fix candidate string for non-ascii

hendrikmuhs · Mar 15, 2024 · 48f2029 · 48f2029
1 parent bca0b3b
commit 48f2029
Show file tree

Hide file tree

Showing 2 changed files with 3 additions and 3 deletions.
diff --git a/keyvi/include/keyvi/dictionary/matching/fuzzy_multiword_completion_matching.h b/keyvi/include/keyvi/dictionary/matching/fuzzy_multiword_completion_matching.h
@@ -251,7 +251,7 @@ class FuzzyMultiwordCompletionMatching final {
 
       if (traverser_ptr_->IsFinalState()) {
         std::string match_str = multiword_boundary_ > 0
-                                    ? distance_metric_->GetCandidate().substr(prefix_length_ + multiword_boundary_)
+                                    ? distance_metric_->GetCandidate(prefix_length_ + multiword_boundary_)
                                     : distance_metric_->GetCandidate();
 
         TRACE("found final state at depth %d %s", prefix_length_ + traverser_ptr_->GetDepth(), match_str.c_str());

diff --git a/keyvi/include/keyvi/stringdistance/needleman_wunsch.h b/keyvi/include/keyvi/stringdistance/needleman_wunsch.h
@@ -198,9 +198,9 @@ class NeedlemanWunsch final {
 
   int32_t GetScore() const { return distance_matrix_.Get(latest_calculated_row_, distance_matrix_.Columns() - 1); }
 
-  std::string GetCandidate() {
+  std::string GetCandidate(size_t pos = 0) {
     std::vector<unsigned char> utf8result;
-    utf8::utf32to8(compare_sequence_.begin(), compare_sequence_.begin() + last_put_position_ + 1,
+    utf8::utf32to8(compare_sequence_.begin() + pos, compare_sequence_.begin() + last_put_position_ + 1,
                    back_inserter(utf8result));
 
     return std::string(utf8result.begin(), utf8result.end());