azure-data-tables-java – Coding skill for AI agents

Azure Tables SDK for Java

Build table storage applications using the Azure Tables SDK for Java. Works with both Azure Table Storage and Cosmos DB Table API.

Installation

<dependency>
  <groupId>com.azure</groupId>
  <artifactId>azure-data-tables</artifactId>
  <version>12.6.0-beta.1</version>
</dependency>

Client Creation

With Connection String

import com.azure.data.tables.TableServiceClient;
import com.azure.data.tables.TableServiceClientBuilder;
import com.azure.data.tables.TableClient;

TableServiceClient serviceClient = new TableServiceClientBuilder()
    .connectionString("<your-connection-string>")
    .buildClient();

With Shared Key

import com.azure.core.credential.AzureNamedKeyCredential;

AzureNamedKeyCredential credential = new AzureNamedKeyCredential(
    "<account-name>",
    "<account-key>");

TableServiceClient serviceClient = new TableServiceClientBuilder()
    .endpoint("<your-table-account-url>")
    .credential(credential)
    .buildClient();

With SAS Token

TableServiceClient serviceClient = new TableServiceClientBuilder()
    .endpoint("<your-table-account-url>")
    .sasToken("<sas-token>")
    .buildClient();

With DefaultAzureCredential (Storage only)

import com.azure.identity.DefaultAzureCredentialBuilder;

TableServiceClient serviceClient = new TableServiceClientBuilder()
    .endpoint("<your-table-account-url>")
    .credential(new DefaultAzureCredentialBuilder().build())
    .buildClient();

Key Concepts

• TableServiceClient: Manage tables (create, list, delete)
• TableClient: Manage entities within a table (CRUD)
• Partition Key: Groups entities for efficient queries
• Row Key: Unique identifier within a partition
• Entity: A row with up to 252 properties (1MB Storage, 2MB Cosmos)

Core Patterns

Examples use placeholder credentials. Select a compatible released SDK version; the beta above is historical. Keep connection strings/SAS in a secret store. Confirm the account/table and deletion scope before writes. Merge/replace/upsert examples do not demonstrate ETag concurrency checks; use conditional updates where concurrent writers matter.

Create Table

// Create table (throws if exists)
TableClient tableClient = serviceClient.createTable("mytable");

// Create if not exists (no exception)
TableClient tableClient = serviceClient.createTableIfNotExists("mytable");

Get Table Client

// From service client
TableClient tableClient = serviceClient.getTableClient("mytable");

// Direct construction
TableClient tableClient = new TableClientBuilder()
    .connectionString("<connection-string>")
    .tableName("mytable")
    .buildClient();

Create Entity

import com.azure.data.tables.models.TableEntity;

TableEntity entity = new TableEntity("partitionKey", "rowKey")
    .addProperty("Name", "Product A")
    .addProperty("Price", 29.99)
    .addProperty("Quantity", 100)
    .addProperty("IsAvailable", true);

tableClient.createEntity(entity);

Get Entity

TableEntity entity = tableClient.getEntity("partitionKey", "rowKey");

String name = (String) entity.getProperty("Name");
Double price = (Double) entity.getProperty("Price");
System.out.printf("Product: %s, Price: %.2f%n", name, price);

Update Entity

import com.azure.data.tables.models.TableEntityUpdateMode;

// Merge (update only specified properties)
TableEntity updateEntity = new TableEntity("partitionKey", "rowKey")
    .addProperty("Price", 24.99);
tableClient.updateEntity(updateEntity, TableEntityUpdateMode.MERGE);

// Replace (replace entire entity)
TableEntity replaceEntity = new TableEntity("partitionKey", "rowKey")
    .addProperty("Name", "Product A Updated")
    .addProperty("Price", 24.99)
    .addProperty("Quantity", 150);
tableClient.updateEntity(replaceEntity, TableEntityUpdateMode.REPLACE);

Upsert Entity

// Insert or update (merge mode)
tableClient.upsertEntity(entity, TableEntityUpdateMode.MERGE);

// Insert or replace
tableClient.upsertEntity(entity, TableEntityUpdateMode.REPLACE);

Delete Entity

tableClient.deleteEntity("partitionKey", "rowKey");

List Entities

import com.azure.data.tables.models.ListEntitiesOptions;

// List all entities
for (TableEntity entity : tableClient.listEntities()) {
    System.out.printf("%s - %s%n",
        entity.getPartitionKey(),
        entity.getRowKey());
}

// With filtering and selection
ListEntitiesOptions options = new ListEntitiesOptions()
    .setFilter("PartitionKey eq 'sales'")
    .setSelect(java.util.Arrays.asList("Name", "Price"));

for (TableEntity entity : tableClient.listEntities(options, null, null)) {
    System.out.printf("%s: %.2f%n",
        entity.getProperty("Name"),
        entity.getProperty("Price"));
}

Query with OData Filter

// Filter by partition key
ListEntitiesOptions options = new ListEntitiesOptions()
    .setFilter("PartitionKey eq 'electronics'");

// Filter with multiple conditions
options.setFilter("PartitionKey eq 'electronics' and Price gt 100");

// Filter with comparison operators
options.setFilter("Quantity ge 10 and Quantity le 100");

// Request page size; stop iteration explicitly if only ten total results are wanted
options.setTop(10);

for (TableEntity entity : tableClient.listEntities(options, null, null)) {
    System.out.println(entity.getRowKey());
}

Batch Operations (Transactions)

import com.azure.data.tables.models.TableTransactionAction;
import com.azure.data.tables.models.TableTransactionActionType;
import java.util.Arrays;

// All entities must have same partition key
List<TableTransactionAction> actions = Arrays.asList(
    new TableTransactionAction(
        TableTransactionActionType.CREATE,
        new TableEntity("batch", "row1").addProperty("Name", "Item 1")),
    new TableTransactionAction(
        TableTransactionActionType.CREATE,
        new TableEntity("batch", "row2").addProperty("Name", "Item 2")),
    new TableTransactionAction(
        TableTransactionActionType.UPSERT_MERGE,
        new TableEntity("batch", "row3").addProperty("Name", "Item 3"))
);

tableClient.submitTransaction(actions);

List Tables

import com.azure.data.tables.models.TableItem;
import com.azure.data.tables.models.ListTablesOptions;

// List all tables
for (TableItem table : serviceClient.listTables()) {
    System.out.println(table.getName());
}

// Filter tables
ListTablesOptions options = new ListTablesOptions()
    .setFilter("TableName eq 'mytable'");

for (TableItem table : serviceClient.listTables(options, null, null)) {
    System.out.println(table.getName());
}

Delete Table

serviceClient.deleteTable("mytable");

Typed Entities

TableEntity is a concrete SDK class, not an interface for a POJO to implement. Map domain values explicitly:

TableEntity product = new TableEntity("electronics", "laptop-001")
    .addProperty("Name", "Laptop")
    .addProperty("Price", 999.99);
tableClient.createEntity(product);

Error Handling

import com.azure.data.tables.models.TableServiceException;

try {
    tableClient.createEntity(entity);
} catch (TableServiceException e) {
    System.out.println("Status: " + e.getResponse().getStatusCode());
    System.out.println("Error: " + e.getMessage());
    // 409 = Conflict (entity exists)
    // 404 = Not Found
}

Environment Variables

# Storage Account
AZURE_TABLES_CONNECTION_STRING=DefaultEndpointsProtocol=https;AccountName=...
AZURE_TABLES_ENDPOINT=https://<account>.table.core.windows.net

# Cosmos DB Table API
COSMOS_TABLE_ENDPOINT=https://<account>.table.cosmosdb.azure.com

Best Practices

1. Partition Key Design: Choose keys that distribute load evenly

2. Batch Operations: Use transactions for atomic multi-entity updates

3. Query Optimization: Always filter by PartitionKey when possible

4. Select Projection: Only select needed properties for performance

5. Entity Size: Keep entities under 1MB (Storage) or 2MB (Cosmos)

Trigger Phrases

• "Azure Tables Java"
• "table storage SDK"
• "Cosmos DB Table API"
• "NoSQL key-value storage"
• "partition key row key"
• "table entity CRUD"

When to Use

This skill is applicable to execute the workflow or actions described in the overview.

Limitations

• Use this skill only when the task clearly matches the scope described above.
• Do not treat the output as a substitute for environment-specific validation, testing, or expert review.
• Stop and ask for clarification if required inputs, permissions, safety boundaries, or success criteria are missing.