From eb6b035cd3ad39bb0d6d342975458890dda55762 Mon Sep 17 00:00:00 2001
From: Niels Bantilan <niels.bantilan@gmail.com>
Date: Wed, 15 Mar 2023 18:25:37 -0400
Subject: [PATCH] fix regression: all missing columns should be reported
 (#1117)

Signed-off-by: Niels Bantilan <niels.bantilan@gmail.com>
---
 pandera/backends/pandas/container.py | 37 +++++++++++++++++++---------
 tests/core/test_schemas.py           | 21 ++++++++++++++++
 2 files changed, 46 insertions(+), 12 deletions(-)

diff --git a/pandera/backends/pandas/container.py b/pandera/backends/pandas/container.py
index dfbb7f74a..82c356dd8 100644
--- a/pandera/backends/pandas/container.py
+++ b/pandera/backends/pandas/container.py
@@ -67,8 +67,12 @@ def validate(
 
         try:
             self.check_column_presence(check_obj, schema, column_info)
-        except SchemaError as exc:
-            error_handler.collect_error(exc.reason_code, exc)
+        except SchemaErrors as exc:
+            for schema_error in exc.schema_errors:
+                error_handler.collect_error(
+                    schema_error["reason_code"],
+                    schema_error["error"],
+                )
 
         # strictness check and filter
         try:
@@ -471,18 +475,27 @@ def check_column_presence(
     ):
         """Check for presence of specified columns in the data object."""
         if column_info.absent_column_names:
-            # NOTE: only report the first absent column for now, need to update
-            # this when backend stuff is complete
-            colname, *_ = column_info.absent_column_names
-            raise SchemaError(
+            reason_code = "column_not_in_dataframe"
+            raise SchemaErrors(
                 schema=schema,
+                schema_errors=[
+                    {
+                        "reason_code": reason_code,
+                        "error": SchemaError(
+                            schema=schema,
+                            data=check_obj,
+                            message=(
+                                f"column '{colname}' not in dataframe"
+                                f"\n{check_obj.head()}"
+                            ),
+                            failure_cases=scalar_failure_case(colname),
+                            check="column_in_dataframe",
+                            reason_code=reason_code,
+                        ),
+                    }
+                    for colname in column_info.absent_column_names
+                ],
                 data=check_obj,
-                message=(
-                    f"column '{colname}' not in dataframe\n{check_obj.head()}"
-                ),
-                failure_cases=scalar_failure_case(colname),
-                check="column_in_dataframe",
-                reason_code="column_not_in_dataframe",
             )
 
     def check_column_values_are_unique(self, check_obj: pd.DataFrame, schema):
diff --git a/tests/core/test_schemas.py b/tests/core/test_schemas.py
index 833473165..eee6a13d6 100644
--- a/tests/core/test_schemas.py
+++ b/tests/core/test_schemas.py
@@ -1965,3 +1965,24 @@ def test_column_set_unique():
     assert not test_schema.columns["a"].unique
     test_schema = test_schema.update_column("a", unique=True)
     assert test_schema.columns["a"].unique
+
+
+def test_missing_columns():
+    """Test that multiple missing columns is correctly reported."""
+    schema = DataFrameSchema(
+        {
+            "column3": Column(int),
+            "column2": Column(float),
+        }
+    )
+
+    df = pd.DataFrame({"column1": [1]})
+
+    try:
+        schema.validate(df, lazy=True)
+    except errors.SchemaErrors as exc:
+        assert (exc.failure_cases["check"] == "column_in_dataframe").all()
+        assert exc.failure_cases["failure_case"].tolist() == [
+            "column3",
+            "column2",
+        ]