matrixone-python-sdk 0.1.9__tar.gz → 0.1.11__tar.gz

{matrixone_python_sdk-0.1.9/matrixone_python_sdk.egg-info → matrixone_python_sdk-0.1.11}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: matrixone-python-sdk
-Version: 0.1.9
+Version: 0.1.11
 Summary: A comprehensive Python SDK for MatrixOne database operations with vector search, fulltext search, and advanced features
 Home-page: https://github.com/matrixorigin/matrixone
 Author: MatrixOne Team
@@ -106,6 +106,7 @@ A comprehensive, high-level Python SDK for MatrixOne that provides SQLAlchemy-li
   - Export to local files or external stages (``stage://`` protocol)
   - Support for raw SQL, SQLAlchemy, and MatrixOne queries
   - Transaction-aware exports for consistency
+- 🔁 **Change Data Capture (CDC)**: Create, monitor, and control CDC tasks directly from Python (`client.cdc`, `async_client.cdc`)
 - 📸 **Snapshot Management**: Create and manage database snapshots at multiple levels
 - ⏰ **Point-in-Time Recovery**: PITR functionality for precise data recovery
 - 🔄 **Table Cloning**: Clone databases and tables efficiently
@@ -171,6 +172,37 @@ conda activate matrixone
 pip install matrixone-python-sdk
 ```
 
+## ⚠️ Important: Column Naming Convention
+
+**🚨 CRITICAL: Always use lowercase with underscores (snake_case) for column names!**
+
+MatrixOne does not support SQL standard double-quoted identifiers in queries, which causes issues with camelCase column names when using SQLAlchemy ORM.
+
+```python
+# ❌ DON'T: CamelCase column names (will fail in SELECT queries)
+class User(Base):
+    userName = Column(String(50))      # CREATE succeeds, SELECT fails!
+    userId = Column(Integer)           # Will cause SQL syntax errors
+
+# ✅ DO: Use lowercase with underscores (snake_case)
+class User(Base):
+    user_name = Column(String(50))     # Works perfectly
+    user_id = Column(Integer)          # All operations succeed
+```
+
+**Why this matters:**
+- ✅ CREATE TABLE works with both styles (uses backticks)
+- ✅ INSERT works with both styles  
+- ❌ **SELECT fails with camelCase** (uses double quotes, not supported by MatrixOne)
+
+**Example of the problem:**
+```python
+# CamelCase generates: SELECT "userName" FROM user  ❌ Fails!
+# snake_case generates: SELECT user_name FROM user  ✅ Works!
+```
+
+---
+
 ## Quick Start
 
 ### Basic Usage
@@ -317,6 +349,152 @@ client.export.to_csv_stage('s3_stage', 'backup2.csv', stmt)
 client.disconnect()
 ```
 
+### CDC Task Management
+
+Create and control Change Data Capture (CDC) tasks directly from the Python SDK:
+
+```python
+from matrixone import Client, build_mysql_uri
+
+client = Client()
+client.connect(database='test')
+
+source_uri = build_mysql_uri('127.0.0.1', 6001, user='admin', password='111', account='acct')
+mysql_sink = build_mysql_uri('192.168.1.100', 3306, user='root', password='111')
+
+# Database-level replication (automatically ensures Level='database')
+client.cdc.create_database_task(
+    task_name='replicate_sales',
+    source_uri=source_uri,
+    sink_type='mysql',
+    sink_uri=mysql_sink,
+    source_database='sales',
+    sink_database='sales_archive',
+    options={'Frequency': '30m'}
+)
+
+# Table-level replication helper (list of tuples/dicts/strings)
+client.cdc.create_table_task(
+    task_name='replicate_key_tables',
+    source_uri=source_uri,
+    sink_type='mysql',
+    sink_uri=mysql_sink,
+    table_mappings=[('sales', 'orders', 'sales_archive', 'orders_backup'), 'sales.customers:sales_archive.customers']
+)
+
+# Pause/resume lifecycle
+client.cdc.pause('replicate_sales')
+client.cdc.resume('replicate_sales')
+
+# Inspect status and per-table watermarks
+tasks = client.cdc.list()
+watermarks = client.cdc.list_watermarks('replicate_sales')
+
+# Drop task when no longer needed
+client.cdc.drop('replicate_sales')
+client.cdc.drop('replicate_key_tables')
+
+client.disconnect()
+```
+
+> Async usage: call the same helpers via `await async_client.cdc.create_database_task(...)` / `create_table_task(...)` or inside `async with client.session()`.
+
+### CDC API Overview
+
+The :class:`matrixone.cdc.CDCManager` class provides high-level helpers for all CDC lifecycle operations. Typical workflows are:
+
+```python
+from matrixone import Client
+from matrixone.cdc import CDCManager
+
+client = Client()
+client.connect(host='127.0.0.1', port=6001, user='root', password='111', database='test')
+
+cdc = CDCManager(client)
+
+# Create a table-level task
+task = cdc.create_table_task(
+    task_name='orders_sync',
+    source_uri='mysql://sys#root:111@127.0.0.1:6001',
+    sink_type='matrixone',
+    sink_uri='mysql://sys#root:111@127.0.0.1:6001',
+    table_mappings=[('sales', 'orders', 'backup', 'orders')],
+    options={'Frequency': '1h', 'NoFull': True},
+)
+
+# Pause and resume
+cdc.pause(task.task_name)
+cdc.resume(task.task_name)
+
+# Inspect metadata
+info = cdc.get(task.task_name)
+print(info.state)
+
+# Clean up
+cdc.drop(task.task_name)
+client.disconnect()
+```
+
+Key helper methods:
+
+* `create()` – create a CDC task from a raw mapping string.
+* `create_database_task()` – convenience wrapper for database-level replication.
+* `create_table_task()` – convenience wrapper for table-level replication using Python data structures.
+* `pause()`/`resume()`/`restart()` – control CDC task state.
+* `list()`/`get()` – retrieve CDC task metadata as :class:`matrixone.cdc.CDCTaskInfo` objects.
+* `list_watermarks()` – read per-table watermarks as :class:`matrixone.cdc.CDCWatermarkInfo`.
+* `list_failing_tasks()` – return tasks whose `err_msg` field is populated.
+* `list_stuck_tasks()` – highlight running tasks that have per-table errors.
+* `list_late_table_watermarks()` – detect tables whose watermarks lag behind expected thresholds (supports custom per-task/per-table overrides).
+
+Refer to :class:`matrixone.cdc.CDCManager` docstrings or ``examples/example_31_cdc_operations.py`` for a comprehensive lifecycle example.
+
+```python
+from matrixone import Client, build_mysql_uri
+
+client = Client()
+client.connect(database='test')
+
+source_uri = build_mysql_uri('127.0.0.1', 6001, user='admin', password='111', account='acct')
+mysql_sink = build_mysql_uri('192.168.1.100', 3306, user='root', password='111')
+
+# Database-level replication (automatically ensures Level='database')
+client.cdc.create_database_task(
+    task_name='replicate_sales',
+    source_uri=source_uri,
+    sink_type='mysql',
+    sink_uri=mysql_sink,
+    source_database='sales',
+    sink_database='sales_archive',
+    options={'Frequency': '30m'}
+)
+
+# Table-level replication helper (list of tuples/dicts/strings)
+client.cdc.create_table_task(
+    task_name='replicate_key_tables',
+    source_uri=source_uri,
+    sink_type='mysql',
+    sink_uri=mysql_sink,
+    table_mappings=[('sales', 'orders', 'sales_archive', 'orders_backup'), 'sales.customers:sales_archive.customers']
+)
+
+# Pause/resume lifecycle
+client.cdc.pause('replicate_sales')
+client.cdc.resume('replicate_sales')
+
+# Inspect status and per-table watermarks
+tasks = client.cdc.list()
+watermarks = client.cdc.list_watermarks('replicate_sales')
+
+# Drop task when no longer needed
+client.cdc.drop('replicate_sales')
+client.cdc.drop('replicate_key_tables')
+
+client.disconnect()
+```
+
+> Async usage: call the same helpers via `await async_client.cdc.create_database_task(...)` / `create_table_task(...)` or inside `async with client.session()`.
+
 ### Wrapping Existing Sessions (For Legacy Code)
 
 If you have existing SQLAlchemy code, wrap your sessions to add MatrixOne features:
@@ -554,6 +732,26 @@ mo-diag -d test -c "sql SELECT COUNT(*) FROM my_table"
 mo-diag -d test -c "flush_table my_table"
 ```
 
+#### CDC Task Shortcuts
+
+CDC management is integrated with the CLI. The interactive commands (`cdc_tasks`,
+`cdc_task`, `cdc_create`, `cdc_drop`) now have non-interactive equivalents
+under `mo-diag cdc`, making automation straightforward.
+
+```bash
+# Summaries (mirrors cdc_tasks)
+mo-diag cdc show
+
+# Inspect a task and tighten watermark thresholds
+mo-diag cdc show nightly_sync --details --threshold=5m
+
+# Start the guided creator in table mode
+mo-diag cdc create --table-level
+
+# Drop a task without interactive prompts
+mo-diag cdc drop nightly_sync --force
+```
+
 ### Available Commands
 
 #### Index Management

{matrixone_python_sdk-0.1.9 → matrixone_python_sdk-0.1.11}/README.md

@@ -70,6 +70,37 @@ A comprehensive Python SDK for MatrixOne that provides SQLAlchemy-like interface
   - Non-interactive mode for scripting and automation
   - Batch operations on tables and indexes
 
+## ⚠️ Important: Column Naming Convention
+
+**🚨 CRITICAL: Always use lowercase with underscores (snake_case) for column names!**
+
+MatrixOne does not support SQL standard double-quoted identifiers in queries, which causes issues with camelCase column names when using SQLAlchemy ORM.
+
+```python
+# ❌ DON'T: CamelCase column names (will fail in SELECT queries)
+class User(Base):
+    userName = Column(String(50))      # CREATE succeeds, SELECT fails!
+    userId = Column(Integer)           # Will cause SQL syntax errors
+
+# ✅ DO: Use lowercase with underscores (snake_case)
+class User(Base):
+    user_name = Column(String(50))     # Works perfectly
+    user_id = Column(Integer)          # All operations succeed
+```
+
+**Why this matters:**
+- ✅ CREATE TABLE works with both styles (uses backticks)
+- ✅ INSERT works with both styles  
+- ❌ **SELECT fails with camelCase** (uses double quotes, not supported by MatrixOne)
+
+**Example of the problem:**
+```python
+# CamelCase generates: SELECT "userName" FROM user  ❌ Fails!
+# snake_case generates: SELECT user_name FROM user  ✅ Works!
+```
+
+---
+
 ## 🚀 Installation
 
 ### From PyPI (Stable Release)
@@ -793,6 +824,26 @@ mo-diag -d test -c "sql SELECT COUNT(*) FROM my_table"
 mo-diag -d test -c "flush_table my_table"
 ```
 
+#### CDC Task Shortcuts
+
+Manage Change Data Capture directly from the shell or scripts. The interactive
+commands (`cdc_tasks`, `cdc_task`, `cdc_create`, `cdc_drop`) now have matching
+non-interactive helpers via `mo-diag cdc`.
+
+```bash
+# Summaries (same as running cdc_tasks inside the shell)
+mo-diag cdc show
+
+# Inspect a task with watermark threshold control
+mo-diag cdc show nightly_sync --details --threshold=5m
+
+# Launch the guided creator with a forced mode
+mo-diag cdc create --table-level
+
+# Drop a task with explicit confirmation override
+mo-diag cdc drop nightly_sync --force
+```
+
 ### Available Commands
 
 #### Index Management

{matrixone_python_sdk-0.1.9 → matrixone_python_sdk-0.1.11}/README_USER.md

@@ -38,6 +38,7 @@ A comprehensive, high-level Python SDK for MatrixOne that provides SQLAlchemy-li
   - Export to local files or external stages (``stage://`` protocol)
   - Support for raw SQL, SQLAlchemy, and MatrixOne queries
   - Transaction-aware exports for consistency
+- 🔁 **Change Data Capture (CDC)**: Create, monitor, and control CDC tasks directly from Python (`client.cdc`, `async_client.cdc`)
 - 📸 **Snapshot Management**: Create and manage database snapshots at multiple levels
 - ⏰ **Point-in-Time Recovery**: PITR functionality for precise data recovery
 - 🔄 **Table Cloning**: Clone databases and tables efficiently
@@ -103,6 +104,37 @@ conda activate matrixone
 pip install matrixone-python-sdk
 ```
 
+## ⚠️ Important: Column Naming Convention
+
+**🚨 CRITICAL: Always use lowercase with underscores (snake_case) for column names!**
+
+MatrixOne does not support SQL standard double-quoted identifiers in queries, which causes issues with camelCase column names when using SQLAlchemy ORM.
+
+```python
+# ❌ DON'T: CamelCase column names (will fail in SELECT queries)
+class User(Base):
+    userName = Column(String(50))      # CREATE succeeds, SELECT fails!
+    userId = Column(Integer)           # Will cause SQL syntax errors
+
+# ✅ DO: Use lowercase with underscores (snake_case)
+class User(Base):
+    user_name = Column(String(50))     # Works perfectly
+    user_id = Column(Integer)          # All operations succeed
+```
+
+**Why this matters:**
+- ✅ CREATE TABLE works with both styles (uses backticks)
+- ✅ INSERT works with both styles  
+- ❌ **SELECT fails with camelCase** (uses double quotes, not supported by MatrixOne)
+
+**Example of the problem:**
+```python
+# CamelCase generates: SELECT "userName" FROM user  ❌ Fails!
+# snake_case generates: SELECT user_name FROM user  ✅ Works!
+```
+
+---
+
 ## Quick Start
 
 ### Basic Usage
@@ -249,6 +281,152 @@ client.export.to_csv_stage('s3_stage', 'backup2.csv', stmt)
 client.disconnect()
 ```
 
+### CDC Task Management
+
+Create and control Change Data Capture (CDC) tasks directly from the Python SDK:
+
+```python
+from matrixone import Client, build_mysql_uri
+
+client = Client()
+client.connect(database='test')
+
+source_uri = build_mysql_uri('127.0.0.1', 6001, user='admin', password='111', account='acct')
+mysql_sink = build_mysql_uri('192.168.1.100', 3306, user='root', password='111')
+
+# Database-level replication (automatically ensures Level='database')
+client.cdc.create_database_task(
+    task_name='replicate_sales',
+    source_uri=source_uri,
+    sink_type='mysql',
+    sink_uri=mysql_sink,
+    source_database='sales',
+    sink_database='sales_archive',
+    options={'Frequency': '30m'}
+)
+
+# Table-level replication helper (list of tuples/dicts/strings)
+client.cdc.create_table_task(
+    task_name='replicate_key_tables',
+    source_uri=source_uri,
+    sink_type='mysql',
+    sink_uri=mysql_sink,
+    table_mappings=[('sales', 'orders', 'sales_archive', 'orders_backup'), 'sales.customers:sales_archive.customers']
+)
+
+# Pause/resume lifecycle
+client.cdc.pause('replicate_sales')
+client.cdc.resume('replicate_sales')
+
+# Inspect status and per-table watermarks
+tasks = client.cdc.list()
+watermarks = client.cdc.list_watermarks('replicate_sales')
+
+# Drop task when no longer needed
+client.cdc.drop('replicate_sales')
+client.cdc.drop('replicate_key_tables')
+
+client.disconnect()
+```
+
+> Async usage: call the same helpers via `await async_client.cdc.create_database_task(...)` / `create_table_task(...)` or inside `async with client.session()`.
+
+### CDC API Overview
+
+The :class:`matrixone.cdc.CDCManager` class provides high-level helpers for all CDC lifecycle operations. Typical workflows are:
+
+```python
+from matrixone import Client
+from matrixone.cdc import CDCManager
+
+client = Client()
+client.connect(host='127.0.0.1', port=6001, user='root', password='111', database='test')
+
+cdc = CDCManager(client)
+
+# Create a table-level task
+task = cdc.create_table_task(
+    task_name='orders_sync',
+    source_uri='mysql://sys#root:111@127.0.0.1:6001',
+    sink_type='matrixone',
+    sink_uri='mysql://sys#root:111@127.0.0.1:6001',
+    table_mappings=[('sales', 'orders', 'backup', 'orders')],
+    options={'Frequency': '1h', 'NoFull': True},
+)
+
+# Pause and resume
+cdc.pause(task.task_name)
+cdc.resume(task.task_name)
+
+# Inspect metadata
+info = cdc.get(task.task_name)
+print(info.state)
+
+# Clean up
+cdc.drop(task.task_name)
+client.disconnect()
+```
+
+Key helper methods:
+
+* `create()` – create a CDC task from a raw mapping string.
+* `create_database_task()` – convenience wrapper for database-level replication.
+* `create_table_task()` – convenience wrapper for table-level replication using Python data structures.
+* `pause()`/`resume()`/`restart()` – control CDC task state.
+* `list()`/`get()` – retrieve CDC task metadata as :class:`matrixone.cdc.CDCTaskInfo` objects.
+* `list_watermarks()` – read per-table watermarks as :class:`matrixone.cdc.CDCWatermarkInfo`.
+* `list_failing_tasks()` – return tasks whose `err_msg` field is populated.
+* `list_stuck_tasks()` – highlight running tasks that have per-table errors.
+* `list_late_table_watermarks()` – detect tables whose watermarks lag behind expected thresholds (supports custom per-task/per-table overrides).
+
+Refer to :class:`matrixone.cdc.CDCManager` docstrings or ``examples/example_31_cdc_operations.py`` for a comprehensive lifecycle example.
+
+```python
+from matrixone import Client, build_mysql_uri
+
+client = Client()
+client.connect(database='test')
+
+source_uri = build_mysql_uri('127.0.0.1', 6001, user='admin', password='111', account='acct')
+mysql_sink = build_mysql_uri('192.168.1.100', 3306, user='root', password='111')
+
+# Database-level replication (automatically ensures Level='database')
+client.cdc.create_database_task(
+    task_name='replicate_sales',
+    source_uri=source_uri,
+    sink_type='mysql',
+    sink_uri=mysql_sink,
+    source_database='sales',
+    sink_database='sales_archive',
+    options={'Frequency': '30m'}
+)
+
+# Table-level replication helper (list of tuples/dicts/strings)
+client.cdc.create_table_task(
+    task_name='replicate_key_tables',
+    source_uri=source_uri,
+    sink_type='mysql',
+    sink_uri=mysql_sink,
+    table_mappings=[('sales', 'orders', 'sales_archive', 'orders_backup'), 'sales.customers:sales_archive.customers']
+)
+
+# Pause/resume lifecycle
+client.cdc.pause('replicate_sales')
+client.cdc.resume('replicate_sales')
+
+# Inspect status and per-table watermarks
+tasks = client.cdc.list()
+watermarks = client.cdc.list_watermarks('replicate_sales')
+
+# Drop task when no longer needed
+client.cdc.drop('replicate_sales')
+client.cdc.drop('replicate_key_tables')
+
+client.disconnect()
+```
+
+> Async usage: call the same helpers via `await async_client.cdc.create_database_task(...)` / `create_table_task(...)` or inside `async with client.session()`.
+
 ### Wrapping Existing Sessions (For Legacy Code)
 
 If you have existing SQLAlchemy code, wrap your sessions to add MatrixOne features:
@@ -486,6 +664,26 @@ mo-diag -d test -c "sql SELECT COUNT(*) FROM my_table"
 mo-diag -d test -c "flush_table my_table"
 ```
 
+#### CDC Task Shortcuts
+
+CDC management is integrated with the CLI. The interactive commands (`cdc_tasks`,
+`cdc_task`, `cdc_create`, `cdc_drop`) now have non-interactive equivalents
+under `mo-diag cdc`, making automation straightforward.
+
+```bash
+# Summaries (mirrors cdc_tasks)
+mo-diag cdc show
+
+# Inspect a task and tighten watermark thresholds
+mo-diag cdc show nightly_sync --details --threshold=5m
+
+# Start the guided creator in table mode
+mo-diag cdc create --table-level
+
+# Drop a task without interactive prompts
+mo-diag cdc drop nightly_sync --force
+```
+
 ### Available Commands
 
 #### Index Management

{matrixone_python_sdk-0.1.9 → matrixone_python_sdk-0.1.11}/examples/example_24_query_update.py

@@ -36,6 +36,7 @@ from matrixone import Client
 from matrixone.orm import declarative_base
 from matrixone.config import get_connection_params
 from sqlalchemy import Column, Integer, String, DECIMAL, TIMESTAMP, func
+from sqlalchemy.dialects.mysql import JSON
 
 # Create MatrixOne logger
 logger = logging.getLogger(__name__)
@@ -59,6 +60,7 @@ class User(Base):
     login_count = Column(Integer, default=0)
     last_login = Column(TIMESTAMP)
     created_at = Column(TIMESTAMP, server_default=func.current_timestamp())
+    preferences = Column(JSON, nullable=True)  # JSON field for user preferences
 
 
 class Product(Base):
@@ -87,7 +89,8 @@ def demo_simple_updates():
         # Create tables
         client.create_all(Base)
 
-        # Insert sample data
+        # Insert sample data with JSON fields
+        # Note: Python dict in JSON column is automatically serialized
         users_data = [
             {
                 "id": 1,
@@ -98,6 +101,7 @@ def demo_simple_updates():
                 "salary": 50000.00,
                 "status": "active",
                 "login_count": 5,
+                "preferences": {"theme": "dark", "language": "en", "notifications": True},
             },
             {
                 "id": 2,
@@ -108,6 +112,7 @@ def demo_simple_updates():
                 "salary": 60000.00,
                 "status": "active",
                 "login_count": 10,
+                "preferences": {"theme": "light", "language": "zh", "notifications": False},
             },
             {
                 "id": 3,
@@ -118,18 +123,22 @@ def demo_simple_updates():
                 "salary": 70000.00,
                 "status": "inactive",
                 "login_count": 2,
+                "preferences": {"theme": "auto", "language": "en", "email_digest": "daily"},
             },
         ]
+        # Using batch_insert with JSON dict - auto-serialization
         client.batch_insert("example_users_update", users_data)
+        print("   ✓ Inserted users with JSON preferences using batch_insert()")
 
         # 1. Simple update with key-value pairs
-        print("1. Simple update with key-value pairs")
+        print("\n1. Simple update with key-value pairs")
         query = client.query(User)
         result = query.update(full_name="Alice Updated", email="alice.updated@example.com").filter(User.id == 1).execute()
 
         # Verify the update
         updated_user = client.query(User).filter(User.id == 1).first()
         print(f"   Updated user: {updated_user.full_name}, {updated_user.email}")
+        print(f"   User preferences (JSON): {updated_user.preferences}")
 
         # 2. Update multiple records with condition
         print("\n2. Update multiple records with condition")
@@ -165,7 +174,7 @@ def demo_sqlalchemy_expressions():
         # Create tables
         client.create_all(Base)
 
-        # Insert sample data
+        # Insert sample data with JSON - Using batch_insert
         users_data = [
             {
                 "id": 1,
@@ -176,7 +185,14 @@ def demo_sqlalchemy_expressions():
                 "salary": 50000.00,
                 "status": "active",
                 "login_count": 5,
+                "preferences": {"theme": "dark", "language": "en"},
             },
+        ]
+        client.batch_insert("example_users_update", users_data)
+
+        # Insert another user using client.insert() with JSON dict
+        client.insert(
+            "example_users_update",
             {
                 "id": 2,
                 "username": "bob",
@@ -186,12 +202,13 @@ def demo_sqlalchemy_expressions():
                 "salary": 60000.00,
                 "status": "active",
                 "login_count": 10,
+                "preferences": {"theme": "light", "language": "zh", "notifications": False},
             },
-        ]
-        client.batch_insert("example_users_update", users_data)
+        )
+        print("   ✓ Inserted users using batch_insert() and insert() with JSON dicts")
 
         # 1. Update with SQLAlchemy expressions
-        print("1. Update with SQLAlchemy expressions")
+        print("\n1. Update with SQLAlchemy expressions")
         query = client.query(User)
         result = query.update(login_count=User.login_count + 1, salary=User.salary * 1.1).filter(User.id == 1).execute()
 

{matrixone_python_sdk-0.1.9 → matrixone_python_sdk-0.1.11}/examples/example_26_stage_operations.py

@@ -225,7 +225,7 @@ class StageOperationsDemo:
             self.results['files_created'].append(csv_file)
 
             # Load data from stage using client.load_data
-            result = client.load_data.from_stage_csv('demo_file_stage', 'stage_users.csv', StageUser)
+            result = client.load_data.read_csv_stage('demo_file_stage', 'stage_users.csv', StageUser)
             self.logger.info(f"✅ Loaded {result.affected_rows} rows from stage")
 
             # Verify data
@@ -392,7 +392,7 @@ class StageOperationsDemo:
             # Load data from stage within session
             with client.session() as tx:
                 # Load data from pre-existing stage in session
-                result = tx.load_data.from_stage_csv('tx_temp_stage', 'tx_data.csv', TxStageData)
+                result = tx.load_data.read_csv_stage('tx_temp_stage', 'tx_data.csv', TxStageData)
                 # Check for both ResultSet and SQLAlchemy Result
                 affected = result.affected_rows if hasattr(result, 'affected_rows') else result.rowcount
                 self.logger.info(f"✅ Loaded {affected} rows in session")
@@ -706,7 +706,7 @@ class StageOperationsDemo:
 
             # Load customers
             print("Loading customers dimension...")
-            stage.load_csv('customers.csv', Customer, ignore_lines=1)
+            stage.load_csv('customers.csv', Customer, skiprows=1)
             count = client.query(Customer).count()
             print(f"✅ Loaded {count} customers")
 
@@ -718,13 +718,13 @@ class StageOperationsDemo:
 
             # Load orders
             print("Loading orders fact table...")
-            stage.load_csv('orders.csv', Order, ignore_lines=1)
+            stage.load_csv('orders.csv', Order, skiprows=1)
             count = client.query(Order).count()
             print(f"✅ Loaded {count} orders")
 
             # Load shipping (TSV)
             print("Loading shipping logistics...")
-            stage.load_tsv('shipping.tsv', Shipping, ignore_lines=1)
+            stage.load_tsv('shipping.tsv', Shipping, skiprows=1)
             count = client.query(Shipping).count()
             print(f"✅ Loaded {count} shipping records")
 
@@ -794,7 +794,7 @@ class StageOperationsDemo:
             self.results['files_created'].append(delta_file)
 
             print("📥 Loading incremental orders...")
-            stage.load_csv('orders_delta.csv', Order, ignore_lines=1)
+            stage.load_csv('orders_delta.csv', Order, skiprows=1)
             count = client.query(Order).count()
             print(f"✅ Total orders now: {count} (+2 new)")