Question

我的hibernate casscade存在问题。我试图坚持一套文件。数据模型如下：

语料库 1 --- n 文档 1 --- n TextBlock n --- 1 演讲者 n --- 1 派对

我的方案如下：

SpeakerFacade sf = new SpeakerFacade();
TextBlockFacade tf = new TextBlockFacade();

Corpus corpus = new Corpus();

Document doc1 = new Document(corpus);
TextBlock tb1 = new TextBlock(new Speaker("David", "Müller", new Party("ASDF")), "TB1", doc1);
tf.createTextBlock(tb1);
TextBlock tb2 = new TextBlock(new Speaker("Benedikt", "Müller", new Party("JKLÖ")), "TB2", doc1);
tf.createTextBlock(tb2);
TextBlock tb3 = new TextBlock(sf.findPersonById(1), "TB3", doc1);
tf.createTextBlock(tb3);

所以在第一个块中我创建了一个新的TextBlock。通过级联，其余部分也应该创建。在第二个块中，我在同一文档中创建另一个文本块。在最后一个块中，我创建了另一个文本块，但使用相同的扬声器。但我不断得到以下例外：

Caused by: org.hibernate.TransientPropertyValueException: object references an unsaved transient instance - save the transient instance before flushing : de.uniba.speechanalyser.persist.model.Document.corpus -> de.uniba.speechanalyser.persist.model.Corpus

在这里你可以看到我的模型类（简称）：

语料库

@Entity
public class Corpus implements Serializable {

private static final long serialVersionUID = 1L;

@Id
@GeneratedValue
@Column(name = "corpus_id")
private int id;

@OneToMany(cascade = { CascadeType.ALL }, mappedBy = "corpus")
private List<Document> documentList;
}

文档类

@Entity
public class Document implements Serializable {

private static final long serialVersionUID = 1L;

@Id
@GeneratedValue
@Column(name = "document_id")
private int id;

@ManyToOne(cascade = { CascadeType.MERGE },fetch= FetchType.EAGER)
private Corpus corpus;

@OneToMany(cascade = { CascadeType.ALL }, mappedBy = "document")
private List<TextBlock> textBlockList;
}

TextBlock类

@Entity
public class TextBlock implements Serializable {

private static final long serialVersionUID = 1L;

@Id
@GeneratedValue
@Column(name = "textblock_id")
private int id;

@Lob
String content;

@ManyToOne(cascade = { CascadeType.MERGE },fetch= FetchType.EAGER)
private Document document;

@ManyToOne(cascade = { CascadeType.MERGE },fetch= FetchType.EAGER)
private Speaker speaker;
}

演讲者班级

@Entity
@NamedQueries({
    @NamedQuery(name = "Speaker.findSpeakerByName", query = "select s from Speaker s where s.firstName = :firstName and s.lastName = :lastName") })
public class Speaker implements Serializable {

private static final long serialVersionUID = 1L;
public static final String FIND_BY_NAME = "Speaker.findSpeakerByName";

@Id
@GeneratedValue
@Column(name = "speaker_id")
private int id;

private String firstName;
private String lastName;

@OneToMany(cascade = { CascadeType.ALL }, mappedBy = "speaker")
private List<TextBlock> textBlock;

@ManyToOne(cascade = { CascadeType.MERGE },fetch= FetchType.EAGER)
private Party party;
}

派对类

@Entity
@NamedQueries({
@NamedQuery(name = "Party.findPartyByName", query = "select p from Party p where p.name = :name")
})

public class Party implements Serializable {

private static final long serialVersionUID = 1L;
public static final String FIND_BY_NAME = "Party.findPartyByName";

@Id
@GeneratedValue
@Column(name = "party_id")
private int id;

private String name;

@OneToMany(cascade = { CascadeType.ALL}, mappedBy = "party")
private List<Speaker> speakerList;
}

我也在使用对象/表之间的关系。尤其是小瀑布。我在stackoverflow上阅读了很多，但没有任何帮助。这是我目前的做法。当创建每个对象时，它们就像：

Speaker speaker= new Speaker("David", "Müller", pf.findById(1));
sf.createSpeaker(speaker);
speaker = sf.findSpeakerById(1);

然后将其添加到TextBlock它没有任何问题。那么有人可以帮助我吗？

问候，大卫

Answer 1

您的Corpus没有被保留。您只能级联MERGE。请参阅Document类中的注释定义：

@ManyToOne(cascade = { CascadeType.MERGE }, fetch= FetchType.EAGER)
private Corpus corpus;

应该是：

@ManyToOne(cascade = { CascadeType.MERGE, CascadeType.PERSIST }, fetch= FetchType.EAGER)
//you can simply say CascadeType.ALL if you are sure what you are doing
private Corpus corpus;

Answer 2

好的，很酷，我把问题缩小到只有关系的发言人 - 党派！

我的实际新课程是： 语料库

@Entity
public class Corpus implements Serializable {

private static final long serialVersionUID = 1L;

@Id
@GeneratedValue
@Column(name = "corpus_id")
private int id;

@OneToMany(cascade = { CascadeType.ALL }, mappedBy = "corpus")
private List<Document> documentList = new ArrayList<Document>();

public void addDocument(Document document) {
    if (documentList != null) {
        documentList.add(document);
    }
}
}

文档类

@Entity
public class Document implements Serializable {

private static final long serialVersionUID = 1L;

@Id
@GeneratedValue
@Column(name = "document_id")
private int id;

@ManyToOne
@JoinColumn(name = "corpus_id")
private Corpus corpus;

@OneToMany(cascade = { CascadeType.ALL }, mappedBy = "document")
private List<TextBlock> textBlockList = new ArrayList<TextBlock>();

public Document() {

}

public Document(Corpus corpus) {
    this.corpus = corpus;
}

public void addTextBlock(TextBlock textBlock) {
    if (textBlockList != null) {
        textBlockList.add(textBlock);
    }
}
}

TextBlock类

@Entity

public class TextBlock implements Serializable {

private static final long serialVersionUID = 1L;

@Id
@GeneratedValue
@Column(name = "textblock_id")
private int id;

@Lob
String content;

@ManyToOne
@JoinColumn(name = "document_id")
private Document document;

@ManyToOne
private Speaker speaker;

public TextBlock() {

}

public TextBlock(Speaker speaker, String content, Document document) {
    this.speaker = speaker;
    this.content = content;
    this.document = document;
}
}

演讲者班级

@Entity
@NamedQueries({
    @NamedQuery(name = "Speaker.findSpeakerByName", query = "select s from Speaker s where s.firstName = :firstName and s.lastName = :lastName") })

public class Speaker implements Serializable {

private static final long serialVersionUID = 1L;
public static final String FIND_BY_NAME = "Speaker.findSpeakerByName";

@Id
@GeneratedValue
@Column(name = "speaker_id")
private int id;

private String firstName;
private String lastName;

@OneToMany(cascade = { CascadeType.ALL }, mappedBy = "speaker")
private List<TextBlock> textBlock;

@ManyToOne
private Party party;

public Speaker() {

}

public Speaker(String firstName, String lastName, Party party) {
    this.firstName = firstName;
    this.lastName = lastName;
    this.party = party;
}
}

派对类

@Entity
@NamedQueries({
@NamedQuery(name = "Party.findPartyByName", query = "select p from Party p where p.name = :name")})

public class Party implements Serializable {

private static final long serialVersionUID = 1L;
public static final String FIND_BY_NAME = "Party.findPartyByName";

@Id
@GeneratedValue
@Column(name = "party_id")
private int id;

private String name;

@OneToMany(cascade = { CascadeType.ALL }, mappedBy = "party")
private List<Speaker> speakerList = new ArrayList<Speaker>();

public Party() {

}

public Party(String name) {
    this.name = name;
}
}

如何在扬声器和派对上设置CascadeType，当我坚持使用语料库时，扬声器和派对也会被持久化。

我的测试代码如下：

CorpusFacade cf = new CorpusFacade();

Corpus corpus = new Corpus();
Document document = new Document(corpus);
corpus.addDocument(document);

Party party = new Party("ASDF");
Speaker speaker = new Speaker("David", "Müller", party);

TextBlock textBlock1 = new TextBlock(speaker, "TB1", document);
document.addTextBlock(textBlock1);
TextBlock textBlock2 = new TextBlock(speaker, "TB1", document);
document.addTextBlock(textBlock2);

cf.createCorpus(corpus);

修改对于那些在我解决问题的过程中有兴趣的人。我首先创建了一个新的演讲者和一个新派对。然后我坚持党（议长自动坚持）。之后，我将Speaker添加到TextBlock。就这样。我无法找到语料库创建议长和派对的解决方案。我希望它适用于某人。

在OneToMany中使用Hibernate保持对象

2 个答案: