java - Solr で深くネストされた構造を更新する

Question

私はまだsolrに慣れていません。以下のようにネストされた構造のインデックスを作成しようとしていますが、SolrJ 6.1 でのインデックス作成に問題があります。

schema.xml

<?xml version="1.0" encoding="UTF-8"?>
<schema name="example" version="1.6">
  <uniqueKey>id</uniqueKey>
  <defaultSearchField>title</defaultSearchField>
  ...
  // Here are described all the fieldType
  ...
  <field name="_root_" type="string" indexed="true" stored="false"/>
  <field name="_version_" type="long" indexed="true" stored="false"/>
  <field name="id" type="string" multiValued="false" indexed="true" required="true" stored="true"/>
  <field name="imdbId" type="string" indexed="true" stored="true"/>
  <field name="rating" type="float" indexed="true" stored="true"/>
  <field name="title" type="text_en" indexed="true" stored="true"/>
  <field name="type" type="string" indexed="true" stored="true"/>
  <field name="userId" type="string" indexed="true" stored="true"/>
</schema>

SolrJの試み

私はそれを3つのステップで行います。

SolrClient solr = new HttpSolrClient.Builder("http://localhost:8983/solr/ml_core").build();

SolrInputDocument doc, childDoc;
String[] line;
CSVReader reader;

// Step 1: Create a document - Very good 
reader = new CSVReader(new FileReader("movies.csv")); // structure of the file: movieId,title
while ((line = reader.readNext()) != null) {
    doc = new SolrInputDocument();
    doc.addField("id", line[0]);
    doc.addField("title", line[1]);
    doc.addField("type", "film");
    solr.add(doc); 
}

// Step 2: Updating a document that I created - Very good 
reader = new CSVReader(new FileReader("links.csv")); // structure of the file: movieId,imdbId 
while ((line = reader.readNext()) != null) {

    doc = new SolrInputDocument();
    doc.addField("id", line[0]);

    Map<String, Object> imdbIdModifier = new HashMap<>(1);
    imdbIdModifier.put("set", line[1]);
    doc.addField("imdbId", imdbIdModifier);  // add the map as the field value

    solr.add(doc); 
}

// Step 3: Updating deeply nested structures - Here is the error
reader = new CSVReader(new FileReader("ratings.csv")); // structure of the file: movieId,userId,rating
while ((line = reader.readNext()) != null) {
    doc = new SolrInputDocument();
    doc.addField("id", line[0]);

    childDoc = new SolrInputDocument();
    childDoc.addField("id", line[0] + "_" + line[1]);
    childDoc.addField("userId", line[1]);
    childDoc.addField("type", "user");
    childDoc.addField("rating", line[2]);
    doc.addChildDocument(childDoc);

    solr.add(doc); 
}

solr.commit();
solr.optimize();

私は以下を受け取りました：

私のクエリ: http://localhost:8983/solr/ml_core/select?indent=on&q=id:1&wt=json

    {
  "responseHeader":{
    "status":0,
    "QTime":1,
    "params":{
      "q":"id:1",
      "indent":"on",
      "wt":"json",
      "_":"1471440200579"}},
  "response":{"numFound":2,"start":0,"docs":[
      {
        "id":"1",
        "title":"Toy Story (1995)",
        "type":"film",
        "imdbId":"0114709",
        "_version_":1542910355358875648},
      {
        "id":"1",
        "_version_":1542910730357964800,
        "_root_":"1"}]
  }}

応答 - 不正解です。「id」フィールドが重複していますが、ファイル schema.xml では、このフィールドは一意としてマークされています。

私のクエリ: http://localhost:8983/solr/ml_core/select?fl= *,[child%20parentFilter=type:film]&indent=on&q={!parent%20which=%27type:film%27}&wt=json

{
  "error":{
    "msg":"Parent query yields document which is not matched by parents filter, docID=19957",
    "trace":"java.lang.IllegalStateException: Parent query yields document which is not matched by parents filter, docID=19957\r\n\tat org.apache.lucene.search.join.ToChildBlockJoinQuery$ToChildBlockJoinScorer.validateParentDoc(ToChildBlockJoinQuery.java:305)\r\n\tat org.apache.lucene.search.join.ToChildBlockJoinQuery$ToChildBlockJoinScorer.access$300(ToChildBlockJoinQuery.java:158)\r\n\tat 
...
org.eclipse.jetty.util.thread.QueuedThreadPool$3.run(QueuedThreadPool.java:572)\r\n\tat java.lang.Thread.run(Thread.java:745)\r\n",
    "code":500}
}

応答 - 不正解です。

私は期待しました：

私のクエリ: http://localhost:8983/solr/ml_core/select?indent=on&q=id:1&wt=json

次の正解が必要です。

  {
  "responseHeader":{
    "status":0,
    "QTime":0,
    "params":{
      "q":"id:1",
      "indent":"on",
      "wt":"json",
      "_":"1471440410850"}},
  "response":{"numFound":1,"start":0,"docs":[
      {
        "id":"1",
        "title":"Toy Story (1995)",
        "type":"film",
        "imdbId":"0114709",
        "_version_":1542910355358875648,
        "_root_":"1"}]
  }}

私のクエリ: http://localhost:8983/solr/ml_core/select?fl= *,[child%20parentFilter=type:film]&indent=on&q={!parent%20which=%27type:film%27}&wt=json

次の正解が必要です。

{
  "responseHeader":{
    "status":0,
    "QTime":7,
    "params":{
      "q":"{!parent which='type:film'}",
      "indent":"on",
      "fl":"*,[child parentFilter=type:film]",
      "wt":"json",
      "_":"1471440410850"}},
  "response":{"numFound":1,"start":0,"docs":[
      {
        "id":"1",
        "title":"Toy Story (1995)",
        "type":"film",
        "imdbId":"0114709",
        "_version_":1542910355358875648,
        "_root_":"1",
        "_childDocuments_":[
        {
          "id":"1_Violet",
          "userId":"Violet",
          "type":"user",
          "rating":5.0,
        {
          "id":"1_Mcka",
          "userId":"Mcka",
          "type":"user",
          "rating":4.0}]}]
  }}

目的のドキュメント構造を取得するにはどうすればよいですか? SolrJでこれを修正するにはどうすればよいですか。ありがとう。