mirror of
https://github.com/meilisearch/MeiliSearch
synced 2024-11-22 21:04:27 +01:00
Merge pull request #484 from meilisearch/fix-reindex-by-chunk
Stop reindexing by chunk during complete reindexing
This commit is contained in:
commit
72450c765d
@ -279,19 +279,17 @@ pub fn reindex_all_documents(writer: &mut heed::RwTxn<MainT>, index: &store::Ind
|
||||
index.postings_lists.clear(writer)?;
|
||||
index.docs_words.clear(writer)?;
|
||||
|
||||
// 3. re-index chunks of documents (otherwise we make the borrow checker unhappy)
|
||||
for documents_ids in documents_ids_to_reindex.chunks(100) {
|
||||
let stop_words = match index.main.stop_words_fst(writer)? {
|
||||
Some(stop_words) => stop_words,
|
||||
None => fst::Set::default(),
|
||||
};
|
||||
|
||||
let number_of_inserted_documents = documents_ids.len();
|
||||
let number_of_inserted_documents = documents_ids_to_reindex.len();
|
||||
let mut indexer = RawIndexer::new(stop_words);
|
||||
let mut ram_store = HashMap::new();
|
||||
|
||||
for document_id in documents_ids {
|
||||
for result in index.documents_fields.document_fields(writer, *document_id)? {
|
||||
for document_id in documents_ids_to_reindex {
|
||||
for result in index.documents_fields.document_fields(writer, document_id)? {
|
||||
let (field_id, bytes) = result?;
|
||||
let value: serde_json::Value = serde_json::from_slice(bytes)?;
|
||||
ram_store.insert((document_id, field_id), value);
|
||||
@ -302,7 +300,7 @@ pub fn reindex_all_documents(writer: &mut heed::RwTxn<MainT>, index: &store::Ind
|
||||
writer,
|
||||
field_id,
|
||||
&schema,
|
||||
*docid,
|
||||
docid,
|
||||
index.documents_fields,
|
||||
index.documents_fields_counts,
|
||||
&mut indexer,
|
||||
@ -320,7 +318,6 @@ pub fn reindex_all_documents(writer: &mut heed::RwTxn<MainT>, index: &store::Ind
|
||||
number_of_inserted_documents,
|
||||
indexer,
|
||||
)?;
|
||||
}
|
||||
|
||||
index.main.put_schema(writer, &schema)?;
|
||||
|
||||
|
Loading…
Reference in New Issue
Block a user