Merge OrientDB：将什么合并到什么？

我对OrientDB的API文档有点困惑。 气味记录法说：

将当前文档与作为参数传递的文档合并。如果该字段已存在，则将根据参数“iUpdateOnlyMode”的值管理冲突

参数： iOther-要合并的其他OdoDocument实例 iUpdateOnlyMode-如果为true，则将始终添加或覆盖其他文档属性。如果为false，则原始文档将删除其他文档中缺少的属性 iMergeSingleItemsOfMultiValueFields-如果为true，则合并多字段集合、映射、数组等的单个项

从这一点上，我不知道什么在什么时候融入了什么。这就是改进的状态，因为有两种方法的描述更不清楚

---

有人对什么合并成什么有正常的解释吗？也许Ron Kittle可能知道？

我做了一些测试，似乎API的名称有点奇怪或记录错误

我希望doc1.mergedoc2将doc2合并到doc1中。因为您通常正在编辑doc1，并且可以连续合并多个内容，如下所示：

doc1.merge(partialDocA);
doc1.merge(partialDocB);
doc1.merge(partialDocC);

partialDocA.merge(doc1, true, false);
partialDocB.merge(doc1, true, false);
partialDocC.merge(doc1, true, false);

public ODocument mergeDocument(Map<String, Object> incomingDocMap)
{
    String sourceUri = (String) incomingDocMap.get(DocumentAttributes.SOURCE_URI.toString());
    if (null == sourceUri)
        throw new IllegalArgumentException("Document sourceUri is null.");
    String docType = (String) incomingDocMap.get(Crawler.Attributes.TYPE);
    if (null == docType)
        throw new IllegalArgumentException("Document docType is null.");

    // Get a document by sourceUri
    String sql = "SELECT * FROM " + docType + " WHERE sourceuri=?";
    activateOnCurrentThread();
    List<ODocument> results = db.command(new OSQLSynchQuery<ODocument>(sql)).execute(sourceUri);
    if (results.size() == 0)
        throw new RuntimeException("No document with sourceUri '"+sourceUri+"'.");

    // Update it from the given map.
    ODocument incomingDoc = new ODocument(docType);
    incomingDoc.fromMap(incomingDocMap);
    //ODocument merged = incomingDoc.merge(results.get(0), true, false);
    ODocument merged = results.get(0).merge(incomingDoc, true, false);
    //merged.save(); /// Duplicated key
    return merged;
}

从merge的文档中可以看出，OrientDB的功能如下：

doc1.merge(partialDocA);
doc1.merge(partialDocB);
doc1.merge(partialDocC);

partialDocA.merge(doc1, true, false);
partialDocB.merge(doc1, true, false);
partialDocC.merge(doc1, true, false);

public ODocument mergeDocument(Map<String, Object> incomingDocMap)
{
    String sourceUri = (String) incomingDocMap.get(DocumentAttributes.SOURCE_URI.toString());
    if (null == sourceUri)
        throw new IllegalArgumentException("Document sourceUri is null.");
    String docType = (String) incomingDocMap.get(Crawler.Attributes.TYPE);
    if (null == docType)
        throw new IllegalArgumentException("Document docType is null.");

    // Get a document by sourceUri
    String sql = "SELECT * FROM " + docType + " WHERE sourceuri=?";
    activateOnCurrentThread();
    List<ODocument> results = db.command(new OSQLSynchQuery<ODocument>(sql)).execute(sourceUri);
    if (results.size() == 0)
        throw new RuntimeException("No document with sourceUri '"+sourceUri+"'.");

    // Update it from the given map.
    ODocument incomingDoc = new ODocument(docType);
    incomingDoc.fromMap(incomingDocMap);
    //ODocument merged = incomingDoc.merge(results.get(0), true, false);
    ODocument merged = results.get(0).merge(incomingDoc, true, false);
    //merged.save(); /// Duplicated key
    return merged;
}

其中true使merge不能删除doc1的其余部分。如果是false，它最终将成为partialdoc。那是什么样的合并？这叫做替换，不是合并

但在我进行了一些测试之后，第一个案例似乎是正确的

如果这是真的，它应该被命名为mergeFromtargetDoc，而不是merge，一些带有多字段的东西

因此，除非我忽略了什么，否则从地图保存文档如下所示：

doc1.merge(partialDocA);
doc1.merge(partialDocB);
doc1.merge(partialDocC);

partialDocA.merge(doc1, true, false);
partialDocB.merge(doc1, true, false);
partialDocC.merge(doc1, true, false);

public ODocument mergeDocument(Map<String, Object> incomingDocMap)
{
    String sourceUri = (String) incomingDocMap.get(DocumentAttributes.SOURCE_URI.toString());
    if (null == sourceUri)
        throw new IllegalArgumentException("Document sourceUri is null.");
    String docType = (String) incomingDocMap.get(Crawler.Attributes.TYPE);
    if (null == docType)
        throw new IllegalArgumentException("Document docType is null.");

    // Get a document by sourceUri
    String sql = "SELECT * FROM " + docType + " WHERE sourceuri=?";
    activateOnCurrentThread();
    List<ODocument> results = db.command(new OSQLSynchQuery<ODocument>(sql)).execute(sourceUri);
    if (results.size() == 0)
        throw new RuntimeException("No document with sourceUri '"+sourceUri+"'.");

    // Update it from the given map.
    ODocument incomingDoc = new ODocument(docType);
    incomingDoc.fromMap(incomingDocMap);
    //ODocument merged = incomingDoc.merge(results.get(0), true, false);
    ODocument merged = results.get(0).merge(incomingDoc, true, false);
    //merged.save(); /// Duplicated key
    return merged;
}

该方法甚至没有填充@return，因此我不知道是否可以继续使用原始的OdoDocument

---

很好，OrientDB。

这在OrientDB 2.2.0中有效吗？