-
Notifications
You must be signed in to change notification settings - Fork 787
/
index.rs
349 lines (319 loc) · 12.3 KB
/
index.rs
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
// Licensed to the Apache Software Foundation (ASF) under one
// or more contributor license agreements. See the NOTICE file
// distributed with this work for additional information
// regarding copyright ownership. The ASF licenses this file
// to you under the Apache License, Version 2.0 (the
// "License"); you may not use this file except in compliance
// with the License. You may obtain a copy of the License at
//
// http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing,
// software distributed under the License is distributed on an
// "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
// KIND, either express or implied. See the License for the
// specific language governing permissions and limitations
// under the License.
//! [`Index`] structures holding decoded [`ColumnIndex`] information
use crate::basic::Type;
use crate::data_type::private::ParquetValueType;
use crate::data_type::{AsBytes, ByteArray, FixedLenByteArray, Int96};
use crate::errors::ParquetError;
use crate::file::metadata::LevelHistogram;
use crate::format::{BoundaryOrder, ColumnIndex};
use std::fmt::Debug;
/// Typed statistics for one data page
///
/// See [`NativeIndex`] for more details
#[derive(Debug, Clone, PartialEq, Eq, Hash)]
pub struct PageIndex<T> {
/// The minimum value, It is None when all values are null
pub min: Option<T>,
/// The maximum value, It is None when all values are null
pub max: Option<T>,
/// Null values in the page
pub null_count: Option<i64>,
/// Repetition level histogram for the page
///
/// `repetition_level_histogram[i]` is a count of how many values are at repetition level `i`.
/// For example, `repetition_level_histogram[0]` indicates how many rows the page contains.
pub repetition_level_histogram: Option<LevelHistogram>,
/// Definition level histogram for the page
///
/// `definition_level_histogram[i]` is a count of how many values are at definition level `i`.
/// For example, `definition_level_histogram[max_definition_level]` indicates how many
/// non-null values are present in the page.
pub definition_level_histogram: Option<LevelHistogram>,
}
impl<T> PageIndex<T> {
pub fn min(&self) -> Option<&T> {
self.min.as_ref()
}
pub fn max(&self) -> Option<&T> {
self.max.as_ref()
}
pub fn null_count(&self) -> Option<i64> {
self.null_count
}
pub fn repetition_level_histogram(&self) -> Option<&LevelHistogram> {
self.repetition_level_histogram.as_ref()
}
pub fn definition_level_histogram(&self) -> Option<&LevelHistogram> {
self.definition_level_histogram.as_ref()
}
}
impl<T> PageIndex<T>
where
T: AsBytes,
{
pub fn max_bytes(&self) -> Option<&[u8]> {
self.max.as_ref().map(|x| x.as_bytes())
}
pub fn min_bytes(&self) -> Option<&[u8]> {
self.min.as_ref().map(|x| x.as_bytes())
}
}
#[derive(Debug, Clone, PartialEq)]
#[allow(non_camel_case_types)]
/// Statistics for data pages in a column chunk.
///
/// See [`NativeIndex`] for more information
pub enum Index {
/// Sometimes reading page index from parquet file
/// will only return pageLocations without min_max index,
/// `NONE` represents this lack of index information
NONE,
BOOLEAN(NativeIndex<bool>),
INT32(NativeIndex<i32>),
INT64(NativeIndex<i64>),
INT96(NativeIndex<Int96>),
FLOAT(NativeIndex<f32>),
DOUBLE(NativeIndex<f64>),
BYTE_ARRAY(NativeIndex<ByteArray>),
FIXED_LEN_BYTE_ARRAY(NativeIndex<FixedLenByteArray>),
}
impl Index {
/// Return min/max elements inside ColumnIndex are ordered or not.
pub fn is_sorted(&self) -> bool {
// 0:UNORDERED, 1:ASCENDING ,2:DESCENDING,
if let Some(order) = self.get_boundary_order() {
order.0 > (BoundaryOrder::UNORDERED.0)
} else {
false
}
}
/// Get boundary_order of this page index.
pub fn get_boundary_order(&self) -> Option<BoundaryOrder> {
match self {
Index::NONE => None,
Index::BOOLEAN(index) => Some(index.boundary_order),
Index::INT32(index) => Some(index.boundary_order),
Index::INT64(index) => Some(index.boundary_order),
Index::INT96(index) => Some(index.boundary_order),
Index::FLOAT(index) => Some(index.boundary_order),
Index::DOUBLE(index) => Some(index.boundary_order),
Index::BYTE_ARRAY(index) => Some(index.boundary_order),
Index::FIXED_LEN_BYTE_ARRAY(index) => Some(index.boundary_order),
}
}
}
/// Strongly typed statistics for data pages in a column chunk.
///
/// This structure is a natively typed, in memory representation of the
/// [`ColumnIndex`] structure in a parquet file footer, as described in the
/// Parquet [PageIndex documentation]. The statistics stored in this structure
/// can be used by query engines to skip decoding pages while reading parquet
/// data.
///
/// # Differences with Row Group Level Statistics
///
/// One significant difference between `NativeIndex` and row group level
/// [`Statistics`] is that page level statistics may not store actual column
/// values as min and max (e.g. they may store truncated strings to save space)
///
/// [PageIndex documentation]: https://github.com/apache/parquet-format/blob/master/PageIndex.md
/// [`Statistics`]: crate::file::statistics::Statistics
#[derive(Debug, Clone, PartialEq, Eq, Hash)]
pub struct NativeIndex<T: ParquetValueType> {
/// The actual column indexes, one item per page
pub indexes: Vec<PageIndex<T>>,
/// If the min/max elements are ordered, and if so in which
/// direction. See [source] for details.
///
/// [source]: https://github.com/apache/parquet-format/blob/bfc549b93e6927cb1fc425466e4084f76edc6d22/src/main/thrift/parquet.thrift#L959-L964
pub boundary_order: BoundaryOrder,
}
impl<T: ParquetValueType> NativeIndex<T> {
pub const PHYSICAL_TYPE: Type = T::PHYSICAL_TYPE;
/// Creates a new [`NativeIndex`]
pub(crate) fn try_new(index: ColumnIndex) -> Result<Self, ParquetError> {
let len = index.min_values.len();
let null_counts = index
.null_counts
.map(|x| x.into_iter().map(Some).collect::<Vec<_>>())
.unwrap_or_else(|| vec![None; len]);
// histograms are a 1D array encoding a 2D num_pages X num_levels matrix.
let to_page_histograms = |opt_hist: Option<Vec<i64>>| {
if let Some(hist) = opt_hist {
// TODO: should we assert (hist.len() % len) == 0?
let num_levels = hist.len() / len;
let mut res = Vec::with_capacity(len);
for i in 0..len {
let page_idx = i * num_levels;
let page_hist = hist[page_idx..page_idx + num_levels].to_vec();
res.push(Some(LevelHistogram::from(page_hist)));
}
res
} else {
vec![None; len]
}
};
let rep_hists: Vec<Option<LevelHistogram>> =
to_page_histograms(index.repetition_level_histograms);
let def_hists: Vec<Option<LevelHistogram>> =
to_page_histograms(index.definition_level_histograms);
let indexes = index
.min_values
.iter()
.zip(index.max_values.iter())
.zip(index.null_pages.into_iter())
.zip(null_counts.into_iter())
.zip(rep_hists.into_iter())
.zip(def_hists.into_iter())
.map(
|(
((((min, max), is_null), null_count), repetition_level_histogram),
definition_level_histogram,
)| {
let (min, max) = if is_null {
(None, None)
} else {
(
Some(T::try_from_le_slice(min)?),
Some(T::try_from_le_slice(max)?),
)
};
Ok(PageIndex {
min,
max,
null_count,
repetition_level_histogram,
definition_level_histogram,
})
},
)
.collect::<Result<Vec<_>, ParquetError>>()?;
Ok(Self {
indexes,
boundary_order: index.boundary_order,
})
}
pub(crate) fn to_thrift(&self) -> ColumnIndex {
let min_values = self
.indexes
.iter()
.map(|x| x.min_bytes().map(|x| x.to_vec()))
.collect::<Option<Vec<_>>>()
.unwrap_or_else(|| vec![vec![]; self.indexes.len()]);
let max_values = self
.indexes
.iter()
.map(|x| x.max_bytes().map(|x| x.to_vec()))
.collect::<Option<Vec<_>>>()
.unwrap_or_else(|| vec![vec![]; self.indexes.len()]);
let null_counts = self
.indexes
.iter()
.map(|x| x.null_count())
.collect::<Option<Vec<_>>>();
// Concatenate page histograms into a single Option<Vec>
let repetition_level_histograms = self
.indexes
.iter()
.map(|x| x.repetition_level_histogram().map(|v| v.values()))
.collect::<Option<Vec<&[i64]>>>()
.map(|hists| hists.concat());
let definition_level_histograms = self
.indexes
.iter()
.map(|x| x.definition_level_histogram().map(|v| v.values()))
.collect::<Option<Vec<&[i64]>>>()
.map(|hists| hists.concat());
ColumnIndex::new(
self.indexes.iter().map(|x| x.min().is_none()).collect(),
min_values,
max_values,
self.boundary_order,
null_counts,
repetition_level_histograms,
definition_level_histograms,
)
}
}
#[cfg(test)]
mod tests {
use super::*;
#[test]
fn test_page_index_min_max_null() {
let page_index = PageIndex {
min: Some(-123),
max: Some(234),
null_count: Some(0),
repetition_level_histogram: Some(LevelHistogram::from(vec![1, 2])),
definition_level_histogram: Some(LevelHistogram::from(vec![1, 2, 3])),
};
assert_eq!(page_index.min().unwrap(), &-123);
assert_eq!(page_index.max().unwrap(), &234);
assert_eq!(page_index.min_bytes().unwrap(), (-123).as_bytes());
assert_eq!(page_index.max_bytes().unwrap(), 234.as_bytes());
assert_eq!(page_index.null_count().unwrap(), 0);
assert_eq!(
page_index.repetition_level_histogram().unwrap().values(),
&vec![1, 2]
);
assert_eq!(
page_index.definition_level_histogram().unwrap().values(),
&vec![1, 2, 3]
);
}
#[test]
fn test_page_index_min_max_null_none() {
let page_index: PageIndex<i32> = PageIndex {
min: None,
max: None,
null_count: None,
repetition_level_histogram: None,
definition_level_histogram: None,
};
assert_eq!(page_index.min(), None);
assert_eq!(page_index.max(), None);
assert_eq!(page_index.min_bytes(), None);
assert_eq!(page_index.max_bytes(), None);
assert_eq!(page_index.null_count(), None);
assert_eq!(page_index.repetition_level_histogram(), None);
assert_eq!(page_index.definition_level_histogram(), None);
}
#[test]
fn test_invalid_column_index() {
let column_index = ColumnIndex {
null_pages: vec![true, false],
min_values: vec![
vec![],
vec![], // this shouldn't be empty as null_pages[1] is false
],
max_values: vec![
vec![],
vec![], // this shouldn't be empty as null_pages[1] is false
],
null_counts: None,
repetition_level_histograms: None,
definition_level_histograms: None,
boundary_order: BoundaryOrder::UNORDERED,
};
let err = NativeIndex::<i32>::try_new(column_index).unwrap_err();
assert_eq!(
err.to_string(),
"Parquet error: error converting value, expected 4 bytes got 0"
);
}
}