The child doc transformer worked great. Thank you.
In my experiment, posting '<delete><id>parent-id</id></delete>' to the update
end point only deleted the parent doc. Do I insert a complex join query from id
to _version_ and delete all the docs of the matching _version_ ?
On 7/24/18 9:27 PM, TK Solr wrote:
Thank you. I'll try the child doc transformer.
On a related question, if I delete a parent document, will its children be
deleted also? Or do I have to have a parent_id field in each child so that the
child docs can be deleted?
On 7/22/18 10:05 AM, Mikhail Khludnev wrote:
Hello,
Check [child]
https://lucene.apache.org/solr/guide/7_4/transforming-result-documents.html#child-childdoctransformerfactory
or [subquery].
Although, it's worth to put reference to it somewhere in blockjoin
qparsers.
Documentation patches are welcome.
On Sun, Jul 22, 2018 at 10:25 AM TK Solr <tksol...@sonic.net> wrote:
https://lucene.apache.org/solr/guide/7_4/other-parsers.html#block-join-parent-query-parser
talks about {!parent which=<query that matches all parent docs>} <query on
child
docs>, which returns parent docs only, and
{!child of=<query that matches all parent docs>} <query on parent docs>,
which
returns child docs only.
Is there a way to retrieve the matched documents in the original, nested
form?
Using the sample document, is there way to get:
<doc>
<field name="id">1</field>
<field name="title">Solr has block join support</field>
<field name="content_type">parentDocument</field>
<doc>
<field name="id">2</field>
<field name="comments">SolrCloud supports it too!</field>
</doc>
</doc>
rather than just the parent or the child docs?
