datafusion_functions/datetime/
to_time.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 crate::datetime::common::*;
19use arrow::array::cast::AsArray;
20use arrow::array::temporal_conversions::time_to_time64ns;
21use arrow::array::types::Time64NanosecondType;
22use arrow::array::{Array, PrimitiveArray, StringArrayType};
23use arrow::datatypes::DataType;
24use arrow::datatypes::DataType::*;
25use chrono::format::{Item, Parsed, StrftimeItems, parse};
26use datafusion_common::{Result, ScalarValue, exec_err};
27use datafusion_expr::{
28    ColumnarValue, Documentation, ScalarFunctionArgs, ScalarUDFImpl, Signature,
29    Volatility,
30};
31use datafusion_macros::user_doc;
32use std::sync::Arc;
33
34/// Default time formats to try when parsing without an explicit format
35const DEFAULT_TIME_FORMATS: &[&str] = &[
36    "%H:%M:%S%.f", // 12:30:45.123456789
37    "%H:%M:%S",    // 12:30:45
38    "%H:%M",       // 12:30
39];
40
41#[user_doc(
42    doc_section(label = "Time and Date Functions"),
43    description = r"Converts a value to a time (`HH:MM:SS.nnnnnnnnn`).
44Supports strings and timestamps as input.
45Strings are parsed as `HH:MM:SS`, `HH:MM:SS.nnnnnnnnn`, or `HH:MM` if no [Chrono format](https://docs.rs/chrono/latest/chrono/format/strftime/index.html)s are provided.
46Timestamps will have the time portion extracted.
47Returns the corresponding time.
48
49Note: `to_time` returns Time64(Nanosecond), which represents the time of day in nanoseconds since midnight.",
50    syntax_example = "to_time(expression[, format1, ..., format_n])",
51    sql_example = r#"```sql
52> select to_time('12:30:45');
53+---------------------------+
54| to_time(Utf8("12:30:45")) |
55+---------------------------+
56| 12:30:45                  |
57+---------------------------+
58> select to_time('12-30-45', '%H-%M-%S');
59+--------------------------------------------+
60| to_time(Utf8("12-30-45"),Utf8("%H-%M-%S")) |
61+--------------------------------------------+
62| 12:30:45                                   |
63+--------------------------------------------+
64> select to_time('2024-01-15 14:30:45'::timestamp);
65+--------------------------------------------------+
66| to_time(Utf8("2024-01-15 14:30:45"))             |
67+--------------------------------------------------+
68| 14:30:45                                         |
69+--------------------------------------------------+
70```
71
72Additional examples can be found [here](https://github.com/apache/datafusion/blob/main/datafusion-examples/examples/builtin_functions/date_time.rs)
73"#,
74    standard_argument(name = "expression", prefix = "String or Timestamp"),
75    argument(
76        name = "format_n",
77        description = r"Optional [Chrono format](https://docs.rs/chrono/latest/chrono/format/strftime/index.html) strings to use to parse the expression. Formats will be tried in the order
78  they appear with the first successful one being returned. If none of the formats successfully parse the expression
79  an error will be returned."
80    )
81)]
82#[derive(Debug, PartialEq, Eq, Hash)]
83pub struct ToTimeFunc {
84    signature: Signature,
85}
86
87impl Default for ToTimeFunc {
88    fn default() -> Self {
89        Self::new()
90    }
91}
92
93impl ToTimeFunc {
94    pub fn new() -> Self {
95        Self {
96            signature: Signature::variadic_any(Volatility::Immutable),
97        }
98    }
99}
100
101impl ScalarUDFImpl for ToTimeFunc {
102    fn name(&self) -> &str {
103        "to_time"
104    }
105
106    fn signature(&self) -> &Signature {
107        &self.signature
108    }
109
110    fn return_type(&self, _arg_types: &[DataType]) -> Result<DataType> {
111        Ok(Time64(arrow::datatypes::TimeUnit::Nanosecond))
112    }
113
114    fn invoke_with_args(&self, args: ScalarFunctionArgs) -> Result<ColumnarValue> {
115        let args = args.args;
116        if args.is_empty() {
117            return exec_err!("to_time function requires 1 or more arguments, got 0");
118        }
119
120        // validate that any args after the first one are Utf8
121        if args.len() > 1 {
122            validate_data_types(&args, "to_time")?;
123        }
124
125        match args[0].data_type() {
126            Utf8View | LargeUtf8 | Utf8 => string_to_time(&args),
127            Null => Ok(ColumnarValue::Scalar(ScalarValue::Time64Nanosecond(None))),
128            // Support timestamp input by extracting time portion using Arrow cast
129            Timestamp(_, _) => timestamp_to_time(&args[0]),
130            other => {
131                exec_err!("Unsupported data type {} for function to_time", other)
132            }
133        }
134    }
135
136    fn documentation(&self) -> Option<&Documentation> {
137        self.doc()
138    }
139}
140
141/// Convert string arguments to time (standalone function, not a method on ToTimeFunc)
142fn string_to_time(args: &[ColumnarValue]) -> Result<ColumnarValue> {
143    let formats = collect_formats(args)?;
144    let formats = compile_formats(&formats);
145
146    match &args[0] {
147        ColumnarValue::Scalar(ScalarValue::Utf8(s))
148        | ColumnarValue::Scalar(ScalarValue::LargeUtf8(s))
149        | ColumnarValue::Scalar(ScalarValue::Utf8View(s)) => {
150            let result = s
151                .as_ref()
152                .map(|s| parse_time_with_formats(s, &formats))
153                .transpose()?;
154            Ok(ColumnarValue::Scalar(ScalarValue::Time64Nanosecond(result)))
155        }
156        ColumnarValue::Array(array) => {
157            let result = match array.data_type() {
158                Utf8 => parse_time_array(&array.as_string::<i32>(), &formats)?,
159                LargeUtf8 => parse_time_array(&array.as_string::<i64>(), &formats)?,
160                Utf8View => parse_time_array(&array.as_string_view(), &formats)?,
161                other => return exec_err!("Unsupported type for to_time: {other}"),
162            };
163            Ok(ColumnarValue::Array(Arc::new(result)))
164        }
165        other => exec_err!("Unsupported argument for to_time: {other:?}"),
166    }
167}
168
169/// Collect format strings from arguments, erroring on non-scalar inputs
170fn collect_formats(args: &[ColumnarValue]) -> Result<Vec<&str>> {
171    if args.len() <= 1 {
172        return Ok(DEFAULT_TIME_FORMATS.to_vec());
173    }
174
175    let mut formats = Vec::with_capacity(args.len() - 1);
176    for (i, arg) in args[1..].iter().enumerate() {
177        match arg {
178            ColumnarValue::Scalar(ScalarValue::Utf8(Some(s)))
179            | ColumnarValue::Scalar(ScalarValue::LargeUtf8(Some(s)))
180            | ColumnarValue::Scalar(ScalarValue::Utf8View(Some(s))) => {
181                formats.push(s.as_str());
182            }
183            ColumnarValue::Scalar(ScalarValue::Utf8(None))
184            | ColumnarValue::Scalar(ScalarValue::LargeUtf8(None))
185            | ColumnarValue::Scalar(ScalarValue::Utf8View(None)) => {
186                // Skip null format strings
187            }
188            ColumnarValue::Array(_) => {
189                return exec_err!(
190                    "to_time format argument {} must be a scalar, not an array",
191                    i + 2 // argument position (1-indexed, +1 for the first arg)
192                );
193            }
194            other => {
195                return exec_err!(
196                    "to_time format argument {} has unsupported type: {:?}",
197                    i + 2,
198                    other.data_type()
199                );
200            }
201        }
202    }
203    Ok(formats)
204}
205
206/// Extract time portion from timestamp using Arrow cast kernel
207fn timestamp_to_time(arg: &ColumnarValue) -> Result<ColumnarValue> {
208    arg.cast_to(&Time64(arrow::datatypes::TimeUnit::Nanosecond), None)
209}
210
211struct CompiledTimeFormat<'a> {
212    source: &'a str,
213    items: Vec<Item<'a>>,
214}
215
216fn compile_formats<'a>(formats: &[&'a str]) -> Vec<CompiledTimeFormat<'a>> {
217    formats
218        .iter()
219        .map(|source| CompiledTimeFormat {
220            source,
221            items: StrftimeItems::new(source).collect(),
222        })
223        .collect()
224}
225
226/// Parse time array using the provided formats
227fn parse_time_array<'a, A: StringArrayType<'a>>(
228    array: &A,
229    formats: &[CompiledTimeFormat<'_>],
230) -> Result<PrimitiveArray<Time64NanosecondType>> {
231    let mut values = Vec::with_capacity(array.len());
232    for i in 0..array.len() {
233        if array.is_null(i) {
234            values.push(0);
235        } else {
236            values.push(parse_time_with_formats(array.value(i), formats)?);
237        }
238    }
239    Ok(PrimitiveArray::new(values.into(), array.nulls().cloned()))
240}
241
242/// Parse time string using provided formats
243fn parse_time_with_formats(s: &str, formats: &[CompiledTimeFormat<'_>]) -> Result<i64> {
244    for format in formats {
245        let mut parsed = Parsed::new();
246        if parse(&mut parsed, s, format.items.iter()).is_ok()
247            && let Ok(time) = parsed.to_naive_time()
248        {
249            return Ok(time_to_time64ns(time));
250        }
251    }
252    exec_err!(
253        "Error parsing '{}' as time. Tried formats: {:?}",
254        s,
255        formats
256            .iter()
257            .map(|format| format.source)
258            .collect::<Vec<_>>()
259    )
260}
datafusion_functions/datetime/to_time.rs

datafusion_functions/datetime/
to_time.rs
