Update extract_facet_number_docids to support deladd obkvs

This commit is contained in:
Clément Renault
2023-10-18 17:40:13 +02:00
committed by Louis Dureuil
parent a82dee21e0
commit fcd3a1434d
4 changed files with 57 additions and 15 deletions

View File

@ -4,11 +4,12 @@ use std::io::{self, BufReader};
use heed::{BytesDecode, BytesEncode};
use super::helpers::{
create_sorter, merge_cbo_roaring_bitmaps, sorter_into_reader, GrenadParameters,
create_sorter, merge_deladd_cbo_roaring_bitmaps, sorter_into_reader, GrenadParameters,
};
use crate::heed_codec::facet::{
FacetGroupKey, FacetGroupKeyCodec, FieldDocIdFacetF64Codec, OrderedF64Codec,
};
use crate::update::del_add::{KvReaderDelAdd, KvWriterDelAdd};
use crate::Result;
/// Extracts the facet number and the documents ids where this facet number appear.
@ -17,8 +18,7 @@ use crate::Result;
/// documents ids from the given chunk of docid facet number positions.
#[logging_timer::time]
pub fn extract_facet_number_docids<R: io::Read + io::Seek>(
// TODO Reader<Key, Obkv<DelAdd, ()>>
docid_fid_facet_number: grenad::Reader<R>,
fid_docid_facet_number: grenad::Reader<R>,
indexer: GrenadParameters,
) -> Result<grenad::Reader<BufReader<File>>> {
puffin::profile_function!();
@ -27,24 +27,30 @@ pub fn extract_facet_number_docids<R: io::Read + io::Seek>(
let mut facet_number_docids_sorter = create_sorter(
grenad::SortAlgorithm::Unstable,
// TODO We must modify the merger to do unions of Del and Add separately
merge_cbo_roaring_bitmaps,
merge_deladd_cbo_roaring_bitmaps,
indexer.chunk_compression_type,
indexer.chunk_compression_level,
indexer.max_nb_chunks,
max_memory,
);
let mut cursor = docid_fid_facet_number.into_cursor()?;
// TODO the value is a Obkv<DelAdd, ()> and must be taken into account
while let Some((key_bytes, _)) = cursor.move_on_next()? {
let mut buffer = Vec::new();
let mut cursor = fid_docid_facet_number.into_cursor()?;
while let Some((key_bytes, deladd_obkv_bytes)) = cursor.move_on_next()? {
let (field_id, document_id, number) =
FieldDocIdFacetF64Codec::bytes_decode(key_bytes).unwrap();
let key = FacetGroupKey { field_id, level: 0, left_bound: number };
let key_bytes = FacetGroupKeyCodec::<OrderedF64Codec>::bytes_encode(&key).unwrap();
// TODO We must put a Obkv<DelAdd, RoaringBitmap>
facet_number_docids_sorter.insert(key_bytes, document_id.to_ne_bytes())?;
buffer.clear();
let mut obkv = KvWriterDelAdd::new(&mut buffer);
for (deladd_key, _) in KvReaderDelAdd::new(deladd_obkv_bytes).iter() {
obkv.insert(deladd_key, document_id.to_ne_bytes())?;
}
obkv.finish()?;
facet_number_docids_sorter.insert(key_bytes, &buffer)?;
}
sorter_into_reader(facet_number_docids_sorter, indexer)