Documentation

Bam record in its native encoding along with virtual address.

Smart constructor. Makes sure we got a at least a full record.

internal representation of a BAM record

Cigar line in BAM coding Bam encodes an operation and a length into a single integer, we keep those integers in an array.

extracts the aligned length from a cigar line This gives the length of an alignment as measured on the reference, which is different from the length on the query or the length of the alignment.

A nucleotide base in an alignment. Experience says we're dealing with Ns and gaps all the type, so purity be damned, they are included as if they were real bases.

To allow Nucleotidess to be unpacked and incorporated into containers, we choose to represent them the same way as the BAM file format: as a 4 bit wide field. Gaps are encoded as 0 where they make sense, N is 15. The contained Word8 is guaranteed to be 0..15.

A vector that packs two Nucleotides into one byte, just like Bam does.

A collection of extension fields. The key is actually only two Chars, but that proved impractical. (Hmm... we could introduce a Key type that is a 16 bit int, then give it an instance IsString... practical?)

Deletes all occurences of some extension field.

Blindly inserts an extension field. This can create duplicates (and there is no telling how other tools react to that).

Deletes all occurences of an extension field, then inserts it with a new value. This is safer than insertE, but also more expensive.

Adjusts a named extension by applying a function.

A simple progress indicator that prints sequence id and position.