Merge pull request #241 from tcezard/EVA3281_synonyms_on_one_line_are…

…_allowed

EVA-3281 - Allow synonyms to be the same value if they are on the same line

inc/assembly_report/assembly_report.hpp

@@ -61,7 +61,11 @@ namespace ebi
 
         void add_synonym(std::string synonym)
         {
-            synonyms.push_back(synonym);
+            // Only add the synonym if there isn't one already
+            if (std::find(synonyms.begin(), synonyms.end(), synonym) == synonyms.end())
+            {
+                synonyms.push_back(synonym);
+            }
         }
 
       private:

test/assembly_report/assembly_report_test.cpp

@@ -42,6 +42,14 @@ namespace ebi
             CHECK_NOTHROW(synonyms_map.parse_assembly_report(input));
         }
 
+        SECTION("Parsing assembly report with same value in one line")
+        {
+            ebi::assembly_report::SynonymsMap synonyms_map;
+            std::string assembly_report_path = "test/input_files/assembly_report/full_assembly_report/passed_with_same_synonym_on_same_line.txt";
+            std::ifstream input{assembly_report_path};
+            CHECK_NOTHROW(synonyms_map.parse_assembly_report(input));
+        }
+
         SECTION("Failing to parse assembly report due to incorrect column size")
         {
             ebi::assembly_report::SynonymsMap synonyms_map;

test/input_files/assembly_report/GCF_000007105.1_ASM710v1_assembly_report.txt

@@ -0,0 +1,27 @@
+# Assembly name:  ASM710v1
+# Organism name:  Zymomonas mobilis subsp. mobilis ZM4 = ATCC 31821 (a-proteobacteria)
+# Infraspecific name:  strain=ZM4
+# Taxid:          264203
+# BioSample:      SAMN02603579
+# BioProject:     PRJNA12354
+# Submitter:      Macrogen Inc.
+# Date:           2010-01-12
+# Assembly type:  na
+# Release type:   major
+# Assembly level: Complete Genome
+# Genome representation: full
+# GenBank assembly accession: GCA_000007105.1
+# RefSeq assembly accession: GCF_000007105.1
+# RefSeq assembly and GenBank assemblies identical: yes
+#
+## Assembly-Units:
+## GenBank Unit Accession	RefSeq Unit Accession	Assembly-Unit name
+## GCA_000007115.1	GCF_000007115.1	Primary Assembly
+#
+# Ordered by chromosome/plasmid; the chromosomes/plasmids are followed by
+# unlocalized scaffolds.
+# Unplaced scaffolds are listed at the end.
+# RefSeq is equal or derived from GenBank object.
+#
+# Sequence-Name	Sequence-Role	Assigned-Molecule	Assigned-Molecule-Location/Type	GenBank-Accn	Relationship	RefSeq-Accn	Assembly-Unit	Sequence-Length	UCSC-style-name
+AE008692.2	assembled-molecule	na	Chromosome	AE008692.2	=	NC_006526.2	Primary Assembly	2056363	na

test/input_files/assembly_report/passed_with_same_synonym_on_same_line.txt

@@ -0,0 +1,3 @@
+# Sequence-Name	Sequence-Role	Assigned-Molecule	Assigned-Molecule-Location/Type	GenBank-Accn	Relationship	RefSeq-Accn	Assembly-Unit	Sequence-Length	UCSC-style-name
+CM000072.5	assembled-molecule	1	Chromosome	CM000072.5	=	NC_005100.4	Primary Assembly	282763074	1
+CM000082.5	assembled-molecule	1	Chromosome	CM000082.5	=	NC_005300.4	Primary Assembly	282763044	2

test/input_files/v4.3/assembly_checker/passed/passed_with_mapping_multi_synonyms/assembly_report.txt

@@ -0,0 +1,3 @@
+# Sequence-Name	Sequence-Role	Assigned-Molecule	Assigned-Molecule-Location/Type	GenBank-Accn	Relationship	RefSeq-Accn	Assembly-Unit	Sequence-Length	UCSC-style-name
+CM000072.5	assembled-molecule	1	Chromosome	CM000072.5	=	NC_005100.4	Primary Assembly	282763074	1
+chr2	assembled-molecule	1	Chromosome	CM000082.5	=	NC_005300.4	Primary Assembly	282763044	2

test/input_files/v4.3/assembly_checker/passed/passed_with_mapping_multi_synonyms/passed_with_mapping_multi_synonyms.fai

@@ -0,0 +1 @@
+CM000072.5	59940	56	60	61

test/vcf/assembly_checker_integration_test.cpp

@@ -74,6 +74,16 @@ namespace ebi
             CHECK(ebi::vcf::assembly_checker::check_vcf_ref(vcf_input, vcf_path, fasta_path, assembly_report_path, outputs));
         }
 
+        SECTION("Mapping contigs but contig has same synonym on the same line, all match")
+        {
+            auto folder = boost::filesystem::path("test/input_files/v4.3/assembly_checker/passed/passed_with_mapping_multi_synonyms/");
+            std::string file_prefix = folder.parent_path().filename().string();
+            std::string vcf_path = folder.string() + file_prefix + ebi::vcf::VCF_EXT;
+            std::ifstream vcf_input{vcf_path};
+            std::string fasta_path{folder.string() + file_prefix + ebi::vcf::FASTA_EXT};
+            std::string assembly_report_path = folder.string() + "assembly_report.txt";
+            CHECK(ebi::vcf::assembly_checker::check_vcf_ref(vcf_input, vcf_path, fasta_path, assembly_report_path, outputs));
+        }
 
         SECTION("compressed VCF, gz compression")
         {