API Documentation

This guide explains how to use the API for validating, analyzing, and submitting CSV files along with metadata. You can interact with the API in two ways:

1. Direct HTTP requests using tools like curl.

2. Drag-and-drop via the provided Streamlit-based user interface.

Usage Options

You can use the API endpoints directly or through a user-friendly Streamlit interface.

Drag-and-Drop Interface

The Streamlit-based interface provides an intuitive way to interact with the API:

1. Navigate to the UI (if hosted publicly or locally).

2. Use the drag-and-drop functionality for: - CSV validation: Upload a file and see validation results. - CSV analysis: Upload a file to get details about its structure. - Dataset submission: Upload both a CSV and metadata file.

Example Workflow in the Interface:

1. Open the app in your browser.

2. Select or upload your CSV file.

3. Review the results for validation, analysis, or submission.

API Endpoints

For more control or automated interaction, you can directly use the API endpoints:

### 1. Validate a CSV File URL: /validate-csv/

Method: POST

Purpose: Validates the structure of the provided CSV file.

Request:

• Headers: Content-Type: multipart/form-data

• Form Data: - file: A .csv file.

Response:

• Success (200):

  {
     "detail": "CSV validation succeeded.",
     "message": "File is valid"
  }
  

• Failure (400):

  {
     "detail": "CSV validation failed: Inconsistent row length at row 3"
  }
  

Example:

curl -X POST "http://127.0.0.1:8000/validate-csv/" -F "file=@correct_csv.csv"

—

### 2. Analyze a CSV File URL: /analyze-csv/

Method: POST

Purpose: Analyzes the structure of a CSV file and provides inferred data types for each column.

Request:

• Headers: Content-Type: multipart/form-data

• Form Data: - file: A .csv file.

Response:

• Success (200):

  {
     "detail": "CSV file is valid",
     "columns": [
       {"column_name": "id", "data_type": "int64"},
       {"column_name": "name", "data_type": "object"}
     ]
  }
  

• Failure (400):

  {
     "detail": "Inconsistent row length at row 3"
  }
  

Example:

curl -X POST "http://127.0.0.1:8000/analyze-csv/" -F "file=@correct_csv.csv"

—

### 3. Submit a Dataset URL: /submit-dataset/

Method: POST

Purpose: Validates and submits a dataset (CSV file) along with its metadata (JSON file).

Request:

• Headers: Content-Type: multipart/form-data

• Form Data: - csv_file: The CSV file. - metadata_file: The metadata file in JSON format.

Response:

• Success (200):

  {
     "detail": "Dataset and metadata validation succeeded.",
     "csv_columns": ["id", "name"],
     "metadata_columns": ["id", "name"]
  }
  

• Failure (400):

  {
     "detail": "Metadata validation failed: Missing 'columns' field in metadata."
  }
  

Example:

curl -X POST "http://127.0.0.1:8000/submit-dataset/" \
-F "csv_file=@correct_csv.csv" \
-F "metadata_file=@correct_csv_metadata.json"

—

Drag-and-Drop Example in Streamlit

When using the Streamlit-based UI, the following actions are available:

### Validating a CSV File

1. Select or upload a CSV file.

2. View the validation result, which shows whether the file is valid or contains structural issues.

### Analyzing a CSV File

1. Drag-and-drop a CSV file into the interface.

2. Review the column names and their inferred data types.

### Submitting a Dataset

1. Upload both a CSV and its metadata JSON file.

2. Receive confirmation about successful submission or errors in the files.

Error Handling

For all API endpoints, the following error codes are used:

• 400 Bad Request: For invalid input or failed validation. The detail field in the response explains the issue.

• 500 Internal Server Error: For unexpected errors during processing. Contact support with the request details if this occurs.