Skip to content

Open Source alternative to Algolia + Pinecone and an Easier-to-Use alternative to ElasticSearch ⚡ 🔍 ✨ Fast, typo tolerant, in-memory fuzzy Search Engine for building delightful search experiences

Show query snippet as a separate field instead of modifying the field value directly.

Kishore Nallan688bd218161f
diff --git a/src/collection.cpp b/src/collection.cpp
index 8b34e67d..f75d6412 100644
--- a/src/collection.cpp
+++ b/src/collection.cpp
@@ -656,25 +656,42 @@ nlohmann::json Collection::search(std::string query, const std::vector<std::stri
const std::string & field_name = search_fields[search_fields.size() - field_order_kv.first];
field search_field = search_schema.at(field_name);
+ // only string fields are supported for now
if(search_field.type == field_types::STRING) {
std::vector<std::string> tokens;
StringUtils::split(document[field_name], tokens, " ");
- for(size_t i = 1; i <= field_order_kv.second.offset_diffs[0]; i++) {
+ std::vector<size_t> token_indices;
+ char num_tokens_found = field_order_kv.second.offset_diffs[0];
+ for(size_t i = 1; i <= num_tokens_found; i++) {
size_t token_index = (size_t)(field_order_kv.second.start_offset + field_order_kv.second.offset_diffs[i]);
- tokens[token_index] = "<mark>" + tokens[token_index] + "</mark>";
+ token_indices.push_back(token_index);
}
- std::stringstream ss;
+ auto minmax = std::minmax_element(token_indices.begin(), token_indices.end());
- for(size_t token_index = 0; token_index < tokens.size(); ++token_index) {
- if(token_index != 0) {
- ss << " ";
+ // pick surrounding tokens within N tokens of min_index and max_index for the snippet
+ const size_t start_index = std::max(0, (int)(*(minmax.first)-5));
+ const size_t end_index = std::min((int)tokens.size(), (int)(*(minmax.second)+5));
+
+ std::stringstream snippet_stream;
+ size_t token_index = 0;
+
+ for(size_t snippet_index = start_index; snippet_index < end_index; snippet_index++) {
+ if(snippet_index != 0) {
+ snippet_stream << " ";
+ }
+
+ if(snippet_index == token_indices[token_index]) {
+ token_index++;
+ snippet_stream << "<mark>" + tokens[snippet_index] + "</mark>";
+ } else {
+ snippet_stream << tokens[snippet_index];
}
- ss << tokens[token_index];
}
- document[field_name] = ss.str();
+ document["_snippets"] = nlohmann::json::object();
+ document["_snippets"][field_name] = snippet_stream.str();
}
result["hits"].push_back(document);