postgres/typesense
1
0
Fork 0
mirror of https://github.com/typesense/typesense.git synced 2025-05-20 21:52:23 +08:00
Code Issues Actions 3 Packages Projects Releases Wiki Activity

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

Browse Source
This commit is contained in:
Kishore Nallan 2017-06-13 14:53:28 -05:00
parent 3efaed15b8
commit 688bd21816
1 changed files with 25 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

33
src/collection.cpp
View File

@ -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);