diff --git a/milli/src/update/index_documents/extract/extract_fid_docid_facet_values.rs b/milli/src/update/index_documents/extract/extract_fid_docid_facet_values.rs index 7d366ae1a..846a023e7 100644 --- a/milli/src/update/index_documents/extract/extract_fid_docid_facet_values.rs +++ b/milli/src/update/index_documents/extract/extract_fid_docid_facet_values.rs @@ -29,9 +29,8 @@ const TRUNCATE_SIZE: usize = size_of::() + size_of::(); /// The extracted facet values stored in grenad files by type. pub struct ExtractedFacetValues { - // TOOD rename into `fid_docid_*` - pub docid_fid_facet_numbers_chunk: grenad::Reader, - pub docid_fid_facet_strings_chunk: grenad::Reader, + pub fid_docid_facet_numbers_chunk: grenad::Reader, + pub fid_docid_facet_strings_chunk: grenad::Reader, pub fid_facet_is_null_docids_chunk: grenad::Reader, pub fid_facet_is_empty_docids_chunk: grenad::Reader, pub fid_facet_exists_docids_chunk: grenad::Reader, @@ -44,7 +43,6 @@ pub struct ExtractedFacetValues { /// We need the fid of the geofields to correctly parse them as numbers if they were sent as strings initially. #[logging_timer::time] pub fn extract_fid_docid_facet_values( - // TODO Reader>> obkv_documents: grenad::Reader, indexer: GrenadParameters, faceted_fields: &HashSet, @@ -83,7 +81,6 @@ pub fn extract_fid_docid_facet_values( let mut cursor = obkv_documents.into_cursor()?; while let Some((docid_bytes, value)) = cursor.move_on_next()? { - // TODO Obkv> let obkv = obkv::KvReader::new(value); for (field_id, field_bytes) in obkv.iter() { @@ -96,7 +93,6 @@ pub fn extract_fid_docid_facet_values( numbers_key_buffer.extend_from_slice(&field_id.to_be_bytes()); strings_key_buffer.extend_from_slice(&field_id.to_be_bytes()); - // Here, we know already that the document must be added to the “field id exists” database let document: [u8; 4] = docid_bytes[..4].try_into().ok().unwrap(); let document = BEU32::from(document).get(); @@ -260,8 +256,8 @@ pub fn extract_fid_docid_facet_values( let facet_is_empty_docids_reader = writer_into_reader(facet_is_empty_docids_writer)?; Ok(ExtractedFacetValues { - docid_fid_facet_numbers_chunk: sorter_into_reader(fid_docid_facet_numbers_sorter, indexer)?, - docid_fid_facet_strings_chunk: sorter_into_reader(fid_docid_facet_strings_sorter, indexer)?, + fid_docid_facet_numbers_chunk: sorter_into_reader(fid_docid_facet_numbers_sorter, indexer)?, + fid_docid_facet_strings_chunk: sorter_into_reader(fid_docid_facet_strings_sorter, indexer)?, fid_facet_is_null_docids_chunk: facet_is_null_docids_reader, fid_facet_is_empty_docids_chunk: facet_is_empty_docids_reader, fid_facet_exists_docids_chunk: facet_exists_docids_reader, diff --git a/milli/src/update/index_documents/extract/mod.rs b/milli/src/update/index_documents/extract/mod.rs index f28af8c2a..3eb46e4a7 100644 --- a/milli/src/update/index_documents/extract/mod.rs +++ b/milli/src/update/index_documents/extract/mod.rs @@ -91,9 +91,9 @@ pub(crate) fn data_from_obkv_documents( let ( docid_word_positions_chunks, ( - docid_fid_facet_numbers_chunks, + fid_docid_facet_numbers_chunks, ( - docid_fid_facet_strings_chunks, + fid_docid_facet_strings_chunks, ( facet_is_null_docids_chunks, (facet_is_empty_docids_chunks, facet_exists_docids_chunks), @@ -201,7 +201,7 @@ pub(crate) fn data_from_obkv_documents( ); spawn_extraction_task::<_, _, Vec>>( - docid_fid_facet_strings_chunks, + fid_docid_facet_strings_chunks, indexer, lmdb_writer_sx.clone(), extract_facet_string_docids, @@ -211,7 +211,7 @@ pub(crate) fn data_from_obkv_documents( ); spawn_extraction_task::<_, _, Vec>>( - docid_fid_facet_numbers_chunks, + fid_docid_facet_numbers_chunks, indexer, lmdb_writer_sx, extract_facet_number_docids, @@ -344,7 +344,7 @@ fn send_and_extract_flattened_documents_data( }); } - let (docid_word_positions_chunk, docid_fid_facet_values_chunks): (Result<_>, Result<_>) = + let (docid_word_positions_chunk, fid_docid_facet_values_chunks): (Result<_>, Result<_>) = rayon::join( || { let (documents_ids, docid_word_positions_chunk, script_language_pair) = @@ -372,8 +372,8 @@ fn send_and_extract_flattened_documents_data( }, || { let ExtractedFacetValues { - docid_fid_facet_numbers_chunk, - docid_fid_facet_strings_chunk, + fid_docid_facet_numbers_chunk, + fid_docid_facet_strings_chunk, fid_facet_is_null_docids_chunk, fid_facet_is_empty_docids_chunk, fid_facet_exists_docids_chunk, @@ -384,26 +384,26 @@ fn send_and_extract_flattened_documents_data( geo_fields_ids, )?; - // send docid_fid_facet_numbers_chunk to DB writer - let docid_fid_facet_numbers_chunk = - unsafe { as_cloneable_grenad(&docid_fid_facet_numbers_chunk)? }; + // send fid_docid_facet_numbers_chunk to DB writer + let fid_docid_facet_numbers_chunk = + unsafe { as_cloneable_grenad(&fid_docid_facet_numbers_chunk)? }; let _ = lmdb_writer_sx.send(Ok(TypedChunk::FieldIdDocidFacetNumbers( - docid_fid_facet_numbers_chunk.clone(), + fid_docid_facet_numbers_chunk.clone(), ))); - // send docid_fid_facet_strings_chunk to DB writer - let docid_fid_facet_strings_chunk = - unsafe { as_cloneable_grenad(&docid_fid_facet_strings_chunk)? }; + // send fid_docid_facet_strings_chunk to DB writer + let fid_docid_facet_strings_chunk = + unsafe { as_cloneable_grenad(&fid_docid_facet_strings_chunk)? }; let _ = lmdb_writer_sx.send(Ok(TypedChunk::FieldIdDocidFacetStrings( - docid_fid_facet_strings_chunk.clone(), + fid_docid_facet_strings_chunk.clone(), ))); Ok(( - docid_fid_facet_numbers_chunk, + fid_docid_facet_numbers_chunk, ( - docid_fid_facet_strings_chunk, + fid_docid_facet_strings_chunk, ( fid_facet_is_null_docids_chunk, (fid_facet_is_empty_docids_chunk, fid_facet_exists_docids_chunk), @@ -413,5 +413,5 @@ fn send_and_extract_flattened_documents_data( }, ); - Ok((docid_word_positions_chunk?, docid_fid_facet_values_chunks?)) + Ok((docid_word_positions_chunk?, fid_docid_facet_values_chunks?)) }