Disambiguation fails for second names

[zepinglee]

I'm recently working on the disambiguation feature of citeproc-lua and I come up with this test case. In the following fixture, both cites have original output "Doe, Roe" and disambiguation is required. Since only the second names are different ("Jane Roe" and "Jack Roe"), they are expanded to long form according to the disambiguation rule. The output is expected to be "Doe, Jane Roe; Doe, Jack Roe". However citeproc-js f56f8c9 produces "Doe, Roe; Doe, Roe" in which the second names are not disambiguated.

      + expected - actual

      -Doe, Roe; Doe, Roe
      +Doe, Jane Roe; Doe, Jack Roe

>>===== MODE =====>>
citation
<<===== MODE =====<<


>>===== RESULT =====>>
Doe, Jane Roe; Doe, Jack Roe
<<===== RESULT =====<<


>>===== CSL =====>>
<style
      xmlns="http://purl.org/net/xbiblio/csl"
      class="note"
      version="1.0">
  <info>
    <id />
    <title />
    <updated>2009-08-10T04:49:00+09:00</updated>
  </info>
  <citation
         disambiguate-add-givenname="true"
         givenname-disambiguation-rule="by-cite">
    <layout delimiter="; ">
      <names variable="author">
        <name form="short" />
      </names>
    </layout>
  </citation>
</style>
<<===== CSL =====<<


>>===== CITATION-ITEMS =====>>
[
    [
        {
            "id": "ITEM-1"
        },
        {
            "id": "ITEM-2"
        }
    ]
]
<<===== CITATION-ITEMS =====<<


>>===== INPUT =====>>
[
    {
        "author": [
            {
                "family": "Doe",
                "given": "John"
            },
            {
                "family": "Roe",
                "given": "Jane"
            }
        ],
        "id": "ITEM-1",
        "type": "book"
    },
    {
        "author": [
            {
                "family": "Doe",
                "given": "John"
            },
            {
                "family": "Roe",
                "given": "Jack"
            }
        ],
        "id": "ITEM-2",
        "type": "book"
    }
]
<<===== INPUT =====<<


>>===== VERSION =====>>
1.0
<<===== VERSION =====<<

[fbennett]

Good catch on this, thank you for this report. I've confirmed the bug and I'm working on a fix. Kind of surprised and embarrassed that we didn't catch this straightforward failure earlier!