Implemented Task B18: CSV preview endpoint

backend/api/chat.py

@@ -1,7 +1,8 @@
+import logging
 import random
 import uuid
 from datetime import datetime
-from typing import Any, Dict, List
+from typing import Any, Dict, List, Optional
 
 from fastapi import APIRouter, Depends, HTTPException, Query
 
@@ -21,6 +22,7 @@
 
 router = APIRouter(prefix="/chat", tags=["chat"])
 project_service = get_project_service()
+logger = logging.getLogger(__name__)
 
 # Mock chat messages database
 MOCK_CHAT_MESSAGES = {}
@@ -378,10 +380,97 @@ async def get_csv_preview(
         if not project_obj:
             raise HTTPException(status_code=404, detail="Project not found")
 
-        # Generate preview from project metadata
-        if not project_obj.columns_metadata:
+        # Check if CSV file exists
+        if not project_obj.csv_path:
+            raise HTTPException(status_code=404, detail="CSV preview not available")
+
+        # Load actual CSV data from storage
+        preview = _load_csv_preview_from_storage(project_obj)
+
+        if not preview:
+            # Fallback to metadata-based preview if file loading fails
+            preview = _generate_preview_from_metadata(project_obj)
+
+        if not preview:
             raise HTTPException(status_code=404, detail="CSV preview not available")
 
+        return ApiResponse(success=True, data=preview)
+
+    except HTTPException:
+        # Re-raise HTTPExceptions (like 404) as-is
+        raise
+    except Exception as e:
+        raise HTTPException(status_code=500, detail=f"Error loading CSV preview: {str(e)}")
+
+
+def _load_csv_preview_from_storage(project_obj) -> Optional[CSVPreview]:
+    """Load CSV preview from actual file in storage"""
+    try:
+        from services.storage_service import storage_service
+        import pandas as pd
+        import io
+
+        # Download CSV file from storage
+        csv_bytes = storage_service.download_file(project_obj.csv_path)
+        if not csv_bytes:
+            return None
+
+        # Read CSV into pandas DataFrame
+        csv_buffer = io.BytesIO(csv_bytes)
+        df = pd.read_csv(csv_buffer)
+
+        # Get first 5 rows for preview
+        preview_df = df.head(5)
+
+        # Extract column information
+        columns = list(df.columns)
+        sample_data = preview_df.values.tolist()
+        total_rows = len(df)
+
+        # Determine data types
+        data_types = {}
+        for col in columns:
+            dtype = str(df[col].dtype)
+            if 'int' in dtype or 'float' in dtype:
+                data_types[col] = 'number'
+            elif 'datetime' in dtype or 'date' in dtype:
+                data_types[col] = 'date'
+            elif 'bool' in dtype:
+                data_types[col] = 'boolean'
+            else:
+                data_types[col] = 'string'
+
+        # Convert any non-serializable values to strings
+        serializable_sample_data = []
+        for row in sample_data:
+            serializable_row = []
+            for value in row:
+                if pd.isna(value):
+                    serializable_row.append(None)
+                elif isinstance(value, (pd.Timestamp, pd.Timedelta)):
+                    serializable_row.append(str(value))
+                else:
+                    serializable_row.append(value)
+            serializable_sample_data.append(serializable_row)
+
+        return CSVPreview(
+            columns=columns,
+            sample_data=serializable_sample_data,
+            total_rows=total_rows,
+            data_types=data_types
+        )
+
+    except Exception as e:
+        logger.error(f"Error loading CSV preview from storage: {str(e)}")
+        return None
+
+
+def _generate_preview_from_metadata(project_obj) -> Optional[CSVPreview]:
+    """Generate preview from project metadata as fallback"""
+    try:
+        if not project_obj.columns_metadata:
+            return None
+
         # Extract column names and types
         columns = [col.get('name', '') for col in project_obj.columns_metadata]
         data_types = {col.get('name', ''): col.get('type', 'unknown') for col in project_obj.columns_metadata}
@@ -405,20 +494,16 @@ async def get_csv_preview(
                         row.append(f"Sample {i+1}")
             sample_data.append(row)
 
-        preview = CSVPreview(
+        return CSVPreview(
             columns=columns,
             sample_data=sample_data,
             total_rows=project_obj.row_count or 0,
             data_types=data_types
         )
 
-        return ApiResponse(success=True, data=preview)
-
-    except HTTPException:
-        # Re-raise HTTPExceptions (like 404) as-is
-        raise
     except Exception as e:
-        raise HTTPException(status_code=500, detail=f"Error loading CSV preview: {str(e)}")
+        logger.error(f"Error generating preview from metadata: {str(e)}")
+        return None
 
 
 @router.get("/{project_id}/suggestions")

backend/test_csv_preview.py

@@ -0,0 +1,126 @@
+#!/usr/bin/env python3
+"""
+Test script for CSV preview endpoint - Task B18
+"""
+
+import io
+import pandas as pd
+from unittest.mock import Mock, patch
+from api.chat import _load_csv_preview_from_storage, _generate_preview_from_metadata
+
+def test_load_csv_preview_from_storage():
+    """Test loading CSV preview from storage"""
+
+    # Create sample CSV data
+    sample_csv = """name,age,city,salary
+Alice,25,New York,75000
+Bob,30,Los Angeles,85000
+Charlie,35,Chicago,90000
+Diana,28,Houston,80000
+Eve,32,Phoenix,77000"""
+
+    # Mock project object
+    mock_project = Mock()
+    mock_project.csv_path = "test/sample.csv"
+
+    # Mock storage service
+    with patch('api.chat.storage_service') as mock_storage:
+        mock_storage.download_file.return_value = sample_csv.encode('utf-8')
+
+        # Test the function
+        result = _load_csv_preview_from_storage(mock_project)
+
+        # Verify results
+        assert result is not None
+        assert result.columns == ["name", "age", "city", "salary"]
+        assert len(result.sample_data) == 5  # Should have 5 rows
+        assert result.total_rows == 5
+        assert result.data_types["name"] == "string"
+        assert result.data_types["age"] == "number"
+        assert result.data_types["salary"] == "number"
+
+        # Check sample data
+        assert result.sample_data[0] == ["Alice", 25, "New York", 75000]
+        assert result.sample_data[1] == ["Bob", 30, "Los Angeles", 85000]
+
+        print("✅ CSV preview from storage test passed!")
+
+def test_generate_preview_from_metadata():
+    """Test generating preview from metadata"""
+
+    # Mock project object with metadata
+    mock_project = Mock()
+    mock_project.row_count = 100
+    mock_project.columns_metadata = [
+        {
+            "name": "product_name",
+            "type": "string",
+            "sample_values": ["Product A", "Product B", "Product C"]
+        },
+        {
+            "name": "sales_amount", 
+            "type": "number",
+            "sample_values": [1500.0, 2300.5, 1890.25]
+        },
+        {
+            "name": "date",
+            "type": "date",
+            "sample_values": ["2024-01-01", "2024-01-02", "2024-01-03"]
+        }
+    ]
+
+    # Test the function
+    result = _generate_preview_from_metadata(mock_project)
+
+    # Verify results
+    assert result is not None
+    assert result.columns == ["product_name", "sales_amount", "date"]
+    assert len(result.sample_data) == 5  # Should have 5 rows
+    assert result.total_rows == 100
+    assert result.data_types["product_name"] == "string"
+    assert result.data_types["sales_amount"] == "number"
+    assert result.data_types["date"] == "date"
+
+    # Check sample data uses actual sample values
+    assert result.sample_data[0] == ["Product A", 1500.0, "2024-01-01"]
+    assert result.sample_data[1] == ["Product B", 2300.5, "2024-01-02"]
+    assert result.sample_data[2] == ["Product C", 1890.25, "2024-01-03"]
+
+    print("✅ CSV preview from metadata test passed!")
+
+def test_csv_data_types_detection():
+    """Test data type detection for different CSV column types"""
+
+    # Create CSV with various data types
+    sample_csv = """id,name,active,price,created_date,rating
+1,Product A,true,19.99,2024-01-01,4.5
+2,Product B,false,29.99,2024-01-02,3.8
+3,Product C,true,39.99,2024-01-03,4.2"""
+
+    mock_project = Mock()
+    mock_project.csv_path = "test/types.csv"
+
+    with patch('api.chat.storage_service') as mock_storage:
+        mock_storage.download_file.return_value = sample_csv.encode('utf-8')
+
+        result = _load_csv_preview_from_storage(mock_project)
+
+        assert result is not None
+        assert result.data_types["id"] == "number"
+        assert result.data_types["name"] == "string"
+        assert result.data_types["active"] == "boolean"
+        assert result.data_types["price"] == "number"
+        assert result.data_types["rating"] == "number"
+
+        print("✅ Data type detection test passed!")
+
+if __name__ == "__main__":
+    print("Testing CSV Preview Endpoint - Task B18")
+    print("=" * 50)
+
+    test_load_csv_preview_from_storage()
+    test_generate_preview_from_metadata() 
+    test_csv_data_types_detection()
+
+    print("\n🎉 All CSV preview tests passed!")
+    print("Task B18 implementation verified!")