tudose-spring-data-hibernate

Imported from zademy/book-to-skill-zademy (.agents/skills/tudose-spring-data-hibernate/SKILL.md). Install upstream with npx skills add zademy/book-to-skill-zademy --skill tudose-spring-data-hibernate. Copyright stays with the author.

Java Persistence with Spring Data and Hibernate

Author: Cătălin Tudose | Pages: ~618 | Chapters: 20 | Generated: 2026-07-26

How to Use This Skill

• Without arguments — load core frameworks for reference
• With a topic — ask about optimistic locking, fetch plans, or another indexed topic; I find and read the relevant chapter
• With chapter — ask for ch11; I load that specific chapter
• Browse — ask "what chapters do you have?" to see the full index

When you ask about a topic not covered in Core Frameworks below, I will read the relevant chapter file before answering.

Core Frameworks & Mental Models

The Paradigm Mismatch (Ch1)

ORM exists to bridge 5 object/relational mismatch problems: granularity, inheritance, identity, associations, data navigation. The #1 performance lever is minimizing database requests — watch for the two classic failures: n+1 selects and the Cartesian product (never eager-fetch two collections at once). The stack: JPA = spec, Hibernate = implementation, Spring Data = convenience layer over both.

Lazy-Global Fetch Discipline (Ch8, Ch12)

Make every association LAZY globally, then opt into eager loading per query. @ManyToOne/@OneToOne default to EAGER — override to FetchType.LAZY. Separate the three concepts: fetch plan (what to load), fetch strategy (how: join/select/subselect/batch), fetch profile (named, reusable plan). Fix n+1 with a JOIN FETCH query, an entity graph, @BatchSize(N), or @Fetch(SUBSELECT) — not by flipping the mapping to EAGER.

Concurrency Default: READ_COMMITTED + @Version (Ch11)

Use READ_COMMITTED isolation + optimistic locking with a numeric @Version as the default recipe; it prevents lost updates without holding locks. Escalate only when needed: OPTIMISTIC_FORCE_INCREMENT for aggregate roots, PESSIMISTIC_WRITE (→ SELECT … FOR UPDATE) for hot rows. Spring @Transactional defaults to REQUIRED; use REQUIRES_NEW for audit logs that must survive a rollback, and rollbackFor to roll back on checked exceptions.

Entity States & Business-Key Equality (Ch10)

Four states: transient, persistent, removed, detached. The persistence context is dirty-checking + first-level cache + identity scope (watch OOM on bulk loads). merge() returns a new managed instance — use its return value. Implement equals()/hashCode() on a business key, never on the generated id (null before flush) and never on all fields.

Identifier & Value-Type Mapping (Ch5, Ch6)

Always use surrogate keys with wrapper types (Long, not long). Prefer the enhanced-sequence generator (allows JDBC batching) over IDENTITY. Map value types up the type-adaptation ladder: built-in type → AttributeConverter → @Embeddable → CompositeUserType. Use @Enumerated(STRING); enforce nullability with @Column(nullable=false) (Bean Validation @NotNull alone does not shape DDL).

Inheritance Decision Rule §7.7 (Ch7)

• No polymorphism / no polymorphic queries → TABLE_PER_CLASS
• Polymorphism + only behavior varies → SINGLE_TABLE (nullable columns)
• Polymorphism + subclass data / need NOT NULL → JOINED

Technology Selection (Ch14–19)

• Full JPA power / custom queries → Spring Data JPA (or a DAO with @PersistenceContext(EXTENDED) when you must bypass Spring Data).
• Aggregates, no lazy/dirty-checking overhead → Spring Data JDBC ("Spring Data JPA minus the persistence context"; its own annotations, not JPA's).
• Repository → hypermedia HTTP API → Spring Data REST (@Version doubles as ETag; PATCH = partial, PUT = full replace; lock down with exported=false).
• Document store, same repository model → Spring Data MongoDB (@DBRef = raw pointer, no cascading; enable auto-index-creation).
• Portable, type-safe, readable queries → Querydsl (Q-types via APT; its DML ignores cascades & 2nd-level cache).
• NoSQL behind JPA, switch backend by config only → Hibernate OGM.

Chapter Index

Topic Index

• AttributeConverter → ch06
• Bag / collections → ch08
• Business key (equals/hashCode) → ch10
• Cascading → ch08, ch13
• Cartesian product → ch09, ch12
• DAO / EXTENDED context → ch14
• Derived queries / @Query → ch04
• Dynamic filters → ch13
• Embeddable → ch05, ch06
• Enhanced-sequence generator → ch05
• Entity graph → ch12
• Entity states → ch10
• Envers (auditing) → ch13
• ETag / optimistic locking → ch11, ch16
• Fetch plan / strategy / profile → ch12
• Inheritance strategies → ch07
• Interceptor (audit) → ch13
• JPQL → ch02
• Join fetch → ch12
• Link entity (many-to-many) → ch09
• MongoTemplate / @DBRef → ch17
• n+1 selects → ch01, ch12
• Optimistic / pessimistic locking → ch11
• Paradigm mismatch → ch01
• Persistence context → ch02, ch10
• Projections → ch04, ch16
• Querydsl / Q-type → ch19
• Query by Example (QBE) → ch04, ch17
• Repository hierarchy → ch04
• REST events / excerpt projection → ch16
• Spring Data JDBC / aggregate → ch15
• Surrogate key → ch05
• Transaction propagation → ch11
• Type-adaptation ladder → ch06
• @Version → ch11, ch16

Supporting Files

• glossary.md — all key terms with definitions and chapter references
• patterns.md — 15 concrete techniques (identifier generation, locking recipes, DAO, auditing, aggregates, testing)
• cheatsheet.md — decision rules, fetch/locking/inheritance selection tables, thresholds

Scope & Limits

This skill covers the book's content only — Spring Data + Hibernate persistence patterns as taught by Tudose. For hands-on implementation in your codebase, combine with project-specific config and schema. For topics beyond this book (e.g. reactive R2DBC, Hibernate Reactive), check related skills or ask the agent directly.