fix: correct date_trunc for times before the epoch

The array-based implementation of date_trunc can produce incorrect
results for negative timestamps (i.e. dates before 1970-01-01). Check
for any such incorrect values and compensate accordingly.

Author: mhilton

datafusion/functions/src/datetime/date_trunc.rs

@@ -482,9 +482,20 @@ fn general_date_trunc_array_fine_granularity<T: ArrowTimestampType>(
 
     if let Some(unit) = unit {
         let original_type = array.data_type();
-        let array = arrow::compute::cast(array, &DataType::Int64)?;
-        let array = arrow::compute::kernels::numeric::div(&array, &unit)?;
+        let input = arrow::compute::cast(array, &DataType::Int64)?;
+        let array = arrow::compute::kernels::numeric::div(&input, &unit)?;
         let array = arrow::compute::kernels::numeric::mul(&array, &unit)?;
+        // For timestamps before 1970-01-01T00:00:00Z (negative values)
+        // it is possible that the truncated value is actually later
+        // than the original value. Correct any such cases by
+        // subtracting `unit`.
+        let too_late = arrow::compute::kernels::cmp::gt(&array, &input)?;
+        let array = if too_late.true_count() > 0 {
+            let earlier = arrow::compute::kernels::numeric::sub(&array, &unit)?;
+            arrow::compute::kernels::zip::zip(&too_late, &earlier, &array)?
+        } else {
+            array
+        };
         let array = arrow::compute::cast(&array, original_type)?;
         Ok(array)
     } else {

datafusion/sqllogictest/test_files/timestamps.slt

@@ -1687,6 +1687,13 @@ SELECT DATE_TRUNC('second', '2022-08-03 14:38:50Z');
 ----
 2022-08-03T14:38:50
 
+# DATE_TRUNC handling of times before the unix epoch (issue 18334)
+query PPP
+SELECT d, DATE_TRUNC('hour', d), DATE_TRUNC('hour', TIMESTAMP '1900-06-15 07:09:00')
+FROM (VALUES (TIMESTAMP '1900-06-15 07:09:00')) AS t(d);
+----
+1900-06-15T07:09:00 1900-06-15T07:00:00 1900-06-15T07:00:00
+
 # Test that interval can add a timestamp
 query P
 SELECT timestamp '2013-07-01 12:00:00' + INTERVAL '8' DAY;
@@ -2244,7 +2251,7 @@ SET TIME ZONE = '+05:00'
 
 statement ok
 CREATE TABLE foo (time TIMESTAMPTZ) AS VALUES
-    ('2020-01-01T00:00:00+05:00'), 
+    ('2020-01-01T00:00:00+05:00'),
     ('2020-01-01T01:00:00+05:00'),
     ('2020-01-01T02:00:00+05:00'),
     ('2020-01-01T03:00:00+05:00')
@@ -2329,17 +2336,17 @@ NULL 1970-01-01T00:00:00 2031-01-19T23:33:25 1970-01-01T00:00:01 1969-12-31T23:5
 # verify timestamp syntax styles are consistent
 query BBBBBBBBBBBBB
 SELECT to_timestamp(null) is null as c1,
-       null::timestamp is null as c2, 
-       cast(null as timestamp) is null as c3, 
-       to_timestamp(0) = 0::timestamp as c4, 
-       to_timestamp(1926632005) = 1926632005::timestamp as c5, 
-       to_timestamp(1) = 1::timestamp as c6, 
-       to_timestamp(-1) = -1::timestamp as c7, 
+       null::timestamp is null as c2,
+       cast(null as timestamp) is null as c3,
+       to_timestamp(0) = 0::timestamp as c4,
+       to_timestamp(1926632005) = 1926632005::timestamp as c5,
+       to_timestamp(1) = 1::timestamp as c6,
+       to_timestamp(-1) = -1::timestamp as c7,
        to_timestamp(0-1) = (0-1)::timestamp as c8,
-       to_timestamp(0) = cast(0 as timestamp) as c9, 
-       to_timestamp(1926632005) = cast(1926632005 as timestamp) as c10, 
-       to_timestamp(1) = cast(1 as timestamp) as c11, 
-       to_timestamp(-1) = cast(-1 as timestamp) as c12, 
+       to_timestamp(0) = cast(0 as timestamp) as c9,
+       to_timestamp(1926632005) = cast(1926632005 as timestamp) as c10,
+       to_timestamp(1) = cast(1 as timestamp) as c11,
+       to_timestamp(-1) = cast(-1 as timestamp) as c12,
        to_timestamp(0-1) = cast(0-1 as timestamp) as c13
 ----
 true true true true true true true true true true true true true
@@ -2352,10 +2359,10 @@ Timestamp(ns) Timestamp(ns) Timestamp(ns)
 
 # verify timestamp output types using timestamp literal syntax
 query BBBBBB
-SELECT arrow_typeof(to_timestamp(1)) = arrow_typeof(1::timestamp) as c1, 
+SELECT arrow_typeof(to_timestamp(1)) = arrow_typeof(1::timestamp) as c1,
        arrow_typeof(to_timestamp(null)) = arrow_typeof(null::timestamp) as c2,
        arrow_typeof(to_timestamp('2023-01-10 12:34:56.000')) = arrow_typeof('2023-01-10 12:34:56.000'::timestamp) as c3,
-       arrow_typeof(to_timestamp(1)) = arrow_typeof(cast(1 as timestamp)) as c4, 
+       arrow_typeof(to_timestamp(1)) = arrow_typeof(cast(1 as timestamp)) as c4,
        arrow_typeof(to_timestamp(null)) = arrow_typeof(cast(null as timestamp)) as c5,
        arrow_typeof(to_timestamp('2023-01-10 12:34:56.000')) = arrow_typeof(cast('2023-01-10 12:34:56.000' as timestamp)) as c6
 ----
@@ -2599,13 +2606,13 @@ drop table table_a
 ##########
 
 statement ok
-create table table_a (ts timestamp) as values 
-    ('2020-09-08T11:42:29Z'::timestamp), 
+create table table_a (ts timestamp) as values
+    ('2020-09-08T11:42:29Z'::timestamp),
     ('2020-09-08T12:42:29Z'::timestamp),
     ('2020-09-08T13:42:29Z'::timestamp)
 
 statement ok
-create table table_b (ts timestamp) as values 
+create table table_b (ts timestamp) as values
     ('2020-09-08T11:42:29.190Z'::timestamp),
     ('2020-09-08T13:42:29.190Z'::timestamp),
     ('2020-09-08T12:42:29.190Z'::timestamp)
@@ -2727,8 +2734,8 @@ statement ok
 drop table t1
 
 statement ok
-create table table_a (val int, ts1 timestamp, ts2 timestamp) as values 
-    (1, '2018-07-01T06:00:00'::timestamp, '2018-07-01T07:00:00'::timestamp), 
+create table table_a (val int, ts1 timestamp, ts2 timestamp) as values
+    (1, '2018-07-01T06:00:00'::timestamp, '2018-07-01T07:00:00'::timestamp),
     (2, '2018-07-01T07:00:00'::timestamp, '2018-07-01T08:00:00'::timestamp)
 
 query I?
@@ -3032,7 +3039,7 @@ NULL
 
 query T
 SELECT to_char(date_column, '%Y-%m-%d')
-FROM (VALUES 
+FROM (VALUES
     (DATE '2020-09-01'),
     (NULL)
 ) AS t(date_column);
@@ -3042,7 +3049,7 @@ NULL
 
 query T
 SELECT to_char(date_column, '%Y-%m-%d')
-FROM (VALUES 
+FROM (VALUES
     (NULL),
     (DATE '2020-09-01')
 ) AS t(date_column);