datafusion_spark/function/datetime/
to_utc_timestamp.rs

1// Licensed to the Apache Software Foundation (ASF) under one
2// or more contributor license agreements.  See the NOTICE file
3// distributed with this work for additional information
4// regarding copyright ownership.  The ASF licenses this file
5// to you under the Apache License, Version 2.0 (the
6// "License"); you may not use this file except in compliance
7// with the License.  You may obtain a copy of the License at
8//
9//   http://www.apache.org/licenses/LICENSE-2.0
10//
11// Unless required by applicable law or agreed to in writing,
12// software distributed under the License is distributed on an
13// "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
14// KIND, either express or implied.  See the License for the
15// specific language governing permissions and limitations
16// under the License.
17
18use std::sync::Arc;
19
20use arrow::array::timezone::Tz;
21use arrow::array::{Array, ArrayRef, AsArray, PrimitiveBuilder, StringArrayType};
22use arrow::datatypes::TimeUnit;
23use arrow::datatypes::{
24    ArrowTimestampType, DataType, Field, FieldRef, TimestampMicrosecondType,
25    TimestampMillisecondType, TimestampNanosecondType, TimestampSecondType,
26};
27use chrono::{DateTime, Offset, TimeZone};
28use datafusion_common::types::{NativeType, logical_string};
29use datafusion_common::utils::take_function_args;
30use datafusion_common::{
31    Result, exec_datafusion_err, exec_err, internal_datafusion_err, internal_err,
32};
33use datafusion_expr::{
34    Coercion, ColumnarValue, ReturnFieldArgs, ScalarFunctionArgs, ScalarUDFImpl,
35    Signature, TypeSignatureClass, Volatility,
36};
37use datafusion_functions::utils::make_scalar_function;
38
39/// Apache Spark `to_utc_timestamp` function.
40///
41/// Interprets the given timestamp in the provided timezone and then converts it to UTC.
42///
43/// Timestamp in Apache Spark represents number of microseconds from the Unix epoch, which is not
44/// timezone-agnostic. So in Apache Spark this function just shift the timestamp value from the given
45/// timezone to UTC timezone.
46///
47/// See <https://spark.apache.org/docs/latest/api/sql/index.html#to_utc_timestamp>
48#[derive(Debug, PartialEq, Eq, Hash)]
49pub struct SparkToUtcTimestamp {
50    signature: Signature,
51}
52
53impl Default for SparkToUtcTimestamp {
54    fn default() -> Self {
55        Self::new()
56    }
57}
58
59impl SparkToUtcTimestamp {
60    pub fn new() -> Self {
61        Self {
62            signature: Signature::coercible(
63                vec![
64                    Coercion::new_implicit(
65                        TypeSignatureClass::Timestamp,
66                        vec![TypeSignatureClass::Native(logical_string())],
67                        NativeType::Timestamp(TimeUnit::Microsecond, None),
68                    ),
69                    Coercion::new_exact(TypeSignatureClass::Native(logical_string())),
70                ],
71                Volatility::Immutable,
72            ),
73        }
74    }
75}
76
77impl ScalarUDFImpl for SparkToUtcTimestamp {
78    fn name(&self) -> &str {
79        "to_utc_timestamp"
80    }
81
82    fn signature(&self) -> &Signature {
83        &self.signature
84    }
85
86    fn return_type(&self, _arg_types: &[DataType]) -> Result<DataType> {
87        internal_err!("return_field_from_args should be used instead")
88    }
89
90    fn return_field_from_args(&self, args: ReturnFieldArgs) -> Result<FieldRef> {
91        let nullable = args.arg_fields.iter().any(|f| f.is_nullable());
92
93        Ok(Arc::new(Field::new(
94            self.name(),
95            args.arg_fields[0].data_type().clone(),
96            nullable,
97        )))
98    }
99
100    fn invoke_with_args(&self, args: ScalarFunctionArgs) -> Result<ColumnarValue> {
101        make_scalar_function(to_utc_timestamp, vec![])(&args.args)
102    }
103}
104
105fn to_utc_timestamp(args: &[ArrayRef]) -> Result<ArrayRef> {
106    let [timestamp, timezone] = take_function_args("to_utc_timestamp", args)?;
107
108    match timestamp.data_type() {
109        DataType::Timestamp(TimeUnit::Nanosecond, tz_opt) => {
110            process_timestamp_with_tz_array::<TimestampNanosecondType>(
111                timestamp,
112                timezone,
113                tz_opt.clone(),
114            )
115        }
116        DataType::Timestamp(TimeUnit::Microsecond, tz_opt) => {
117            process_timestamp_with_tz_array::<TimestampMicrosecondType>(
118                timestamp,
119                timezone,
120                tz_opt.clone(),
121            )
122        }
123        DataType::Timestamp(TimeUnit::Millisecond, tz_opt) => {
124            process_timestamp_with_tz_array::<TimestampMillisecondType>(
125                timestamp,
126                timezone,
127                tz_opt.clone(),
128            )
129        }
130        DataType::Timestamp(TimeUnit::Second, tz_opt) => {
131            process_timestamp_with_tz_array::<TimestampSecondType>(
132                timestamp,
133                timezone,
134                tz_opt.clone(),
135            )
136        }
137        ts_type => {
138            exec_err!("`to_utc_timestamp`: unsupported argument types: {ts_type}")
139        }
140    }
141}
142
143fn process_timestamp_with_tz_array<T: ArrowTimestampType>(
144    ts_array: &ArrayRef,
145    tz_array: &ArrayRef,
146    tz_opt: Option<Arc<str>>,
147) -> Result<ArrayRef> {
148    match tz_array.data_type() {
149        DataType::Utf8 => {
150            process_arrays::<T, _>(tz_opt, ts_array, tz_array.as_string::<i32>())
151        }
152        DataType::LargeUtf8 => {
153            process_arrays::<T, _>(tz_opt, ts_array, tz_array.as_string::<i64>())
154        }
155        DataType::Utf8View => {
156            process_arrays::<T, _>(tz_opt, ts_array, tz_array.as_string_view())
157        }
158        other => {
159            exec_err!("`to_utc_timestamp`: timezone must be a string type, got {other}")
160        }
161    }
162}
163
164fn process_arrays<'a, T: ArrowTimestampType, S>(
165    return_tz_opt: Option<Arc<str>>,
166    ts_array: &ArrayRef,
167    tz_array: &'a S,
168) -> Result<ArrayRef>
169where
170    &'a S: StringArrayType<'a>,
171{
172    let ts_primitive = ts_array.as_primitive::<T>();
173    let mut builder = PrimitiveBuilder::<T>::with_capacity(ts_array.len());
174
175    for (ts_opt, tz_opt) in ts_primitive.iter().zip(tz_array.iter()) {
176        match (ts_opt, tz_opt) {
177            (Some(ts), Some(tz_str)) => {
178                let tz: Tz = tz_str.parse().map_err(|e| {
179                    exec_datafusion_err!(
180                        "`to_utc_timestamp`: invalid timezone '{tz_str}': {e}"
181                    )
182                })?;
183                let val = adjust_to_utc_time::<T>(ts, tz)?;
184                builder.append_value(val);
185            }
186            _ => builder.append_null(),
187        }
188    }
189
190    builder = builder.with_timezone_opt(return_tz_opt);
191    Ok(Arc::new(builder.finish()))
192}
193
194fn adjust_to_utc_time<T: ArrowTimestampType>(ts: i64, tz: Tz) -> Result<i64> {
195    let dt = match T::UNIT {
196        TimeUnit::Nanosecond => Some(DateTime::from_timestamp_nanos(ts)),
197        TimeUnit::Microsecond => DateTime::from_timestamp_micros(ts),
198        TimeUnit::Millisecond => DateTime::from_timestamp_millis(ts),
199        TimeUnit::Second => DateTime::from_timestamp(ts, 0),
200    }
201    .ok_or_else(|| internal_datafusion_err!("Invalid timestamp"))?;
202    let naive_dt = dt.naive_utc();
203
204    let offset_seconds = tz
205        .offset_from_utc_datetime(&naive_dt)
206        .fix()
207        .local_minus_utc() as i64;
208
209    let offset_in_unit = match T::UNIT {
210        TimeUnit::Nanosecond => offset_seconds.checked_mul(1_000_000_000),
211        TimeUnit::Microsecond => offset_seconds.checked_mul(1_000_000),
212        TimeUnit::Millisecond => offset_seconds.checked_mul(1_000),
213        TimeUnit::Second => Some(offset_seconds),
214    }
215    .ok_or_else(|| internal_datafusion_err!("Offset overflow"))?;
216
217    ts.checked_sub(offset_in_unit).ok_or_else(|| {
218        internal_datafusion_err!("Timestamp overflow during timezone adjustment")
219    })
220}
datafusion_spark/function/datetime/to_utc_timestamp.rs

datafusion_spark/function/datetime/
to_utc_timestamp.rs
