Dataflow 写入组

使用 Dataflow 连接器和 MutationGroup 类写入数据,从而确保以原子方式应用一组变更。

深入探索

如需查看包含此代码示例的详细文档,请参阅以下内容:

代码示例

Java

如需了解如何安装和使用 Spanner 客户端库,请参阅 Spanner 客户端库

如需向 Spanner 进行身份验证,请设置应用默认凭据。 如需了解详情,请参阅为本地开发环境设置身份验证

PCollection<MutationGroup> mutations =
    suspiciousUserIds.apply(
        MapElements.via(
            new SimpleFunction<>() {

              @Override
              public MutationGroup apply(String userId) {
                // Immediately block the user.
                Mutation userMutation =
                    Mutation.newUpdateBuilder("Users")
                        .set("id")
                        .to(userId)
                        .set("state")
                        .to("BLOCKED")
                        .build();
                long generatedId =
                    Hashing.sha1()
                        .newHasher()
                        .putString(userId, Charsets.UTF_8)
                        .putLong(timestamp.getSeconds())
                        .putLong(timestamp.getNanos())
                        .hash()
                        .asLong();

                // Add an entry to pending review requests.
                Mutation pendingReview =
                    Mutation.newInsertOrUpdateBuilder("PendingReviews")
                        .set("id")
                        .to(generatedId) // Must be deterministically generated.
                        .set("userId")
                        .to(userId)
                        .set("action")
                        .to("REVIEW ACCOUNT")
                        .set("note")
                        .to("Suspicious activity detected.")
                        .build();

                return MutationGroup.create(userMutation, pendingReview);
              }
            }));

mutations.apply(SpannerIO.write()
    .withInstanceId(instanceId)
    .withDatabaseId(databaseId)
    .grouped());

后续步骤

如需搜索和过滤其他 Google Cloud 产品的代码示例,请参阅 Google Cloud 示例浏览器