add sam_tag_dictionary

Author: smehringer

include/bio/format/sam_input_handler.hpp

@@ -32,6 +32,7 @@
 #include <bio/map_io/header.hpp>
 #include <bio/map_io/misc.hpp>
 #include <bio/map_io/sam_flag.hpp>
+#include <bio/map_io/sam_tag_dictionary.hpp>
 
 namespace bio
 {
@@ -109,6 +110,13 @@ class format_input_handler<sam> : public format_input_handler_base<format_input_
         get<field::tlen>(raw_record)  = (*file_it).fields[8];
         get<field::seq>(raw_record)   = (*file_it).fields[9];
         get<field::qual>(raw_record)  = (*file_it).fields[10];
+
+        // fields[10].end() that is guaranteed to be char*
+        char const * end_qual             = (*file_it).fields[10].data() + (*file_it).fields[10].size() + 1/*\t or \n*/;
+        // line.end() that is guaranteed to be char*
+        char const * end_line             = (*file_it).line.data() + (*file_it).line.size();
+        // SAM tags go from end of qual til end of line (possibly empty)
+        get<field::tags>(raw_record) = std::string_view{end_qual, static_cast<size_t>(end_line - end_qual)};
     }
 
     /* PARSED RECORD HANDLING */
@@ -160,7 +168,7 @@ class format_input_handler<sam> : public format_input_handler_base<format_input_
         }
     }
 
-    /* POS, MAPQ are handled correctly by default, unless we want pos to be read into an std:optional */
+    /* POS, MAPQ are handled correctly by default */
 
     //!\brief Overload for parsing CIGAR.
     void parse_field(vtag_t<field::cigar> const & /**/, std::vector<seqan3::cigar> & cigar_vector)
@@ -242,6 +250,16 @@ class format_input_handler<sam> : public format_input_handler_base<format_input_
             parse_field_aux(raw_field, parsed_field); // reading into e.g. dna4 vector
     }
 
+    //!\brief Overload for parsing the SAM tag dictionary.
+    void parse_field(vtag_t<field::tags> const & /**/, map_io::sam_tag_dictionary & dictionary)
+    {
+        std::string_view raw_field = get<field::tags>(raw_record);
+
+        if (!raw_field.empty())
+            for (std::string_view const tag_field : raw_field | detail::eager_split('\t'))
+                dictionary.parse_and_emplace(tag_field);
+    }
+
     //!\brief Overload for parsing the private data.
     void parse_field(vtag_t<field::_private> const & /**/, map_io::record_private_data & parsed_field)
     {

include/bio/map_io/misc.hpp

@@ -33,6 +33,7 @@ inline constexpr auto default_field_ids = vtag<field::qname,
                                                field::tlen,
                                                field::seq,
                                                field::qual,
+                                               field::tags,
                                                field::_private>;
 
 } // namespace bio::map_io

include/bio/map_io/reader_options.hpp

@@ -28,6 +28,7 @@
 #include <bio/map_io/header.hpp>
 #include <bio/map_io/misc.hpp>
 #include <bio/map_io/sam_flag.hpp>
+#include <bio/map_io/sam_tag_dictionary.hpp>
 
 namespace bio::map_io
 {
@@ -68,6 +69,7 @@ inline constexpr auto field_types_sam =
        int32_t,                                                                 // field::tlen,
        decltype(std::string_view{} | seqan3::views::char_to<seqan3::dna5>),     // field::seq
        decltype(std::string_view{} | seqan3::views::char_to<seqan3::phred42>),  // field::qual
+       sam_tag_dictionary,                                                      // field::tags
        record_private_data>;                                                    // field::_private
 //!\}