From 0ae40221e66382ac330748c3eefe916de5e26975 Mon Sep 17 00:00:00 2001
From: Fabian Engelniederhammer <fabian.engelniederhammer@tngtech.com>
Date: Tue, 16 Jul 2024 22:16:45 +0200
Subject: [PATCH] feat: set the default sequence when there is only a single
 sequence in the reference genomes

closes #454
---
 include/silo/preprocessing/preprocessor.h |  1 +
 src/silo/preprocessing/preprocessor.cpp   | 13 ++++++++++++-
 2 files changed, 13 insertions(+), 1 deletion(-)

diff --git a/include/silo/preprocessing/preprocessor.h b/include/silo/preprocessing/preprocessor.h
index 3bce03cde..446bb6666 100644
--- a/include/silo/preprocessing/preprocessor.h
+++ b/include/silo/preprocessing/preprocessor.h
@@ -43,6 +43,7 @@ class Preprocessor {
    Database preprocess();
 
   private:
+   void finalizeConfig();
    void validateConfig();
 
    static std::string makeNonNullKey(const std::string& field);
diff --git a/src/silo/preprocessing/preprocessor.cpp b/src/silo/preprocessing/preprocessor.cpp
index 74a963fb6..fca68c830 100644
--- a/src/silo/preprocessing/preprocessor.cpp
+++ b/src/silo/preprocessing/preprocessor.cpp
@@ -63,7 +63,7 @@ Preprocessor::Preprocessor(
 }
 
 Database Preprocessor::preprocess() {
-   validateConfig();
+   finalizeConfig();
 
    SPDLOG_INFO(
       "preprocessing - creating intermediate results directory '{}'",
@@ -128,6 +128,17 @@ Database Preprocessor::preprocess() {
    );
 }
 
+void Preprocessor::finalizeConfig() {
+   if (nuc_sequences.size() == 1 && !database_config.default_nucleotide_sequence.has_value()) {
+      database_config.default_nucleotide_sequence = nuc_sequences.at(0);
+   }
+   if (aa_sequences.size() == 1 && !database_config.default_amino_acid_sequence.has_value()) {
+      database_config.default_amino_acid_sequence = aa_sequences.at(0);
+   }
+
+   validateConfig();
+}
+
 void Preprocessor::validateConfig() {
    const bool default_nucleotide_sequence_is_not_in_reference =
       database_config.default_nucleotide_sequence.has_value() &&