Use an u16 field id instead of one byte

milli/src/update/delete_documents.rs

@@ -430,8 +430,10 @@ where
     C: heed::BytesDecode<'a, DItem = K> + heed::BytesEncode<'a, EItem = K>,
     F: Fn(K) -> DocumentId,
 {
-    let mut iter =
-        db.remap_key_type::<ByteSlice>().prefix_iter_mut(wtxn, &[field_id])?.remap_key_type::<C>();
+    let mut iter = db
+        .remap_key_type::<ByteSlice>()
+        .prefix_iter_mut(wtxn, &field_id.to_be_bytes())?
+        .remap_key_type::<C>();
 
     while let Some(result) = iter.next() {
         let (key, ()) = result?;

milli/src/update/facets.rs

@@ -15,7 +15,7 @@ use crate::heed_codec::CboRoaringBitmapCodec;
 use crate::update::index_documents::{
     create_writer, write_into_lmdb_database, writer_into_reader, WriteMethod,
 };
-use crate::{Index, Result};
+use crate::{FieldId, Index, Result};
 
 pub struct Facets<'t, 'u, 'i> {
     wtxn: &'t mut heed::RwTxn<'i, 'u>,
@@ -119,7 +119,7 @@ impl<'t, 'u, 'i> Facets<'t, 'u, 'i> {
 fn clear_field_number_levels<'t>(
     wtxn: &'t mut heed::RwTxn,
     db: heed::Database<FacetLevelValueF64Codec, CboRoaringBitmapCodec>,
-    field_id: u8,
+    field_id: FieldId,
 ) -> heed::Result<()> {
     let left = (field_id, 1, f64::MIN, f64::MIN);
     let right = (field_id, u8::MAX, f64::MAX, f64::MAX);
@@ -135,11 +135,11 @@ fn compute_facet_number_levels<'t>(
     shrink_size: Option<u64>,
     level_group_size: NonZeroUsize,
     min_level_size: NonZeroUsize,
-    field_id: u8,
+    field_id: FieldId,
 ) -> Result<Reader<FileFuse>> {
     let first_level_size = db
         .remap_key_type::<ByteSlice>()
-        .prefix_iter(rtxn, &[field_id])?
+        .prefix_iter(rtxn, &field_id.to_be_bytes())?
         .remap_types::<DecodeIgnore, DecodeIgnore>()
         .fold(Ok(0usize), |count, result| result.and(count).map(|c| c + 1))?;
 
@@ -196,11 +196,11 @@ fn compute_facet_number_levels<'t>(
 fn compute_faceted_documents_ids(
     rtxn: &heed::RoTxn,
     db: heed::Database<ByteSlice, CboRoaringBitmapCodec>,
-    field_id: u8,
+    field_id: FieldId,
 ) -> Result<RoaringBitmap> {
     let mut documents_ids = RoaringBitmap::new();
 
-    for result in db.prefix_iter(rtxn, &[field_id])? {
+    for result in db.prefix_iter(rtxn, &field_id.to_be_bytes())? {
         let (_key, docids) = result?;
         documents_ids |= docids;
     }
@@ -210,7 +210,7 @@ fn compute_faceted_documents_ids(
 
 fn write_number_entry(
     writer: &mut Writer<File>,
-    field_id: u8,
+    field_id: FieldId,
     level: u8,
     left: f64,
     right: f64,

milli/src/update/index_documents/merge_function.rs

@@ -40,7 +40,7 @@ pub fn keep_first(_key: &[u8], values: &[Cow<[u8]>]) -> Result<Vec<u8>> {
     Ok(values.first().unwrap().to_vec())
 }
 
-pub fn merge_two_obkvs(base: obkv::KvReader, update: obkv::KvReader, buffer: &mut Vec<u8>) {
+pub fn merge_two_obkvs(base: obkv::KvReaderU16, update: obkv::KvReaderU16, buffer: &mut Vec<u8>) {
     use itertools::merge_join_by;
     use itertools::EitherOrBoth::{Both, Left, Right};
 

milli/src/update/index_documents/store.rs

@@ -7,6 +7,7 @@ use std::time::Instant;
 use std::{cmp, iter};
 
 use bstr::ByteSlice as _;
+use concat_arrays::concat_arrays;
 use fst::Set;
 use grenad::{CompressionType, FileFuse, Reader, Sorter, Writer};
 use heed::BytesEncode;
@@ -776,7 +777,8 @@ impl<'s, A: AsRef<[u8]>> Store<'s, A> {
         for ((fid, count), docids) in self.field_id_word_count_docids {
             docids_buffer.clear();
             CboRoaringBitmapCodec::serialize_into(&docids, &mut docids_buffer);
-            self.field_id_word_count_docids_sorter.insert([fid, count], &docids_buffer)?;
+            let key: [u8; 3] = concat_arrays!(fid.to_be_bytes(), [count]);
+            self.field_id_word_count_docids_sorter.insert(key, &docids_buffer)?;
         }
 
         let fst = builder.into_set();

milli/src/update/index_documents/transform.rs

@@ -626,7 +626,7 @@ mod test {
                 Some("tata".to_string()),
                 false,
             );
-            assert_eq!(result.unwrap(), (0u8, "toto".to_string()));
+            assert_eq!(result.unwrap(), (0, "toto".to_string()));
             assert_eq!(fields_map.len(), 1);
         }
 
@@ -635,7 +635,7 @@ mod test {
             let mut fields_map = FieldsIdsMap::new();
             let result =
                 compute_primary_key_pair(None, &mut fields_map, Some("tata".to_string()), false);
-            assert_eq!(result.unwrap(), (0u8, "tata".to_string()));
+            assert_eq!(result.unwrap(), (0, "tata".to_string()));
             assert_eq!(fields_map.len(), 1);
         }
 
@@ -643,7 +643,7 @@ mod test {
         fn should_return_default_if_both_are_none() {
             let mut fields_map = FieldsIdsMap::new();
             let result = compute_primary_key_pair(None, &mut fields_map, None, true);
-            assert_eq!(result.unwrap(), (0u8, "id".to_string()));
+            assert_eq!(result.unwrap(), (0, "id".to_string()));
             assert_eq!(fields_map.len(), 1);
         }