AAB20418 (partial ldh-a human) inherits cross-reference to full-length uniprot sequence

A peculiarity in ENA:
The AAB20418 record is a coding sequence for a peptide fragment:
<?xml version="1.0" encoding="UTF-8"?>
<ROOT request="AAB20418&amp;display=xml">
<entry accession="AAB20418" version="1" entryVersion="4" dataClass="STD" taxonomicDivision="HUM" moleculeType="genomic DNA" sequenceLength="39" topology="linear" firstPublic="1993-07-22" firstPublicRelease="36" lastUpdated="2006-11-14" lastUpdatedRelease="89">
<description>Homo sapiens (human) lactate dehydrogenase A</description>
<reference type="article" number="1">
<title>Genotypic analysis of families with lactate dehydrogenase A (M) deficiency by selective DNA amplification</title>
<author>Maekawa M.</author>
<author>Sudo K.</author>
<author>Li S.S.</author>
<author>Kanno T.</author>
<journal>Hum. Genet.</journal>
<year>1991</year>
<volume>88</volume>
<issue>1</issue>
<firstPage>34</firstPage>
<lastPage>38</lastPage>
<comment>GenBank staff at the National Library of Medicine created this entry [NCBI gibbsq 66853] from the original journal article. This sequence comes from Fig.3.</comment>
<referenceLocation>
Hum. Genet. 88(1):34-38(1991).
</referenceLocation>
<xref db="DOI" id="10.1007/BF00204925"/>
<xref db="PUBMED" id="1959923"/>
</reference>
<xref db="EMBL" id="S66853.1"/>
<xref db="MD5" id="96db7a51fbe12ed81993bc47e3b2fad8"/>
<feature name="source" location="1..39">
<taxon scientificName="Homo sapiens" commonName="human" taxId="9606">
<lineage>
<taxon scientificName="Eukaryota"/>
<taxon scientificName="Metazoa"/>
<taxon scientificName="Chordata"/>
<taxon scientificName="Craniata"/>
<taxon scientificName="Vertebrata"/>
<taxon scientificName="Euteleostomi"/>
<taxon scientificName="Mammalia"/>
<taxon scientificName="Eutheria"/>
<taxon scientificName="Euarchontoglires"/>
<taxon scientificName="Primates"/>
<taxon scientificName="Haplorrhini"/>
<taxon scientificName="Catarrhini"/>
<taxon scientificName="Hominidae"/>
<taxon scientificName="Homo"/>
</lineage>
</taxon>
<qualifier name="organism">
<value>
Homo sapiens
</value>
</qualifier>
</feature>
<feature name="CDS" location="S66853.1:1..39">
<xref db="GOA" id="P00338"/>
<xref db="HGNC" id="HGNC:6535"/>
<xref db="InterPro" id="IPR001236"/>
<xref db="InterPro" id="IPR001557"/>
<xref db="InterPro" id="IPR011304"/>
<xref db="InterPro" id="IPR015955"/>
<xref db="InterPro" id="IPR016040"/>
<xref db="InterPro" id="IPR018177"/>
<xref db="InterPro" id="IPR022383"/>
<xref db="PDB" id="1I10"/>
<xref db="PDB" id="4AJP"/>
<xref db="PDB" id="4JNK"/>
<xref db="PDB" id="4L4R"/>
<xref db="PDB" id="4L4S"/>
<xref db="PDB" id="4M49"/>
<xref db="PDB" id="4OJN"/>
<xref db="PDB" id="4OKN"/>
<xref db="PDB" id="4QO7"/>
<xref db="PDB" id="4QO8"/>
<xref db="PDB" id="4QSM"/>
<xref db="PDB" id="4QT0"/>
<xref db="PDB" id="4R68"/>
<xref db="PDB" id="4R69"/>
<xref db="PDB" id="4RLS"/>
<xref db="PDB" id="4ZVV"/>
<xref db="PDB" id="5IXS"/>
<xref db="PDB" id="5IXY"/>
<xref db="UniProtKB/Swiss-Prot" id="P00338"/>
<qualifier name="codon_start">
<value>
1
</value>
</qualifier>
<qualifier name="gene">
<value>
lactate dehydrogenase A, LDH-A
</value>
</qualifier>
<qualifier name="product">
<value>
lactate dehydrogenase A
</value>
</qualifier>
<qualifier name="note">
<value>
This sequence comes from Fig.3; LDH-A
</value>
</qualifier>
<qualifier name="protein_id">
<value>
AAB20418.1
</value>
</qualifier>
<qualifier name="translation">
<value>
SWAIGREYNEES
</value>
</qualifier>
<qualifier name="partial">
</qualifier>
</feature>
<sequence>
tcctgggctattggcagagagtataatgaagaatcttag
</sequence>
</entry>
</ROOT>

This fragment is a variant observed somewhere on P00338 (Jalview (2.10.2 and earlier) will add the Uniprot ID P00338 to the product, since there's no additional information to indicate that this product is not, in fact, in direct 1:1 correspondence with the canonical sequence of P00338.

Currently, this case causes CrossRef2xmlTests.testRetrieveAndShowCrossref to fail for UNIPROT:P00338, because this fragment's peptide is not added to the dataset (presumably because there's already a sequence with this mapping)