2 WHERE t1.t1_id > t1.t1_int)
+----
+DataFusion error: Invalid (non-executable) plan after Analyzer
+caused by
+Error during planning: In/Exist subquery can only be used in Projection, Filter, TableScan, Window functions, Aggregate and Join plan nodes, but was used in [Sort: t1.t1_int IN (<subquery>) ASC NULLS LAST]
+
 
 #non_aggregated_correlated_scalar_subquery
-statement error DataFusion error: Invalid plan after Analyzer\ncaused by\nError during planning: Correlated scalar subquery must be aggregated to return at most one row
+statement error
 SELECT t1_id, (SELECT t2_int FROM t2 WHERE t2.t2_int = t1.t1_int) as t2_int from t1
+----
+DataFusion error: Invalid (non-executable) plan after Analyzer
+caused by
+Error during planning: Correlated scalar subquery must be aggregated to return at most one row
+
 
 #non_aggregated_correlated_scalar_subquery_unique
 query II rowsort
@@ -456,12 +471,22 @@ SELECT t1_id, (SELECT t3_int FROM t3 WHERE t3.t3_id = t1.t1_id) as t3_int from t
 
 
 #non_aggregated_correlated_scalar_subquery
-statement error DataFusion error: Invalid plan after Analyzer\ncaused by\nError during planning: Correlated scalar subquery must be aggregated to return at most one row
+statement error
 SELECT t1_id, (SELECT t2_int FROM t2 WHERE t2.t2_int = t1_int group by t2_int) as t2_int from t1
+----
+DataFusion error: Invalid (non-executable) plan after Analyzer
+caused by
+Error during planning: Correlated scalar subquery must be aggregated to return at most one row
+
 
 #non_aggregated_correlated_scalar_subquery_with_limit
-statement error DataFusion error: Invalid plan after Analyzer\ncaused by\nError during planning: Correlated scalar subquery must be aggregated to return at most one row
+statement error
 SELECT t1_id, (SELECT t2_int FROM t2 WHERE t2.t2_int = t1.t1_int limit 2) as t2_int from t1
+----
+DataFusion error: Invalid (non-executable) plan after Analyzer
+caused by
+Error during planning: Correlated scalar subquery must be aggregated to return at most one row
+
 
 #non_aggregated_correlated_scalar_subquery_with_single_row
 query TT
@@ -523,8 +548,13 @@ logical_plan
 07)--TableScan: t1 projection=[t1_id]
 
 #aggregated_correlated_scalar_subquery_with_extra_group_by_columns
-statement error DataFusion error: Invalid plan after Analyzer\ncaused by\nError during planning: A GROUP BY clause in a scalar correlated subquery cannot contain non-correlated columns
+statement error
 SELECT t1_id, (SELECT sum(t2_int) FROM t2 WHERE t2.t2_id = t1.t1_id group by t2_name) as t2_sum from t1
+----
+DataFusion error: Invalid (non-executable) plan after Analyzer
+caused by
+Error during planning: A GROUP BY clause in a scalar correlated subquery cannot contain non-correlated columns
+
 
 #support_agg_correlated_columns
 query TT