GenBank entries in the CON division contain CONTIG fields, rather than sequences. Where the ORIGIN line would normally be, followed by sequence we lines such as:

/translation="SGAFKSLISSAFVSWKTTGKLQQTVRDSVERTGRGLHSGEISTV
KILPAAARFGRRFLFRSTVIAASIDNVVKETPLCTTLSKDGCTIRTVEHLLSALEASG
VDNCLIEIAGSGDCQRSIEV"
CONTIG join(AUSU01008588.1:1..7393)
//

Entries such as this are saved by NCBI Entrez if the 'Genbank' option is chosen, and the complete sequence is saved if 'Genbank (full) is chosen.

Currently, BioLegato will stop reading a GenBank entry upon the first instance of a CONTIG entry, and fails to read the remaining entry. BioLegato should be able to at least read such files as empty sequences. (Even better would be if the Open function could go to NCBI and get the full entry, but that is too complicated.) At the very least, BioLegato
should always terminate reading the current sequence
upon encountering the // line, and begin a new sequence at that point.

Examples of GenBank entries with CONTIGs instead of sequences include accession numbers KE535216, NW_003763341, and GL982899.

So the fix should be to read a 0 length sequence when encountering CONTIGs.