/Users/andrewlamb/Software/arrow-rs/arrow-data/src/equal/variable_size.rs
Line | Count | Source |
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 | | |
18 | | use crate::data::{ArrayData, contains_nulls}; |
19 | | use arrow_buffer::ArrowNativeType; |
20 | | use num_integer::Integer; |
21 | | |
22 | | use super::utils::equal_len; |
23 | | |
24 | 97.2k | fn offset_value_equal<T: ArrowNativeType + Integer>( |
25 | 97.2k | lhs_values: &[u8], |
26 | 97.2k | rhs_values: &[u8], |
27 | 97.2k | lhs_offsets: &[T], |
28 | 97.2k | rhs_offsets: &[T], |
29 | 97.2k | lhs_pos: usize, |
30 | 97.2k | rhs_pos: usize, |
31 | 97.2k | len: usize, |
32 | 97.2k | ) -> bool { |
33 | 97.2k | let lhs_start = lhs_offsets[lhs_pos].as_usize(); |
34 | 97.2k | let rhs_start = rhs_offsets[rhs_pos].as_usize(); |
35 | 97.2k | let lhs_len = (lhs_offsets[lhs_pos + len] - lhs_offsets[lhs_pos]) |
36 | 97.2k | .to_usize() |
37 | 97.2k | .unwrap(); |
38 | 97.2k | let rhs_len = (rhs_offsets[rhs_pos + len] - rhs_offsets[rhs_pos]) |
39 | 97.2k | .to_usize() |
40 | 97.2k | .unwrap(); |
41 | | |
42 | 97.2k | if lhs_len == 0 && rhs_len == 03 { |
43 | 3 | return true; |
44 | 97.2k | } |
45 | | |
46 | 97.2k | lhs_len == rhs_len && equal_len(lhs_values, rhs_values, lhs_start, rhs_start, lhs_len) |
47 | 97.2k | } |
48 | | |
49 | 80.1k | pub(super) fn variable_sized_equal<T: ArrowNativeType + Integer>( |
50 | 80.1k | lhs: &ArrayData, |
51 | 80.1k | rhs: &ArrayData, |
52 | 80.1k | lhs_start: usize, |
53 | 80.1k | rhs_start: usize, |
54 | 80.1k | len: usize, |
55 | 80.1k | ) -> bool { |
56 | 80.1k | let lhs_offsets = lhs.buffer::<T>(0); |
57 | 80.1k | let rhs_offsets = rhs.buffer::<T>(0); |
58 | | |
59 | | // the offsets of the `ArrayData` are ignored as they are only applied to the offset buffer. |
60 | 80.1k | let lhs_values = lhs.buffers()[1].as_slice(); |
61 | 80.1k | let rhs_values = rhs.buffers()[1].as_slice(); |
62 | | |
63 | | // Only checking one null mask here because by the time the control flow reaches |
64 | | // this point, the equality of the two masks would have already been verified. |
65 | 80.1k | if !contains_nulls(lhs.nulls(), lhs_start, len) { |
66 | 80.0k | offset_value_equal( |
67 | 80.0k | lhs_values, |
68 | 80.0k | rhs_values, |
69 | 80.0k | lhs_offsets, |
70 | 80.0k | rhs_offsets, |
71 | 80.0k | lhs_start, |
72 | 80.0k | rhs_start, |
73 | 80.0k | len, |
74 | | ) |
75 | | } else { |
76 | 20.1k | (0..len)50 .all50 (|i| { |
77 | 20.1k | let lhs_pos = lhs_start + i; |
78 | 20.1k | let rhs_pos = rhs_start + i; |
79 | | |
80 | | // the null bits can still be `None`, indicating that the value is valid. |
81 | 20.1k | let lhs_is_null = lhs.nulls().map(|v| v.is_null(lhs_pos)).unwrap_or_default(); |
82 | 20.1k | let rhs_is_null = rhs.nulls().map(|v| v.is_null(rhs_pos)).unwrap_or_default(); |
83 | | |
84 | 20.1k | lhs_is_null |
85 | 17.2k | || (lhs_is_null == rhs_is_null) |
86 | 17.2k | && offset_value_equal( |
87 | 17.2k | lhs_values, |
88 | 17.2k | rhs_values, |
89 | 17.2k | lhs_offsets, |
90 | 17.2k | rhs_offsets, |
91 | 17.2k | lhs_pos, |
92 | 17.2k | rhs_pos, |
93 | | 1, |
94 | | ) |
95 | 20.1k | }) |
96 | | } |
97 | 80.1k | } |