[FIX] genbank/embl parsing

seqan · Nov 12, 2023 · e746fcb · e746fcb
1 parent 285f8e8
commit e746fcb
Show file tree

Hide file tree

Showing 3 changed files with 9 additions and 11 deletions.
diff --git a/include/seqan3/io/sequence_file/format_embl.hpp b/include/seqan3/io/sequence_file/format_embl.hpp
@@ -197,12 +197,10 @@ class format_embl
         }
         else
         {
-            detail::consume(stream_view | detail::take_until(is_end));
+            detail::consume(stream_view | detail::take_until_or_throw(is_end)); // consume until "//"
         }
-        //Jump over // and cntrl
-        ++stream_it;
-        ++stream_it;
-        ++stream_it;
+
+        std::ranges::advance(stream_it, 3u, std::ranges::end(stream_view)); // Skip `//` and potentially '\n'
     }
 
     //!\copydoc sequence_file_output_format::write_sequence_record

diff --git a/include/seqan3/io/sequence_file/format_genbank.hpp b/include/seqan3/io/sequence_file/format_genbank.hpp
@@ -164,8 +164,8 @@ class format_genbank
         {
             constexpr auto is_legal_alph = char_is_valid_for<seq_legal_alph_type>;
             std::ranges::copy(
-                stream_view | std::views::filter(!(is_space || is_digit))
-                    | detail::take_until_or_throw_and_consume(is_end) // consume "//"
+                stream_view | std::views::filter(!(is_space || is_digit)) // ignore whitespace and numbers
+                    | detail::take_until_or_throw(is_end)                 // until //
                     | std::views::transform(
                         [is_legal_alph](char const c) // enforce legal alphabet
                         {
@@ -183,9 +183,10 @@ class format_genbank
         }
         else
         {
-            detail::consume(stream_view | detail::take_until_or_throw_and_consume(is_end)); // consume until "//"
-            ++stream_it;                                                                    // consume "/n"
+            detail::consume(stream_view | detail::take_until_or_throw(is_end)); // consume until "//"
         }
+
+        std::ranges::advance(stream_it, 3u, std::ranges::end(stream_view)); // Skip `//` and potentially '\n'
     }
 
     //!\copydoc sequence_file_output_format::write_sequence_record

diff --git a/test/unit/io/sequence_file/sequence_file_format_genbank_test.cpp b/test/unit/io/sequence_file/sequence_file_format_genbank_test.cpp
@@ -50,8 +50,7 @@ DEFINITION  Homo sapiens mRNA for prepro cortistatin like peptide, complete
 ACCESSION   ID3
 ORIGIN
         1 ACGTTTA
-//
-)"};
+//)"};
 
     std::string illegal_alphabet_character_input{
         R"(LOCUS ID1
-Original file line number
+Diff line change
@@ Expand Up @@
     ACCESSION   ID3
     ORIGIN
 ACGTTTA
-    //
-    )"};
+    //)"};
         std::string illegal_alphabet_character_input{
             R"(LOCUS ID1
@@ Expand Down @@