Compare commits

...

1 Commits

Author SHA1 Message Date
Paul Masurel
7d551fb580 Refactoring dyn Column 2022-09-02 00:41:46 +09:00
30 changed files with 403 additions and 484 deletions

View File

@@ -7,11 +7,12 @@
// Of course, you can have a look at the tantivy's built-in collectors
// such as the `CountCollector` for more examples.
use std::sync::Arc;
use fastfield_codecs::Column;
// ---
// Importing tantivy...
use tantivy::collector::{Collector, SegmentCollector};
use tantivy::fastfield::DynamicFastFieldReader;
use tantivy::query::QueryParser;
use tantivy::schema::{Field, Schema, FAST, INDEXED, TEXT};
use tantivy::{doc, Index, Score, SegmentReader};
@@ -96,7 +97,7 @@ impl Collector for StatsCollector {
}
struct StatsSegmentCollector {
fast_field_reader: DynamicFastFieldReader<u64>,
fast_field_reader: Arc<dyn Column<u64>>,
stats: Stats,
}

View File

@@ -2,7 +2,6 @@ use std::cmp::Reverse;
use std::collections::{HashMap, HashSet};
use std::sync::{Arc, RwLock, Weak};
use fastfield_codecs::Column;
use tantivy::collector::TopDocs;
use tantivy::query::QueryParser;
use tantivy::schema::{Field, Schema, FAST, TEXT};

View File

@@ -1,3 +1,5 @@
use std::marker::PhantomData;
pub trait Column<T = u64> {
/// Return the value associated to the given idx.
///
@@ -42,8 +44,103 @@ pub trait Column<T = u64> {
fn max_value(&self) -> T;
fn num_vals(&self) -> u64;
/// Returns a iterator over the data
fn iter<'a>(&'a self) -> Box<dyn Iterator<Item = T> + 'a> {
Box::new((0..self.num_vals()).map(|idx| self.get_val(idx)))
}
}
struct VecColumn<'a>(&'a [u64]);
impl<'a> Column for VecColumn<'a> {
fn get_val(&self, position: u64) -> u64 {
self.0[position as usize]
}
fn iter<'b>(&'b self) -> Box<dyn Iterator<Item = u64> + 'b> {
Box::new(self.0.iter().cloned())
}
fn min_value(&self) -> u64 {
self.0.iter().min().cloned().unwrap_or(0)
}
fn max_value(&self) -> u64 {
self.0.iter().max().cloned().unwrap_or(0)
}
fn num_vals(&self) -> u64 {
self.0.len() as u64
}
}
impl<'a> From<&'a [u64]> for VecColumn<'a> {
fn from(data: &'a [u64]) -> Self {
Self(data)
}
}
struct MonotonicMappingColumn<C, T, Input> {
from_column: C,
monotonic_mapping: T,
_phantom: PhantomData<Input>,
}
/// Creates a view of a column transformed by a monotonic mapping.
pub fn monotonic_map_column<C, T, Input, Output>(
from_column: C,
monotonic_mapping: T,
) -> impl Column<Output>
where
C: Column<Input>,
T: Fn(Input) -> Output,
{
MonotonicMappingColumn {
from_column,
monotonic_mapping,
_phantom: PhantomData,
}
}
impl<C, T, Input, Output> Column<Output> for MonotonicMappingColumn<C, T, Input>
where
C: Column<Input>,
T: Fn(Input) -> Output,
{
fn get_val(&self, idx: u64) -> Output {
let from_val = self.from_column.get_val(idx);
(self.monotonic_mapping)(from_val)
}
fn min_value(&self) -> Output {
let from_min_value = self.from_column.min_value();
(self.monotonic_mapping)(from_min_value)
}
fn max_value(&self) -> Output {
let from_max_value = self.from_column.max_value();
(self.monotonic_mapping)(from_max_value)
}
fn num_vals(&self) -> u64 {
self.from_column.num_vals()
}
}
#[cfg(test)]
mod tests {
use super::*;
#[test]
fn test_monotonic_mapping() {
let vals = &[1u64, 3u64][..];
let col = VecColumn::from(vals);
let mapped = monotonic_map_column(col, |el| el + 4);
assert_eq!(mapped.min_value(), 5u64);
assert_eq!(mapped.max_value(), 7u64);
assert_eq!(mapped.num_vals(), 2);
assert_eq!(mapped.num_vals(), 2);
assert_eq!(mapped.get_val(0), 5);
assert_eq!(mapped.get_val(0), 7);
}
}

View File

@@ -14,7 +14,7 @@ pub mod linear;
mod column;
pub use self::column::Column;
pub use self::column::{monotonic_map_column, Column};
#[derive(PartialEq, Eq, PartialOrd, Ord, Debug, Clone, Copy)]
#[repr(u8)]
@@ -56,12 +56,12 @@ impl FastFieldCodecType {
/// The FastFieldSerializerEstimate trait is required on all variants
/// of fast field compressions, to decide which one to choose.
pub trait FastFieldCodec {
pub trait FastFieldCodec: 'static {
/// A codex needs to provide a unique name and id, which is
/// used for debugging and de/serialization.
const CODEC_TYPE: FastFieldCodecType;
type Reader: Column<u64>;
type Reader: Column<u64> + 'static;
/// Reads the metadata and returns the CodecReader
fn open_from_bytes(bytes: OwnedBytes) -> io::Result<Self::Reader>;
@@ -90,35 +90,6 @@ pub struct FastFieldStats {
pub num_vals: u64,
}
struct VecColum<'a>(&'a [u64]);
impl<'a> Column for VecColum<'a> {
fn get_val(&self, position: u64) -> u64 {
self.0[position as usize]
}
fn iter<'b>(&'b self) -> Box<dyn Iterator<Item = u64> + 'b> {
Box::new(self.0.iter().cloned())
}
fn min_value(&self) -> u64 {
self.0.iter().min().cloned().unwrap_or(0)
}
fn max_value(&self) -> u64 {
self.0.iter().max().cloned().unwrap_or(0)
}
fn num_vals(&self) -> u64 {
self.0.len() as u64
}
}
impl<'a> From<&'a [u64]> for VecColum<'a> {
fn from(data: &'a [u64]) -> Self {
Self(data)
}
}
#[cfg(test)]
mod tests {
use proptest::arbitrary::any;

View File

@@ -4,14 +4,14 @@ use std::rc::Rc;
use std::sync::atomic::AtomicU32;
use std::sync::Arc;
use fastfield_codecs::Column;
use super::agg_req::{Aggregation, Aggregations, BucketAggregationType, MetricAggregation};
use super::bucket::{HistogramAggregation, RangeAggregation, TermsAggregation};
use super::metric::{AverageAggregation, StatsAggregation};
use super::segment_agg_result::BucketCount;
use super::VecWithNames;
use crate::fastfield::{
type_and_cardinality, DynamicFastFieldReader, FastType, MultiValuedFastFieldReader,
};
use crate::fastfield::{type_and_cardinality, FastType, MultiValuedFastFieldReader};
use crate::schema::{Cardinality, Type};
use crate::{InvertedIndexReader, SegmentReader, TantivyError};
@@ -37,10 +37,16 @@ impl AggregationsWithAccessor {
#[derive(Clone)]
pub(crate) enum FastFieldAccessor {
Multi(MultiValuedFastFieldReader<u64>),
Single(DynamicFastFieldReader<u64>),
Single(Arc<dyn Column<u64>>),
}
impl FastFieldAccessor {
pub fn as_single(&self) -> Option<&DynamicFastFieldReader<u64>> {
pub fn as_single(&self) -> Option<&dyn Column<u64>> {
match self {
FastFieldAccessor::Multi(_) => None,
FastFieldAccessor::Single(reader) => Some(&**reader),
}
}
pub fn into_single(self) -> Option<Arc<dyn Column<u64>>> {
match self {
FastFieldAccessor::Multi(_) => None,
FastFieldAccessor::Single(reader) => Some(reader),
@@ -118,7 +124,7 @@ impl BucketAggregationWithAccessor {
pub struct MetricAggregationWithAccessor {
pub metric: MetricAggregation,
pub field_type: Type,
pub accessor: DynamicFastFieldReader<u64>,
pub accessor: Arc<dyn Column>,
}
impl MetricAggregationWithAccessor {
@@ -134,9 +140,8 @@ impl MetricAggregationWithAccessor {
Ok(MetricAggregationWithAccessor {
accessor: accessor
.as_single()
.expect("unexpected fast field cardinality")
.clone(),
.into_single()
.expect("unexpected fast field cardinality"),
field_type,
metric: metric.clone(),
})

View File

@@ -15,7 +15,6 @@ use crate::aggregation::intermediate_agg_result::{
IntermediateAggregationResults, IntermediateBucketResult, IntermediateHistogramBucketEntry,
};
use crate::aggregation::segment_agg_result::SegmentAggregationResultsCollector;
use crate::fastfield::DynamicFastFieldReader;
use crate::schema::Type;
use crate::{DocId, TantivyError};
@@ -264,7 +263,7 @@ impl SegmentHistogramCollector {
req: &HistogramAggregation,
sub_aggregation: &AggregationsWithAccessor,
field_type: Type,
accessor: &DynamicFastFieldReader<u64>,
accessor: &dyn Column<u64>,
) -> crate::Result<Self> {
req.validate()?;
let min = f64_from_fastfield_u64(accessor.min_value(), &field_type);

View File

@@ -1,7 +1,6 @@
use std::fmt::Debug;
use std::ops::Range;
use fastfield_codecs::Column;
use fnv::FnvHashMap;
use serde::{Deserialize, Serialize};

View File

@@ -4,7 +4,6 @@ use fastfield_codecs::Column;
use serde::{Deserialize, Serialize};
use crate::aggregation::f64_from_fastfield_u64;
use crate::fastfield::DynamicFastFieldReader;
use crate::schema::Type;
use crate::DocId;
@@ -58,7 +57,7 @@ impl SegmentAverageCollector {
data: Default::default(),
}
}
pub(crate) fn collect_block(&mut self, doc: &[DocId], field: &DynamicFastFieldReader<u64>) {
pub(crate) fn collect_block(&mut self, doc: &[DocId], field: &dyn Column<u64>) {
let mut iter = doc.chunks_exact(4);
for docs in iter.by_ref() {
let val1 = field.get_val(docs[0] as u64);

View File

@@ -2,7 +2,6 @@ use fastfield_codecs::Column;
use serde::{Deserialize, Serialize};
use crate::aggregation::f64_from_fastfield_u64;
use crate::fastfield::DynamicFastFieldReader;
use crate::schema::Type;
use crate::{DocId, TantivyError};
@@ -164,7 +163,7 @@ impl SegmentStatsCollector {
stats: IntermediateStats::default(),
}
}
pub(crate) fn collect_block(&mut self, doc: &[DocId], field: &DynamicFastFieldReader<u64>) {
pub(crate) fn collect_block(&mut self, doc: &[DocId], field: &dyn Column<u64>) {
let mut iter = doc.chunks_exact(4);
for docs in iter.by_ref() {
let val1 = field.get_val(docs[0] as u64);

View File

@@ -185,10 +185,10 @@ impl SegmentMetricResultCollector {
pub(crate) fn collect_block(&mut self, doc: &[DocId], metric: &MetricAggregationWithAccessor) {
match self {
SegmentMetricResultCollector::Average(avg_collector) => {
avg_collector.collect_block(doc, &metric.accessor);
avg_collector.collect_block(doc, &*metric.accessor);
}
SegmentMetricResultCollector::Stats(stats_collector) => {
stats_collector.collect_block(doc, &metric.accessor);
stats_collector.collect_block(doc, &*metric.accessor);
}
}
}

View File

@@ -10,11 +10,12 @@
// ---
// Importing tantivy...
use std::marker::PhantomData;
use std::sync::Arc;
use fastfield_codecs::Column;
use crate::collector::{Collector, SegmentCollector};
use crate::fastfield::{DynamicFastFieldReader, FastValue};
use crate::fastfield::FastValue;
use crate::schema::Field;
use crate::{Score, SegmentReader, TantivyError};
@@ -160,7 +161,7 @@ where
TPredicate: 'static,
TPredicateValue: FastValue,
{
fast_field_reader: DynamicFastFieldReader<TPredicateValue>,
fast_field_reader: Arc<dyn Column<TPredicateValue>>,
segment_collector: TSegmentCollector,
predicate: TPredicate,
t_predicate_value: PhantomData<TPredicateValue>,

View File

@@ -1,8 +1,10 @@
use std::sync::Arc;
use fastdivide::DividerU64;
use fastfield_codecs::Column;
use crate::collector::{Collector, SegmentCollector};
use crate::fastfield::{DynamicFastFieldReader, FastValue};
use crate::fastfield::FastValue;
use crate::schema::{Field, Type};
use crate::{DocId, Score};
@@ -85,7 +87,7 @@ impl HistogramComputer {
}
pub struct SegmentHistogramCollector {
histogram_computer: HistogramComputer,
ff_reader: DynamicFastFieldReader<u64>,
ff_reader: Arc<dyn Column<u64>>,
}
impl SegmentCollector for SegmentHistogramCollector {

View File

@@ -1,9 +1,11 @@
use std::sync::Arc;
use fastfield_codecs::Column;
use super::*;
use crate::collector::{Count, FilterCollector, TopDocs};
use crate::core::SegmentReader;
use crate::fastfield::{BytesFastFieldReader, DynamicFastFieldReader};
use crate::fastfield::BytesFastFieldReader;
use crate::query::{AllQuery, QueryParser};
use crate::schema::{Field, Schema, FAST, TEXT};
use crate::time::format_description::well_known::Rfc3339;
@@ -158,7 +160,7 @@ pub struct FastFieldTestCollector {
pub struct FastFieldSegmentCollector {
vals: Vec<u64>,
reader: DynamicFastFieldReader<u64>,
reader: Arc<dyn Column<u64>>,
}
impl FastFieldTestCollector {

View File

@@ -1,6 +1,7 @@
use std::collections::BinaryHeap;
use std::fmt;
use std::marker::PhantomData;
use std::sync::Arc;
use fastfield_codecs::Column;
@@ -11,7 +12,7 @@ use crate::collector::tweak_score_top_collector::TweakedScoreTopCollector;
use crate::collector::{
CustomScorer, CustomSegmentScorer, ScoreSegmentTweaker, ScoreTweaker, SegmentCollector,
};
use crate::fastfield::{DynamicFastFieldReader, FastValue};
use crate::fastfield::FastValue;
use crate::query::Weight;
use crate::schema::Field;
use crate::{DocAddress, DocId, Score, SegmentOrdinal, SegmentReader, TantivyError};
@@ -131,7 +132,7 @@ impl fmt::Debug for TopDocs {
}
struct ScorerByFastFieldReader {
ff_reader: DynamicFastFieldReader<u64>,
ff_reader: Arc<dyn Column<u64>>,
}
impl CustomSegmentScorer<u64> for ScorerByFastFieldReader {
@@ -409,7 +410,6 @@ impl TopDocs {
/// # use tantivy::query::QueryParser;
/// use tantivy::SegmentReader;
/// use tantivy::collector::TopDocs;
/// use tantivy::fastfield::Column;
/// use tantivy::schema::Field;
///
/// fn create_schema() -> Schema {
@@ -517,7 +517,6 @@ impl TopDocs {
/// use tantivy::SegmentReader;
/// use tantivy::collector::TopDocs;
/// use tantivy::schema::Field;
/// use fastfield_codecs::Column;
///
/// # fn create_schema() -> Schema {
/// # let mut schema_builder = Schema::builder();

View File

@@ -1,7 +1,9 @@
use std::sync::Arc;
use fastfield_codecs::Column;
use crate::directory::{FileSlice, OwnedBytes};
use crate::fastfield::{DynamicFastFieldReader, MultiValueLength};
use crate::fastfield::MultiValueLength;
use crate::DocId;
/// Reader for byte array fast fields
@@ -16,13 +18,13 @@ use crate::DocId;
/// and the start index for the next document, and keeping the bytes in between.
#[derive(Clone)]
pub struct BytesFastFieldReader {
idx_reader: DynamicFastFieldReader<u64>,
idx_reader: Arc<dyn Column<u64>>,
values: OwnedBytes,
}
impl BytesFastFieldReader {
pub(crate) fn open(
idx_reader: DynamicFastFieldReader<u64>,
idx_reader: Arc<dyn Column<u64>>,
values_file: FileSlice,
) -> crate::Result<BytesFastFieldReader> {
let values = values_file.read_bytes()?;

View File

@@ -3,20 +3,11 @@ use std::num::NonZeroU64;
use common::BinarySerializable;
use fastdivide::DividerU64;
use fastfield_codecs::{Column, FastFieldCodec};
use fastfield_codecs::{monotonic_map_column, Column, FastFieldCodec};
use ownedbytes::OwnedBytes;
pub const GCD_DEFAULT: u64 = 1;
/// Wrapper for accessing a fastfield.
///
/// Holds the data and the codec to the read the data.
#[derive(Clone)]
pub struct GCDReader<CodecReader: Column> {
gcd_params: GCDParams,
reader: CodecReader,
}
#[derive(Debug, Clone, Copy)]
struct GCDParams {
gcd: u64,
@@ -24,12 +15,6 @@ struct GCDParams {
num_vals: u64,
}
impl GCDParams {
pub fn eval(&self, val: u64) -> u64 {
self.min_value + self.gcd * val
}
}
impl BinarySerializable for GCDParams {
fn serialize<W: Write>(&self, writer: &mut W) -> io::Result<()> {
self.gcd.serialize(writer)?;
@@ -52,31 +37,13 @@ impl BinarySerializable for GCDParams {
pub fn open_gcd_from_bytes<WrappedCodec: FastFieldCodec>(
bytes: OwnedBytes,
) -> io::Result<GCDReader<WrappedCodec::Reader>> {
) -> io::Result<impl Column> {
let footer_offset = bytes.len() - 24;
let (body, mut footer) = bytes.split(footer_offset);
let gcd_params = GCDParams::deserialize(&mut footer)?;
let gcd_remap = move |val: u64| gcd_params.min_value + gcd_params.gcd * val;
let reader: WrappedCodec::Reader = WrappedCodec::open_from_bytes(body)?;
Ok(GCDReader { gcd_params, reader })
}
impl<C: Column + Clone> Column for GCDReader<C> {
#[inline]
fn get_val(&self, doc: u64) -> u64 {
let val = self.reader.get_val(doc);
self.gcd_params.eval(val)
}
fn min_value(&self) -> u64 {
self.gcd_params.eval(self.reader.min_value())
}
fn max_value(&self) -> u64 {
self.gcd_params.eval(self.reader.max_value())
}
fn num_vals(&self) -> u64 {
self.gcd_params.num_vals
}
Ok(monotonic_map_column(reader, gcd_remap))
}
pub fn write_gcd_header<W: Write>(
@@ -134,6 +101,7 @@ mod tests {
use std::collections::HashMap;
use std::num::NonZeroU64;
use std::path::Path;
use std::sync::Arc;
use std::time::{Duration, SystemTime};
use common::HasLen;
@@ -141,11 +109,11 @@ mod tests {
use crate::directory::{CompositeFile, RamDirectory, WritePtr};
use crate::fastfield::gcd::compute_gcd;
use crate::fastfield::reader::open_fast_field;
use crate::fastfield::serializer::FastFieldCodecEnableCheck;
use crate::fastfield::tests::{FIELD, FIELDI64, SCHEMA, SCHEMAI64};
use crate::fastfield::tests::{encode_decode_fast_field, FIELD, FIELDI64, SCHEMA, SCHEMAI64};
use crate::fastfield::{
find_gcd, CompositeFastFieldSerializer, DynamicFastFieldReader, FastFieldCodecType,
FastFieldsWriter, ALL_CODECS,
find_gcd, CompositeFastFieldSerializer, FastFieldCodecType, FastFieldsWriter, ALL_CODECS,
};
use crate::schema::{Cardinality, Schema};
use crate::{DateOptions, DatePrecision, DateTime, Directory};
@@ -187,8 +155,7 @@ mod tests {
let file = directory.open_read(path).unwrap();
let composite_file = CompositeFile::open(&file)?;
let file = composite_file.open_read(*FIELD).unwrap();
let fast_field_reader = DynamicFastFieldReader::<i64>::open(file)?;
let fast_field_reader: Arc<dyn Column<i64>> = open_fast_field(file.read_bytes()?)?;
assert_eq!(fast_field_reader.get_val(0), -4000i64);
assert_eq!(fast_field_reader.get_val(1), -3000i64);
assert_eq!(fast_field_reader.get_val(2), -2000i64);
@@ -229,7 +196,7 @@ mod tests {
let file = directory.open_read(path).unwrap();
let composite_file = CompositeFile::open(&file)?;
let file = composite_file.open_read(*FIELD).unwrap();
let fast_field_reader = DynamicFastFieldReader::<u64>::open(file)?;
let fast_field_reader = open_fast_field::<u64>(file.read_bytes()?)?;
assert_eq!(fast_field_reader.get_val(0), 1000u64);
assert_eq!(fast_field_reader.get_val(1), 2000u64);
assert_eq!(fast_field_reader.get_val(2), 3000u64);
@@ -258,7 +225,7 @@ mod tests {
#[test]
pub fn test_fastfield2() {
let test_fastfield = DynamicFastFieldReader::<u64>::from(vec![100, 200, 300]);
let test_fastfield = encode_decode_fast_field(&[100u64, 200u64, 300u64]);
assert_eq!(test_fastfield.get_val(0), 100);
assert_eq!(test_fastfield.get_val(1), 200);
assert_eq!(test_fastfield.get_val(2), 300);
@@ -324,7 +291,7 @@ mod tests {
let composite_file = CompositeFile::open(&file)?;
let file = composite_file.open_read(*FIELD).unwrap();
let len = file.len();
let test_fastfield = DynamicFastFieldReader::<DateTime>::open(file)?;
let test_fastfield = open_fast_field::<DateTime>(file.read_bytes()?)?;
assert_eq!(test_fastfield.get_val(0), time1.truncate(precision));
assert_eq!(test_fastfield.get_val(1), time2.truncate(precision));

View File

@@ -26,9 +26,8 @@ pub use self::alive_bitset::{intersect_alive_bitsets, write_alive_bitset, AliveB
pub use self::bytes::{BytesFastFieldReader, BytesFastFieldWriter};
pub use self::error::{FastFieldNotAvailableError, Result};
pub use self::facet_reader::FacetReader;
pub(crate) use self::gcd::{find_gcd, GCDReader, GCD_DEFAULT};
pub(crate) use self::gcd::{find_gcd, GCD_DEFAULT};
pub use self::multivalued::{MultiValuedFastFieldReader, MultiValuedFastFieldWriter};
pub use self::reader::DynamicFastFieldReader;
pub use self::readers::FastFieldReaders;
pub(crate) use self::readers::{type_and_cardinality, FastType};
pub use self::serializer::{Column, CompositeFastFieldSerializer, FastFieldStats};
@@ -266,6 +265,7 @@ mod tests {
use std::collections::HashMap;
use std::ops::Range;
use std::path::Path;
use std::sync::Arc;
use common::HasLen;
use once_cell::sync::Lazy;
@@ -275,6 +275,7 @@ mod tests {
use super::*;
use crate::directory::{CompositeFile, Directory, RamDirectory, WritePtr};
use crate::fastfield::reader::open_fast_field;
use crate::merge_policy::NoMergePolicy;
use crate::schema::{Document, Field, Schema, FAST, STRING, TEXT};
use crate::time::OffsetDateTime;
@@ -295,9 +296,51 @@ mod tests {
pub static FIELD: Lazy<Field> = Lazy::new(|| SCHEMA.get_field("field").unwrap());
pub static FIELDI64: Lazy<Field> = Lazy::new(|| SCHEMAI64.get_field("field").unwrap());
/// Encode values using the most appropriate codec and and then loads it
/// right away.
///
/// This is useful in tests and bench.
pub(crate) fn encode_decode_fast_field<Item: FastValue>(
vals: &[Item],
) -> Arc<dyn Column<Item>> {
let mut schema_builder = Schema::builder();
let field = schema_builder.add_u64_field("field", FAST);
let schema = schema_builder.build();
let path = Path::new("__dummy__");
let directory: RamDirectory = RamDirectory::create();
{
let write: WritePtr = directory
.open_write(path)
.expect("With a RamDirectory, this should never fail.");
let mut serializer = CompositeFastFieldSerializer::from_write(write)
.expect("With a RamDirectory, this should never fail.");
let mut fast_field_writers = FastFieldsWriter::from_schema(&schema);
{
let fast_field_writer = fast_field_writers
.get_field_writer_mut(field)
.expect("With a RamDirectory, this should never fail.");
for val in vals {
fast_field_writer.add_val(val.to_u64());
}
}
fast_field_writers
.serialize(&mut serializer, &HashMap::new(), None)
.unwrap();
serializer.close().unwrap();
}
let file = directory.open_read(path).expect("Failed to open the file");
let composite_file = CompositeFile::open(&file).expect("Failed to read the composite file");
let field_bytes = composite_file
.open_read(field)
.expect("File component not found")
.read_bytes()
.unwrap();
open_fast_field(field_bytes).unwrap()
}
#[test]
pub fn test_fastfield() {
let test_fastfield = DynamicFastFieldReader::<u64>::from(vec![100, 200, 300]);
let test_fastfield = encode_decode_fast_field(&[100u64, 200u64, 300u64]);
assert_eq!(test_fastfield.get_val(0u64), 100);
assert_eq!(test_fastfield.get_val(1u64), 200);
assert_eq!(test_fastfield.get_val(2u64), 300);
@@ -328,8 +371,8 @@ mod tests {
let file = directory.open_read(path).unwrap();
assert_eq!(file.len(), 45);
let composite_file = CompositeFile::open(&file)?;
let file = composite_file.open_read(*FIELD).unwrap();
let fast_field_reader = DynamicFastFieldReader::<u64>::open(file)?;
let fast_field_bytes = composite_file.open_read(*FIELD).unwrap().read_bytes()?;
let fast_field_reader = open_fast_field::<u64>(fast_field_bytes)?;
assert_eq!(fast_field_reader.get_val(0), 13u64);
assert_eq!(fast_field_reader.get_val(1), 14u64);
assert_eq!(fast_field_reader.get_val(2), 2u64);
@@ -360,8 +403,11 @@ mod tests {
assert_eq!(file.len(), 70);
{
let fast_fields_composite = CompositeFile::open(&file)?;
let data = fast_fields_composite.open_read(*FIELD).unwrap();
let fast_field_reader = DynamicFastFieldReader::<u64>::open(data)?;
let data = fast_fields_composite
.open_read(*FIELD)
.unwrap()
.read_bytes()?;
let fast_field_reader = open_fast_field::<u64>(data)?;
assert_eq!(fast_field_reader.get_val(0), 4u64);
assert_eq!(fast_field_reader.get_val(1), 14_082_001u64);
assert_eq!(fast_field_reader.get_val(2), 3_052u64);
@@ -396,8 +442,11 @@ mod tests {
assert_eq!(file.len(), 43);
{
let fast_fields_composite = CompositeFile::open(&file).unwrap();
let data = fast_fields_composite.open_read(*FIELD).unwrap();
let fast_field_reader = DynamicFastFieldReader::<u64>::open(data)?;
let data = fast_fields_composite
.open_read(*FIELD)
.unwrap()
.read_bytes()?;
let fast_field_reader = open_fast_field::<u64>(data)?;
for doc in 0..10_000 {
assert_eq!(fast_field_reader.get_val(doc), 100_000u64);
}
@@ -428,8 +477,11 @@ mod tests {
assert_eq!(file.len(), 80051);
{
let fast_fields_composite = CompositeFile::open(&file)?;
let data = fast_fields_composite.open_read(*FIELD).unwrap();
let fast_field_reader = DynamicFastFieldReader::<u64>::open(data)?;
let data = fast_fields_composite
.open_read(*FIELD)
.unwrap()
.read_bytes()?;
let fast_field_reader = open_fast_field::<u64>(data)?;
assert_eq!(fast_field_reader.get_val(0), 0u64);
for doc in 1..10_001 {
assert_eq!(
@@ -469,8 +521,11 @@ mod tests {
assert_eq!(file.len(), 75_usize); // linear interpol size after calc improvement
{
let fast_fields_composite = CompositeFile::open(&file)?;
let data = fast_fields_composite.open_read(i64_field).unwrap();
let fast_field_reader = DynamicFastFieldReader::<i64>::open(data)?;
let data = fast_fields_composite
.open_read(i64_field)
.unwrap()
.read_bytes()?;
let fast_field_reader = open_fast_field::<i64>(data)?;
assert_eq!(fast_field_reader.min_value(), -100i64);
assert_eq!(fast_field_reader.max_value(), 9_999i64);
@@ -509,8 +564,11 @@ mod tests {
let file = directory.open_read(path).unwrap();
{
let fast_fields_composite = CompositeFile::open(&file).unwrap();
let data = fast_fields_composite.open_read(i64_field).unwrap();
let fast_field_reader = DynamicFastFieldReader::<i64>::open(data)?;
let data = fast_fields_composite
.open_read(i64_field)
.unwrap()
.read_bytes()?;
let fast_field_reader = open_fast_field::<i64>(data)?;
assert_eq!(fast_field_reader.get_val(0), 0i64);
}
Ok(())
@@ -547,8 +605,11 @@ mod tests {
let file = directory.open_read(path)?;
{
let fast_fields_composite = CompositeFile::open(&file)?;
let data = fast_fields_composite.open_read(*FIELD).unwrap();
let fast_field_reader = DynamicFastFieldReader::<u64>::open(data)?;
let data = fast_fields_composite
.open_read(*FIELD)
.unwrap()
.read_bytes()?;
let fast_field_reader = open_fast_field::<u64>(data)?;
for a in 0..n {
assert_eq!(fast_field_reader.get_val(a as u64), permutation[a as usize]);
@@ -607,7 +668,7 @@ mod tests {
let mut all = vec![];
for doc in docs {
let mut out = vec![];
let mut out: Vec<u64> = vec![];
ff.get_vals(doc, &mut out);
all.extend(out);
}
@@ -865,7 +926,7 @@ mod tests {
#[test]
pub fn test_fastfield_bool() {
let test_fastfield = DynamicFastFieldReader::<bool>::from(vec![true, false, true, false]);
let test_fastfield = encode_decode_fast_field::<bool>(&[true, false, true, false]);
assert_eq!(test_fastfield.get_val(0), true);
assert_eq!(test_fastfield.get_val(1), false);
assert_eq!(test_fastfield.get_val(2), true);
@@ -898,8 +959,8 @@ mod tests {
let file = directory.open_read(path).unwrap();
assert_eq!(file.len(), 44);
let composite_file = CompositeFile::open(&file)?;
let file = composite_file.open_read(field).unwrap();
let fast_field_reader = DynamicFastFieldReader::<bool>::open(file)?;
let data = composite_file.open_read(field).unwrap().read_bytes()?;
let fast_field_reader = open_fast_field::<bool>(data)?;
assert_eq!(fast_field_reader.get_val(0), true);
assert_eq!(fast_field_reader.get_val(1), false);
assert_eq!(fast_field_reader.get_val(2), true);
@@ -934,8 +995,8 @@ mod tests {
let file = directory.open_read(path).unwrap();
assert_eq!(file.len(), 56);
let composite_file = CompositeFile::open(&file)?;
let file = composite_file.open_read(field).unwrap();
let fast_field_reader = DynamicFastFieldReader::<bool>::open(file)?;
let data = composite_file.open_read(field).unwrap().read_bytes()?;
let fast_field_reader = open_fast_field::<bool>(data)?;
for i in 0..25 {
assert_eq!(fast_field_reader.get_val(i * 2), true);
assert_eq!(fast_field_reader.get_val(i * 2 + 1), false);
@@ -968,8 +1029,8 @@ mod tests {
let file = directory.open_read(path).unwrap();
assert_eq!(file.len(), 43);
let composite_file = CompositeFile::open(&file)?;
let file = composite_file.open_read(field).unwrap();
let fast_field_reader = DynamicFastFieldReader::<bool>::open(file)?;
let data = composite_file.open_read(field).unwrap().read_bytes()?;
let fast_field_reader = open_fast_field::<bool>(data)?;
assert_eq!(fast_field_reader.get_val(0), false);
Ok(())
@@ -978,32 +1039,17 @@ mod tests {
#[cfg(all(test, feature = "unstable"))]
mod bench {
use std::collections::HashMap;
use std::path::Path;
use std::sync::Arc;
use fastfield_codecs::Column;
use test::{self, Bencher};
use super::tests::{generate_permutation, FIELD, SCHEMA};
use super::*;
use crate::directory::{CompositeFile, Directory, RamDirectory, WritePtr};
use crate::fastfield::tests::generate_permutation_gcd;
use crate::fastfield::tests::{
encode_decode_fast_field, generate_permutation, generate_permutation_gcd,
};
#[bench]
fn bench_intfastfield_linear_veclookup(b: &mut Bencher) {
let permutation = generate_permutation();
b.iter(|| {
let n = test::black_box(7000u32);
let mut a = 0u64;
for i in (0u32..n / 7).map(|v| v * 7) {
a ^= permutation[i as usize];
}
a
});
}
#[bench]
fn bench_intfastfield_veclookup(b: &mut Bencher) {
fn bench_intfastfield_jumpy_veclookup(b: &mut Bencher) {
let permutation = generate_permutation();
b.iter(|| {
let n = test::black_box(1000u32);
@@ -1016,102 +1062,81 @@ mod bench {
}
#[bench]
fn bench_intfastfield_linear_fflookup(b: &mut Bencher) {
let path = Path::new("test");
fn bench_intfastfield_jumpy_fflookup(b: &mut Bencher) {
let permutation = generate_permutation();
let directory: RamDirectory = RamDirectory::create();
{
let write: WritePtr = directory.open_write(Path::new("test")).unwrap();
let mut serializer = CompositeFastFieldSerializer::from_write(write).unwrap();
let mut fast_field_writers = FastFieldsWriter::from_schema(&SCHEMA);
for &x in &permutation {
fast_field_writers.add_document(&doc!(*FIELD=>x));
let column: Arc<dyn Column<u64>> = encode_decode_fast_field(&permutation);
b.iter(|| {
let n = test::black_box(1000u32);
let mut a = 0u64;
for _ in 0u32..n {
a = column.get_val(a as u64);
}
fast_field_writers
.serialize(&mut serializer, &HashMap::new(), None)
.unwrap();
serializer.close().unwrap();
}
let file = directory.open_read(&path).unwrap();
{
let fast_fields_composite = CompositeFile::open(&file).unwrap();
let data = fast_fields_composite.open_read(*FIELD).unwrap();
let fast_field_reader = DynamicFastFieldReader::<u64>::open(data).unwrap();
a
});
}
b.iter(|| {
let n = test::black_box(7000u32);
let mut a = 0u64;
for i in (0u32..n / 7).map(|val| val * 7) {
a ^= fast_field_reader.get_val(i as u64);
}
a
});
}
#[bench]
fn bench_intfastfield_linear_fflookup(b: &mut Bencher) {
let permutation = generate_permutation();
let column: Arc<dyn Column<u64>> = encode_decode_fast_field(&permutation);
b.iter(|| {
let n = test::black_box(7000u32);
let mut a = 0u64;
for i in (0..n / 7).map(|val| val * 7) {
a += column.get_val(i as u64);
}
a
});
}
#[bench]
fn bench_intfastfield_linear_vec(b: &mut Bencher) {
let permutation = generate_permutation();
b.iter(|| {
let n = test::black_box(7000);
let mut a = 0u64;
for i in (0..n / 7).map(|val| val * 7) {
a += permutation[i];
}
a
});
}
#[bench]
fn bench_intfastfield_fflookup(b: &mut Bencher) {
let path = Path::new("test");
let permutation = generate_permutation();
let directory: RamDirectory = RamDirectory::create();
{
let write: WritePtr = directory.open_write(Path::new("test")).unwrap();
let mut serializer = CompositeFastFieldSerializer::from_write(write).unwrap();
let mut fast_field_writers = FastFieldsWriter::from_schema(&SCHEMA);
for &x in &permutation {
fast_field_writers.add_document(&doc!(*FIELD=>x));
let column: Arc<dyn Column<u64>> = encode_decode_fast_field(&permutation);
b.iter(|| {
let mut a = 0u64;
for i in 0u64..permutation.len() as u64 {
a = column.get_val(i);
}
fast_field_writers
.serialize(&mut serializer, &HashMap::new(), None)
.unwrap();
serializer.close().unwrap();
}
let file = directory.open_read(&path).unwrap();
{
let fast_fields_composite = CompositeFile::open(&file).unwrap();
let data = fast_fields_composite.open_read(*FIELD).unwrap();
let fast_field_reader = DynamicFastFieldReader::<u64>::open(data).unwrap();
b.iter(|| {
let mut a = 0u32;
for i in 0u64..permutation.len() as u64 {
a = fast_field_reader.get_val(i) as u32;
}
a
});
}
a
});
}
#[bench]
fn bench_intfastfield_fflookup_gcd(b: &mut Bencher) {
let path = Path::new("test");
let permutation = generate_permutation_gcd();
let directory: RamDirectory = RamDirectory::create();
{
let write: WritePtr = directory.open_write(Path::new("test")).unwrap();
let mut serializer = CompositeFastFieldSerializer::from_write(write).unwrap();
let mut fast_field_writers = FastFieldsWriter::from_schema(&SCHEMA);
for &x in &permutation {
fast_field_writers.add_document(&doc!(*FIELD=>x));
let column: Arc<dyn Column<u64>> = encode_decode_fast_field(&permutation);
b.iter(|| {
let mut a = 0u64;
for i in 0..permutation.len() as u64 {
a += column.get_val(i);
}
fast_field_writers
.serialize(&mut serializer, &HashMap::new(), None)
.unwrap();
serializer.close().unwrap();
}
let file = directory.open_read(&path).unwrap();
{
let fast_fields_composite = CompositeFile::open(&file).unwrap();
let data = fast_fields_composite.open_read(*FIELD).unwrap();
let fast_field_reader = DynamicFastFieldReader::<u64>::open(data).unwrap();
a
});
}
b.iter(|| {
let mut a = 0u32;
for i in 0u32..permutation.len() as u32 {
a = fast_field_reader.get_val(i as u64) as u32;
}
a
});
}
#[bench]
fn bench_intfastfield_vec(b: &mut Bencher) {
let permutation = generate_permutation_gcd();
b.iter(|| {
let mut a = 0u64;
for i in 0..permutation.len() {
a += permutation[i as usize] as u64;
}
a
});
}
}

View File

@@ -346,6 +346,7 @@ mod tests {
assert!(test_multivalued_no_panic(&ops[..]).is_ok());
}
}
#[test]
fn test_multivalued_proptest_gcd() {
use IndexingOp::*;

View File

@@ -1,8 +1,9 @@
use std::ops::Range;
use std::sync::Arc;
use fastfield_codecs::Column;
use crate::fastfield::{DynamicFastFieldReader, FastValue, MultiValueLength};
use crate::fastfield::{FastValue, MultiValueLength};
use crate::DocId;
/// Reader for a multivalued `u64` fast field.
@@ -14,14 +15,14 @@ use crate::DocId;
/// The `idx_reader` associated, for each document, the index of its first value.
#[derive(Clone)]
pub struct MultiValuedFastFieldReader<Item: FastValue> {
idx_reader: DynamicFastFieldReader<u64>,
vals_reader: DynamicFastFieldReader<Item>,
idx_reader: Arc<dyn Column<u64>>,
vals_reader: Arc<dyn Column<Item>>,
}
impl<Item: FastValue> MultiValuedFastFieldReader<Item> {
pub(crate) fn open(
idx_reader: DynamicFastFieldReader<u64>,
vals_reader: DynamicFastFieldReader<Item>,
idx_reader: Arc<dyn Column<u64>>,
vals_reader: Arc<dyn Column<Item>>,
) -> MultiValuedFastFieldReader<Item> {
MultiValuedFastFieldReader {
idx_reader,

View File

@@ -1,144 +1,65 @@
use std::collections::HashMap;
use std::marker::PhantomData;
use std::path::Path;
use std::sync::Arc;
use common::BinarySerializable;
use fastfield_codecs::bitpacked::{BitpackedCodec, BitpackedReader};
use fastfield_codecs::blockwise_linear::{BlockwiseLinearCodec, BlockwiseLinearReader};
use fastfield_codecs::linear::{LinearCodec, LinearReader};
use fastfield_codecs::{Column, FastFieldCodec, FastFieldCodecType};
use fastfield_codecs::bitpacked::BitpackedCodec;
use fastfield_codecs::blockwise_linear::BlockwiseLinearCodec;
use fastfield_codecs::linear::LinearCodec;
use fastfield_codecs::{monotonic_map_column, Column, FastFieldCodec, FastFieldCodecType};
use super::gcd::open_gcd_from_bytes;
use super::FastValue;
use crate::directory::{CompositeFile, Directory, FileSlice, OwnedBytes, RamDirectory, WritePtr};
use crate::directory::OwnedBytes;
use crate::error::DataCorruption;
use crate::fastfield::{CompositeFastFieldSerializer, FastFieldsWriter, GCDReader};
use crate::schema::{Schema, FAST};
#[derive(Clone)]
/// DynamicFastFieldReader wraps different readers to access
/// the various encoded fastfield data
pub enum DynamicFastFieldReader<Item: FastValue> {
/// Bitpacked compressed fastfield data.
Bitpacked(FastFieldReaderCodecWrapper<Item, BitpackedReader>),
/// Linear interpolated values + bitpacked
Linear(FastFieldReaderCodecWrapper<Item, LinearReader>),
/// Blockwise linear interpolated values + bitpacked
BlockwiseLinear(FastFieldReaderCodecWrapper<Item, BlockwiseLinearReader>),
/// GCD and Bitpacked compressed fastfield data.
BitpackedGCD(FastFieldReaderCodecWrapper<Item, GCDReader<BitpackedReader>>),
/// GCD and Linear interpolated values + bitpacked
LinearGCD(FastFieldReaderCodecWrapper<Item, GCDReader<LinearReader>>),
/// GCD and Blockwise linear interpolated values + bitpacked
BlockwiseLinearGCD(FastFieldReaderCodecWrapper<Item, GCDReader<BlockwiseLinearReader>>),
fn open_codec_from_bytes<C: FastFieldCodec, Item: FastValue>(
bytes: OwnedBytes,
) -> crate::Result<Arc<dyn Column<Item>>> {
let reader = C::open_from_bytes(bytes)?;
Ok(Arc::new(monotonic_map_column(reader, Item::from_u64)))
}
impl<Item: FastValue> DynamicFastFieldReader<Item> {
/// Returns correct the reader wrapped in the `DynamicFastFieldReader` enum for the data.
pub fn open_from_id(
mut bytes: OwnedBytes,
codec_type: FastFieldCodecType,
) -> crate::Result<DynamicFastFieldReader<Item>> {
let reader = match codec_type {
FastFieldCodecType::Bitpacked => {
DynamicFastFieldReader::Bitpacked(BitpackedCodec::open_from_bytes(bytes)?.into())
}
FastFieldCodecType::Linear => {
DynamicFastFieldReader::Linear(LinearCodec::open_from_bytes(bytes)?.into())
}
FastFieldCodecType::BlockwiseLinear => DynamicFastFieldReader::BlockwiseLinear(
BlockwiseLinearCodec::open_from_bytes(bytes)?.into(),
),
FastFieldCodecType::Gcd => {
let codec_type = FastFieldCodecType::deserialize(&mut bytes)?;
match codec_type {
FastFieldCodecType::Bitpacked => DynamicFastFieldReader::BitpackedGCD(
open_gcd_from_bytes::<BitpackedCodec>(bytes)?.into(),
),
FastFieldCodecType::Linear => DynamicFastFieldReader::LinearGCD(
open_gcd_from_bytes::<LinearCodec>(bytes)?.into(),
),
FastFieldCodecType::BlockwiseLinear => {
DynamicFastFieldReader::BlockwiseLinearGCD(
open_gcd_from_bytes::<BlockwiseLinearCodec>(bytes)?.into(),
)
}
FastFieldCodecType::Gcd => {
return Err(DataCorruption::comment_only(
"Gcd codec wrapped into another gcd codec. This combination is not \
allowed.",
)
.into())
}
fn open_codec_with_gcd<C: FastFieldCodec, Item: FastValue>(
bytes: OwnedBytes,
) -> crate::Result<Arc<dyn Column<Item>>> {
let reader = open_gcd_from_bytes::<C>(bytes)?;
Ok(Arc::new(monotonic_map_column(reader, Item::from_u64)))
}
/// Returns correct the reader wrapped in the `DynamicFastFieldReader` enum for the data.
fn open_from_id<Item: FastValue>(
mut bytes: OwnedBytes,
codec_type: FastFieldCodecType,
) -> crate::Result<Arc<dyn Column<Item>>> {
match codec_type {
FastFieldCodecType::Bitpacked => open_codec_from_bytes::<BitpackedCodec, _>(bytes),
FastFieldCodecType::Linear => open_codec_from_bytes::<LinearCodec, _>(bytes),
FastFieldCodecType::BlockwiseLinear => {
open_codec_from_bytes::<BlockwiseLinearCodec, _>(bytes)
}
FastFieldCodecType::Gcd => {
let codec_type = FastFieldCodecType::deserialize(&mut bytes)?;
match codec_type {
FastFieldCodecType::Bitpacked => open_codec_with_gcd::<BitpackedCodec, _>(bytes),
FastFieldCodecType::Linear => open_codec_with_gcd::<LinearCodec, _>(bytes),
FastFieldCodecType::BlockwiseLinear => {
open_codec_with_gcd::<BlockwiseLinearCodec, _>(bytes)
}
FastFieldCodecType::Gcd => Err(DataCorruption::comment_only(
"Gcd codec wrapped into another gcd codec. This combination is not allowed.",
)
.into()),
}
};
Ok(reader)
}
/// Returns correct the reader wrapped in the `DynamicFastFieldReader` enum for the data.
pub fn open(file: FileSlice) -> crate::Result<DynamicFastFieldReader<Item>> {
let mut bytes = file.read_bytes()?;
let codec_type = FastFieldCodecType::deserialize(&mut bytes)?;
Self::open_from_id(bytes, codec_type)
}
}
}
impl<Item: FastValue> Column<Item> for DynamicFastFieldReader<Item> {
#[inline]
fn get_val(&self, idx: u64) -> Item {
match self {
Self::Bitpacked(reader) => reader.get_val(idx),
Self::Linear(reader) => reader.get_val(idx),
Self::BlockwiseLinear(reader) => reader.get_val(idx),
Self::BitpackedGCD(reader) => reader.get_val(idx),
Self::LinearGCD(reader) => reader.get_val(idx),
Self::BlockwiseLinearGCD(reader) => reader.get_val(idx),
}
}
#[inline]
fn get_range(&self, start: u64, output: &mut [Item]) {
match self {
Self::Bitpacked(reader) => reader.get_range(start, output),
Self::Linear(reader) => reader.get_range(start, output),
Self::BlockwiseLinear(reader) => reader.get_range(start, output),
Self::BitpackedGCD(reader) => reader.get_range(start, output),
Self::LinearGCD(reader) => reader.get_range(start, output),
Self::BlockwiseLinearGCD(reader) => reader.get_range(start, output),
}
}
fn min_value(&self) -> Item {
match self {
Self::Bitpacked(reader) => reader.min_value(),
Self::Linear(reader) => reader.min_value(),
Self::BlockwiseLinear(reader) => reader.min_value(),
Self::BitpackedGCD(reader) => reader.min_value(),
Self::LinearGCD(reader) => reader.min_value(),
Self::BlockwiseLinearGCD(reader) => reader.min_value(),
}
}
fn max_value(&self) -> Item {
match self {
Self::Bitpacked(reader) => reader.max_value(),
Self::Linear(reader) => reader.max_value(),
Self::BlockwiseLinear(reader) => reader.max_value(),
Self::BitpackedGCD(reader) => reader.max_value(),
Self::LinearGCD(reader) => reader.max_value(),
Self::BlockwiseLinearGCD(reader) => reader.max_value(),
}
}
fn num_vals(&self) -> u64 {
match self {
Self::Bitpacked(reader) => reader.num_vals(),
Self::Linear(reader) => reader.num_vals(),
Self::BlockwiseLinear(reader) => reader.num_vals(),
Self::BitpackedGCD(reader) => reader.num_vals(),
Self::LinearGCD(reader) => reader.num_vals(),
Self::BlockwiseLinearGCD(reader) => reader.num_vals(),
}
}
/// Returns correct the reader wrapped in the `DynamicFastFieldReader` enum for the data.
pub fn open_fast_field<Item: FastValue>(
mut bytes: OwnedBytes,
) -> crate::Result<Arc<dyn Column<Item>>> {
let codec_type = FastFieldCodecType::deserialize(&mut bytes)?;
open_from_id(bytes, codec_type)
}
/// Wrapper for accessing a fastfield.
@@ -239,40 +160,3 @@ impl<Item: FastValue, C: Column + Clone> Column<Item> for FastFieldReaderCodecWr
self.reader.num_vals()
}
}
impl<Item: FastValue> From<Vec<Item>> for DynamicFastFieldReader<Item> {
fn from(vals: Vec<Item>) -> DynamicFastFieldReader<Item> {
let mut schema_builder = Schema::builder();
let field = schema_builder.add_u64_field("field", FAST);
let schema = schema_builder.build();
let path = Path::new("__dummy__");
let directory: RamDirectory = RamDirectory::create();
{
let write: WritePtr = directory
.open_write(path)
.expect("With a RamDirectory, this should never fail.");
let mut serializer = CompositeFastFieldSerializer::from_write(write)
.expect("With a RamDirectory, this should never fail.");
let mut fast_field_writers = FastFieldsWriter::from_schema(&schema);
{
let fast_field_writer = fast_field_writers
.get_field_writer_mut(field)
.expect("With a RamDirectory, this should never fail.");
for val in vals {
fast_field_writer.add_val(val.to_u64());
}
}
fast_field_writers
.serialize(&mut serializer, &HashMap::new(), None)
.unwrap();
serializer.close().unwrap();
}
let file = directory.open_read(path).expect("Failed to open the file");
let composite_file = CompositeFile::open(&file).expect("Failed to read the composite file");
let field_file = composite_file
.open_read(field)
.expect("File component not found");
DynamicFastFieldReader::open(field_file).unwrap()
}
}

View File

@@ -1,5 +1,9 @@
use super::reader::DynamicFastFieldReader;
use std::sync::Arc;
use fastfield_codecs::Column;
use crate::directory::{CompositeFile, FileSlice};
use crate::fastfield::reader::open_fast_field;
use crate::fastfield::{
BytesFastFieldReader, FastFieldNotAvailableError, FastValue, MultiValuedFastFieldReader,
};
@@ -109,14 +113,16 @@ impl FastFieldReaders {
&self,
field: Field,
index: usize,
) -> crate::Result<DynamicFastFieldReader<TFastValue>> {
) -> crate::Result<Arc<dyn Column<TFastValue>>> {
let fast_field_slice = self.fast_field_data(field, index)?;
DynamicFastFieldReader::open(fast_field_slice)
let bytes = fast_field_slice.read_bytes()?;
open_fast_field(bytes)
}
pub(crate) fn typed_fast_field_reader<TFastValue: FastValue>(
&self,
field: Field,
) -> crate::Result<DynamicFastFieldReader<TFastValue>> {
) -> crate::Result<Arc<dyn Column<TFastValue>>> {
self.typed_fast_field_reader_with_idx(field, 0)
}
@@ -132,7 +138,7 @@ impl FastFieldReaders {
/// Returns the `u64` fast field reader reader associated to `field`.
///
/// If `field` is not a u64 fast field, this method returns an Error.
pub fn u64(&self, field: Field) -> crate::Result<DynamicFastFieldReader<u64>> {
pub fn u64(&self, field: Field) -> crate::Result<Arc<dyn Column<u64>>> {
self.check_type(field, FastType::U64, Cardinality::SingleValue)?;
self.typed_fast_field_reader(field)
}
@@ -142,14 +148,14 @@ impl FastFieldReaders {
///
/// If not, the fastfield reader will returns the u64-value associated to the original
/// FastValue.
pub fn u64_lenient(&self, field: Field) -> crate::Result<DynamicFastFieldReader<u64>> {
pub fn u64_lenient(&self, field: Field) -> crate::Result<Arc<dyn Column<u64>>> {
self.typed_fast_field_reader(field)
}
/// Returns the `i64` fast field reader reader associated to `field`.
///
/// If `field` is not a i64 fast field, this method returns an Error.
pub fn i64(&self, field: Field) -> crate::Result<DynamicFastFieldReader<i64>> {
pub fn i64(&self, field: Field) -> crate::Result<Arc<dyn Column<i64>>> {
self.check_type(field, FastType::I64, Cardinality::SingleValue)?;
self.typed_fast_field_reader(field)
}
@@ -157,7 +163,7 @@ impl FastFieldReaders {
/// Returns the `date` fast field reader reader associated to `field`.
///
/// If `field` is not a date fast field, this method returns an Error.
pub fn date(&self, field: Field) -> crate::Result<DynamicFastFieldReader<DateTime>> {
pub fn date(&self, field: Field) -> crate::Result<Arc<dyn Column<DateTime>>> {
self.check_type(field, FastType::Date, Cardinality::SingleValue)?;
self.typed_fast_field_reader(field)
}
@@ -165,7 +171,7 @@ impl FastFieldReaders {
/// Returns the `f64` fast field reader reader associated to `field`.
///
/// If `field` is not a f64 fast field, this method returns an Error.
pub fn f64(&self, field: Field) -> crate::Result<DynamicFastFieldReader<f64>> {
pub fn f64(&self, field: Field) -> crate::Result<Arc<dyn Column<f64>>> {
self.check_type(field, FastType::F64, Cardinality::SingleValue)?;
self.typed_fast_field_reader(field)
}
@@ -173,7 +179,7 @@ impl FastFieldReaders {
/// Returns the `bool` fast field reader reader associated to `field`.
///
/// If `field` is not a bool fast field, this method returns an Error.
pub fn bool(&self, field: Field) -> crate::Result<DynamicFastFieldReader<bool>> {
pub fn bool(&self, field: Field) -> crate::Result<Arc<dyn Column<bool>>> {
self.check_type(field, FastType::Bool, Cardinality::SingleValue)?;
self.typed_fast_field_reader(field)
}
@@ -241,7 +247,8 @@ impl FastFieldReaders {
)));
}
let fast_field_idx_file = self.fast_field_data(field, 0)?;
let idx_reader = DynamicFastFieldReader::open(fast_field_idx_file)?;
let fast_field_idx_bytes = fast_field_idx_file.read_bytes()?;
let idx_reader = open_fast_field(fast_field_idx_bytes)?;
let data = self.fast_field_data(field, 1)?;
BytesFastFieldReader::open(idx_reader, data)
} else {

View File

@@ -6,7 +6,7 @@ use fastdivide::DividerU64;
pub use fastfield_codecs::bitpacked::{BitpackedCodec, BitpackedSerializerLegacy};
use fastfield_codecs::blockwise_linear::BlockwiseLinearCodec;
use fastfield_codecs::linear::LinearCodec;
use fastfield_codecs::FastFieldCodecType;
use fastfield_codecs::{monotonic_map_column, FastFieldCodecType};
pub use fastfield_codecs::{Column, FastFieldCodec, FastFieldStats};
use super::{find_gcd, ALL_CODECS, GCD_DEFAULT};
@@ -136,56 +136,22 @@ impl CompositeFastFieldSerializer {
}
Self::write_header(field_write, FastFieldCodecType::Gcd)?;
struct GCDWrappedFFAccess<T: Column> {
fastfield_accessor: T,
base_value: u64,
max_value: u64,
num_vals: u64,
gcd: DividerU64,
}
impl<T: Column> Column for GCDWrappedFFAccess<T> {
fn get_val(&self, position: u64) -> u64 {
self.gcd
.divide(self.fastfield_accessor.get_val(position) - self.base_value)
}
fn iter(&self) -> Box<dyn Iterator<Item = u64> + '_> {
Box::new(
self.fastfield_accessor
.iter()
.map(|val| self.gcd.divide(val - self.base_value)),
)
}
fn min_value(&self) -> u64 {
0
}
fn max_value(&self) -> u64 {
self.max_value
}
fn num_vals(&self) -> u64 {
self.num_vals
}
}
let num_vals = fastfield_accessor.num_vals();
let base_value = fastfield_accessor.min_value();
let max_value = (fastfield_accessor.max_value() - fastfield_accessor.min_value()) / gcd;
let fastfield_accessor = GCDWrappedFFAccess {
fastfield_accessor,
base_value,
max_value,
num_vals,
gcd: DividerU64::divide_by(gcd),
};
let gcd_divider = DividerU64::divide_by(gcd);
let divided_fastfield_accessor = monotonic_map_column(fastfield_accessor, |val: u64| {
gcd_divider.divide(val - base_value)
});
let num_vals = divided_fastfield_accessor.num_vals();
Self::create_auto_detect_u64_fast_field_with_idx_gcd(
self.codec_enable_checker.clone(),
field,
field_write,
fastfield_accessor,
divided_fastfield_accessor,
)?;
write_gcd_header(field_write, base_value, gcd, num_vals)?;
Ok(())

View File

@@ -143,8 +143,6 @@ pub(crate) fn get_doc_id_mapping_from_field(
#[cfg(test)]
mod tests_indexsorting {
use fastfield_codecs::Column;
use crate::collector::TopDocs;
use crate::indexer::doc_id_mapping::DocIdMapping;
use crate::query::QueryParser;

View File

@@ -775,7 +775,6 @@ impl Drop for IndexWriter {
mod tests {
use std::collections::{HashMap, HashSet};
use fastfield_codecs::Column;
use proptest::prelude::*;
use proptest::prop_oneof;
use proptest::strategy::Strategy;

View File

@@ -9,8 +9,8 @@ use crate::core::{Segment, SegmentReader};
use crate::docset::{DocSet, TERMINATED};
use crate::error::DataCorruption;
use crate::fastfield::{
AliveBitSet, Column, CompositeFastFieldSerializer, DynamicFastFieldReader, FastFieldStats,
MultiValueLength, MultiValuedFastFieldReader,
AliveBitSet, Column, CompositeFastFieldSerializer, FastFieldStats, MultiValueLength,
MultiValuedFastFieldReader,
};
use crate::fieldnorm::{FieldNormReader, FieldNormReaders, FieldNormsSerializer, FieldNormsWriter};
use crate::indexer::doc_id_mapping::{expect_field_id_for_sort_field, SegmentDocIdMapping};
@@ -87,7 +87,7 @@ pub struct IndexMerger {
}
fn compute_min_max_val(
u64_reader: &impl Column<u64>,
u64_reader: &dyn Column<u64>,
segment_reader: &SegmentReader,
) -> Option<(u64, u64)> {
if segment_reader.max_doc() == 0 {
@@ -341,12 +341,12 @@ impl IndexMerger {
.readers
.iter()
.filter_map(|reader| {
let u64_reader: DynamicFastFieldReader<u64> =
let u64_reader: Arc<dyn Column<u64>> =
reader.fast_fields().typed_fast_field_reader(field).expect(
"Failed to find a reader for single fast field. This is a tantivy bug and \
it should never happen.",
);
compute_min_max_val(&u64_reader, reader)
compute_min_max_val(&*u64_reader, reader)
})
.reduce(|a, b| (a.0.min(b.0), a.1.max(b.1)))
.expect("Unexpected error, empty readers in IndexMerger");
@@ -355,7 +355,7 @@ impl IndexMerger {
.readers
.iter()
.map(|reader| {
let u64_reader: DynamicFastFieldReader<u64> =
let u64_reader: Arc<dyn Column<u64>> =
reader.fast_fields().typed_fast_field_reader(field).expect(
"Failed to find a reader for single fast field. This is a tantivy bug and \
it should never happen.",
@@ -372,7 +372,7 @@ impl IndexMerger {
#[derive(Clone)]
struct SortedDocIdFieldAccessProvider<'a> {
doc_id_mapping: &'a SegmentDocIdMapping,
fast_field_readers: &'a Vec<DynamicFastFieldReader<u64>>,
fast_field_readers: &'a Vec<Arc<dyn Column<u64>>>,
stats: FastFieldStats,
}
impl<'a> Column for SortedDocIdFieldAccessProvider<'a> {
@@ -443,7 +443,7 @@ impl IndexMerger {
pub(crate) fn get_sort_field_accessor(
reader: &SegmentReader,
sort_by_field: &IndexSortByField,
) -> crate::Result<impl Column> {
) -> crate::Result<Arc<dyn Column>> {
let field_id = expect_field_id_for_sort_field(reader.schema(), sort_by_field)?; // for now expect fastfield, but not strictly required
let value_accessor = reader.fast_fields().u64_lenient(field_id)?;
Ok(value_accessor)
@@ -452,7 +452,7 @@ impl IndexMerger {
pub(crate) fn get_reader_with_sort_field_accessor(
&self,
sort_by_field: &IndexSortByField,
) -> crate::Result<Vec<(SegmentOrdinal, impl Column)>> {
) -> crate::Result<Vec<(SegmentOrdinal, Arc<dyn Column>)>> {
let reader_ordinal_and_field_accessors = self
.readers
.iter()
@@ -618,7 +618,7 @@ impl IndexMerger {
.map(|reader| {
let u64s_reader: MultiValuedFastFieldReader<u64> = reader
.fast_fields()
.typed_fast_field_multi_reader(field)
.typed_fast_field_multi_reader::<u64>(field)
.expect(
"Failed to find index for multivalued field. This is a bug in tantivy, \
please report.",
@@ -668,7 +668,7 @@ impl IndexMerger {
{
let mut serialize_vals =
fast_field_serializer.new_u64_fast_field_with_idx(field, 0u64, max_term_ord, 1)?;
let mut vals = Vec::with_capacity(100);
let mut vals: Vec<u64> = Vec::with_capacity(100);
for old_doc_addr in doc_id_mapping.iter_old_doc_addrs() {
let term_ordinal_mapping: &[TermOrdinal] =
@@ -742,7 +742,7 @@ impl IndexMerger {
for reader in &self.readers {
let ff_reader: MultiValuedFastFieldReader<u64> = reader
.fast_fields()
.typed_fast_field_multi_reader(field)
.typed_fast_field_multi_reader::<u64>(field)
.expect(
"Failed to find multivalued fast field reader. This is a bug in tantivy. \
Please report.",
@@ -1199,7 +1199,6 @@ impl IndexMerger {
#[cfg(test)]
mod tests {
use byteorder::{BigEndian, ReadBytesExt};
use fastfield_codecs::Column;
use schema::FAST;
use crate::collector::tests::{

View File

@@ -1,7 +1,5 @@
#[cfg(test)]
mod tests {
use fastfield_codecs::Column;
use crate::collector::TopDocs;
use crate::core::Index;
use crate::fastfield::{AliveBitSet, MultiValuedFastFieldReader};
@@ -480,11 +478,12 @@ mod tests {
#[cfg(all(test, feature = "unstable"))]
mod bench_sorted_index_merge {
use std::sync::Arc;
use fastfield_codecs::Column;
use test::{self, Bencher};
use crate::core::Index;
use crate::fastfield::DynamicFastFieldReader;
use crate::indexer::merger::IndexMerger;
use crate::schema::{Cardinality, NumericOptions, Schema};
use crate::{IndexSettings, IndexSortByField, IndexWriter, Order};
@@ -536,7 +535,7 @@ mod bench_sorted_index_merge {
b.iter(|| {
let sorted_doc_ids = doc_id_mapping.iter_old_doc_addrs().map(|doc_addr| {
let reader = &merger.readers[doc_addr.segment_ord as usize];
let u64_reader: DynamicFastFieldReader<u64> =
let u64_reader: Arc<dyn Column<u64>> =
reader.fast_fields().typed_fast_field_reader(field).expect(
"Failed to find a reader for single fast field. This is a tantivy bug and \
it should never happen.",

View File

@@ -421,7 +421,6 @@ pub struct DocAddress {
#[cfg(test)]
pub mod tests {
use common::{BinarySerializable, FixedSize};
use fastfield_codecs::Column;
use rand::distributions::{Bernoulli, Uniform};
use rand::rngs::StdRng;
use rand::{Rng, SeedableRng};

View File

@@ -339,7 +339,7 @@ impl StoreReader {
async fn read_block_async(&self, checkpoint: &Checkpoint) -> crate::AsyncIoResult<Block> {
let cache_key = checkpoint.byte_range.start;
if let Some(block) = self.cache.get_from_cache(checkpoint.byte_range.start) {
return Ok(block.clone());
return Ok(block);
}
let compressed_block = self

View File

@@ -172,8 +172,7 @@ where TValueReader: value::ValueReader
}
pub fn suffix(&self) -> &[u8] {
&self
.block_reader
self.block_reader
.buffer_from_to(self.suffix_start, self.suffix_end)
}

View File

@@ -50,7 +50,7 @@ pub struct SSTableIndexBuilder {
/// matches `left <= left' < right`.
fn find_shorter_str_in_between(left: &mut Vec<u8>, right: &[u8]) {
assert!(&left[..] < right);
let common_len = common_prefix_len(&left, right);
let common_len = common_prefix_len(left, right);
if left.len() == common_len {
return;
}