Firebase Auth & Realtime Database Apps · 课时

事务性数据操作

学习使用事务防止竞态条件，并确保关键数据以原子方式更新

事务性数据操作 是 CoddyKit 上的免费 Firebase Auth & Realtime Database Apps 课时。这是第 2 节课，共 4 节。你可以在下方免费阅读本课时的完整内容 — 然后在浏览器中使用内置代码编辑器和全天候 AI 导师进行实践。这是 Firebase Auth & Realtime Database Apps 学习路径的一部分，你的进度在网页和 CoddyKit 应用中同步。 Firebase Auth & Realtime Database Apps 课程共包含 4 节课。

本课时的部分内容尚未翻译，以英文显示。

Why Data Integrity Matters

In real-time applications, multiple users might try to update the same data simultaneously. This can lead to serious issues like data corruption or inconsistencies if not handled correctly.

Imagine a simple counter on a website. If two users click 'Like' at the exact same moment, the counter should increment by two, not just one.

The Race Condition Problem

Without proper synchronization, a common scenario called a race condition can occur. This happens when the outcome of an operation depends on the unpredictable sequence or timing of other operations.

For example, if you read a value, increment it, and then write it back, another user might read the original value before you write your incremented one, causing an update to be lost.

Introducing Firebase Transactions

Firebase Realtime Database provides a powerful feature called transactions to solve race conditions and ensure data integrity. A transaction guarantees an atomic update.

Atomic means the operation either completes entirely or doesn't happen at all. It's like a single, unbreakable step.

How `runTransaction` Works

You initiate a transaction using the runTransaction() method on a DatabaseReference. This method takes a Transaction.Handler callback.

Firebase passes the current state of the data to your handler.
You modify this data within the handler.
If another client writes to the same location while your transaction is running, Firebase automatically retries your transaction with the new current data.

Implementing a Safe Counter

Let's see how to safely increment a counter using a transaction. This ensures that even if multiple users try to increment simultaneously, the count will always be correct.

Try running this example:

public class Main {

  // Mock Firebase classes for demonstration
  static class MockFirebaseDatabase {
    private Integer value = 0; // Simulate data at a path
    public Integer get() { return value; }
    public void set(Integer val) { value = val; }

    public interface TransactionHandler {
      TransactionResult doTransaction(MutableData currentData);
    }

    public static class MutableData {
      private Integer data;
      public MutableData(Integer data) { this.data = data; }
      public Integer getValue() { return data; }
      public void setValue(Integer data) { this.data = data; }
    }

    public static class TransactionResult {
      private boolean success;
      private MutableData newData;
      private TransactionResult(boolean success, MutableData newData) {
        this.success = success;
        this.newData = newData;
      }
      public static TransactionResult success(MutableData newData) {
        return new TransactionResult(true, newData);
      }
      public static TransactionResult abort() {
        return new TransactionResult(false, null);
      }
      public boolean isSuccess() { return success; }
      public MutableData getNewData() { return newData; }
    }

    public void runTransaction(TransactionHandler handler) {
      // Simulate read, modify, and retry logic
      MutableData currentData = new MutableData(this.get());
      TransactionResult result = handler.doTransaction(currentData);

      if (result.isSuccess()) {
        this.set(result.getNewData().getValue());
        System.out.println("Transaction committed. New value: " + this.get());
      } else {
        System.out.println("Transaction aborted.");
      }
    }
  }

  public static void main(String[] args) {
    MockFirebaseDatabase counterRef = new MockFirebaseDatabase();
    counterRef.set(5); // Initial value

    counterRef.runTransaction(new MockFirebaseDatabase.TransactionHandler() {
      @Override
      public MockFirebaseDatabase.TransactionResult doTransaction(MockFirebaseDatabase.MutableData currentData) {
        Integer currentValue = currentData.getValue();
        if (currentValue == null) {
          currentValue = 0;
        }
        currentData.setValue(currentValue + 1);
        return MockFirebaseDatabase.TransactionResult.success(currentData);
      }
    });
  }
}

Understanding `MutableData`

Inside your Transaction.Handler, the MutableData object represents the data at the database location you're trying to modify.

Use currentData.getValue() to read the existing value.
Use currentData.setValue(newValue) to set the new value you want to write.

Remember, this is the data Firebase will try to commit. If a conflict occurs, your handler will be called again with the updated MutableData.

`TransactionResult` and Aborting

After processing the MutableData, your handler must return a Transaction.Result:

Transaction.Result.success(mutableData): Tells Firebase to try to commit the new value in mutableData.
Transaction.Result.abort(): Tells Firebase to cancel the transaction. This is useful if the data is in an unexpected state or if your logic determines the transaction shouldn't proceed.

Here's an example of aborting a transaction:

public class Main {

  // Mock Firebase classes (repeated for full program requirement)
  static class MockFirebaseDatabase {
    private Integer value = 0;
    public Integer get() { return value; }
    public void set(Integer val) { value = val; }

    public interface TransactionHandler {
      TransactionResult doTransaction(MutableData currentData);
    }

    public static class MutableData {
      private Integer data;
      public MutableData(Integer data) { this.data = data; }
      public Integer getValue() { return data; }
      public void setValue(Integer data) { this.data = data; }
    }

    public static class TransactionResult {
      private boolean success;
      private MutableData newData;
      private TransactionResult(boolean success, MutableData newData) {
        this.success = success;
        this.newData = newData;
      }
      public static TransactionResult success(MutableData newData) {
        return new TransactionResult(true, newData);
      }
      public static TransactionResult abort() {
        return new TransactionResult(false, null);
      }
      public boolean isSuccess() { return success; }
      public MutableData getNewData() { return newData; }
    }

    public void runTransaction(TransactionHandler handler) {
      MutableData currentData = new MutableData(this.get());
      TransactionResult result = handler.doTransaction(currentData);

      if (result.isSuccess()) {
        this.set(result.getNewData().getValue());
        System.out.println("Transaction committed. New value: " + this.get());
      } else {
        System.out.println("Transaction aborted.");
      }
    }
  }

  public static void main(String[] args) {
    MockFirebaseDatabase statusRef = new MockFirebaseDatabase();
    statusRef.set(1); // 1 = Active, 0 = Inactive

    // Try to change status, but abort if it's already Inactive (0)
    statusRef.runTransaction(new MockFirebaseDatabase.TransactionHandler() {
      @Override
      public MockFirebaseDatabase.TransactionResult doTransaction(MockFirebaseDatabase.MutableData currentData) {
        Integer status = currentData.getValue();
        if (status != null && status == 0) {
          System.out.println("Status is already Inactive. Aborting transaction.");
          return MockFirebaseDatabase.TransactionResult.abort();
        }
        // Change status to 0 (Inactive)
        currentData.setValue(0);
        return MockFirebaseDatabase.TransactionResult.success(currentData);
      }
    });
  }
}

Handling Transaction Completion

After calling runTransaction(), you'll typically want to know if it succeeded or failed. Firebase provides an onComplete callback for this.

This callback gives you:

error: If the transaction failed.
committed: A boolean indicating if the transaction was committed.
currentData: The final state of the data.

Use this callback to update your UI or handle any post-transaction logic.

Beyond Simple Counters

Transactions are invaluable for any scenario requiring strong data consistency:

Unique Usernames: Ensure a username is truly unique before assigning it.
Voting Systems: Prevent double-voting or ensure vote counts are accurate.
Inventory Management: Safely decrement stock levels without overselling.
Game Scores: Update high scores reliably in multiplayer games.

Quick Check

Transactions are crucial for maintaining data integrity in concurrent environments. Which of the following best describes the primary benefit of using Firebase Realtime Database transactions?

Recap & Next Steps

You've learned about the critical role of transactional data operations in maintaining data integrity in real-time applications.

We explored race conditions and why they're problematic.
You now understand how Firebase's runTransaction() method ensures atomic updates.
We saw examples of safely incrementing counters and using Transaction.Result.abort().

Transactions are a powerful tool, but use them judiciously as they can be slower than direct writes. In the next lesson, we'll dive into atomic counters and queues!

常见问题解答

「事务性数据操作」课时是免费的吗？

是的 — 「事务性数据操作」的完整文本可在网页上免费阅读。要进行交互式练习（内置代码编辑器和全天候 AI 导师）并解锁 Firebase Auth & Realtime Database Apps 课程的其余内容，请升级到 CoddyKit PRO。 Firebase Auth & Realtime Database Apps 课程共包含 4 节课。

「事务性数据操作」这节课中我会学到什么？

学习使用事务防止竞态条件，并确保关键数据以原子方式更新你通过在浏览器中直接运行的动手代码来练习 Firebase Auth & Realtime Database Apps，全天候 AI 导师会在你学习这节课的过程中回答你的问题。

学习 Firebase Auth & Realtime Database Apps 需要有经验吗？

无需任何先前经验。CoddyKit 上的 Firebase Auth & Realtime Database Apps 课程适合初学者到高级学习者，你可以从这里开始或从头开始，按照自己的节奏学习。这是第 2 节课，共 4 节。

「事务性数据操作」课时需要多长时间？

大多数 CoddyKit 课程大约需要 5–10 分钟。每节课都很精短且互动，所以你能稳步进步，并在网页和应用中从离开的地方继续。

我能在这节 Firebase Auth & Realtime Database Apps 课中编写并运行代码吗？

能。每节 Firebase Auth & Realtime Database Apps 课都包含内置代码编辑器，你可以在浏览器中直接编写并运行真实代码，并获得即时 AI 反馈 — 无需本地设置。

此课程中的所有课时

← 返回 Firebase Auth & Realtime Database Apps