Optional Components¶

Starting with version 0.8.0, Pytanis has been refactored to make external dependencies optional. This allows you to install only the components you need, reducing the overall dependency footprint.

Installation¶

Core Installation¶

The core installation includes only the essential dependencies:

pip install pytanis

This gives you: - Pretalx client for conference management - Local file storage (CSV/Excel) - Core utilities and configuration

Optional Components¶

Install additional components as needed:

# Google Sheets support
pip install pytanis[google]

# HelpDesk support
pip install pytanis[helpdesk]

# Mailgun support
pip install pytanis[mailgun]

# Jupyter notebooks and visualization
pip install pytanis[jupyter]

# Schedule optimization
pip install pytanis[optimization]

# Everything
pip install pytanis[all]

Storage Abstraction¶

Pytanis now provides a storage abstraction layer that allows you to choose between different storage backends.

Configuration¶

Configure your storage provider in config.toml:

[Storage]
provider = "local"  # or "google"
local_path = "./data"  # for local storage

[Google]  # Only needed if using Google storage
client_secret_json = "client_secret.json"
token_json = "token.json"

Using the Storage API¶

from pytanis import get_storage_client
import pandas as pd

# Get storage client based on configuration
storage = get_storage_client()

# Create a DataFrame
df = pd.DataFrame({'name': ['Alice', 'Bob'], 'score': [95, 87]})

# Write to storage (works with both local and Google)
storage.write_sheet('my_data', df, sheet_name='results')

# Read from storage
df_read = storage.read_sheet('my_data', 'results')

# List available sheets
sheets = storage.list_sheets('my_data')

Local Storage¶

When using local storage: - Files are stored as Excel (.xlsx) or CSV files - Spreadsheet IDs map to filenames - Multiple sheets are supported for Excel files

Google Sheets Storage¶

When using Google Sheets: - Requires pytanis[google] installation - Uses the same API as local storage - Spreadsheet IDs are Google Sheets IDs

Communication Abstraction¶

Pytanis provides abstraction for email and ticket systems.

Configuration¶

[Communication]
email_provider = "mailgun"  # or "helpdesk"
ticket_provider = "helpdesk"

[Mailgun]  # If using Mailgun
token = "your-token"
from_address = "noreply@example.com"

[HelpDesk]  # If using HelpDesk
account = "your-account"
entity_id = "your-entity"
token = "your-token"

Using the Communication API¶

from pytanis import get_mail_client
from pytanis.communication import EmailMessage

# Get mail client based on configuration
mail_client = get_mail_client()

# Create and send an email
message = EmailMessage(
    to=['recipient@example.com'],
    subject='Conference Update',
    body='Your talk has been accepted!',
    html_body='<p>Your talk has been accepted!</p>'
)

message_id = mail_client.send_email(message)

Backward Compatibility¶

The original APIs (GSheetsClient, HelpDeskClient) are still available and work as before. They are now lazy-loaded, meaning dependencies are only required when you actually use them.

# This still works but requires pytanis[google]
from pytanis import GSheetsClient
client = GSheetsClient()

# This still works but requires pytanis[helpdesk]
from pytanis import HelpDeskClient
client = HelpDeskClient()

Migration Guide¶

For Existing Projects¶

1. Update your installation:

   # If you use all features
   pip install pytanis[all]
   
   # Or install only what you need
   pip install pytanis[google,mailgun]
   

2. Update your configuration (optional): Add the new sections to use the abstraction layer:

   [Storage]
   provider = "google"  # Keep using Google Sheets
   
   [Communication]
   email_provider = "mailgun"  # Keep using Mailgun
   

3. Update your code (optional): You can continue using the old APIs or migrate to the new abstraction:

   # Old way (still works)
   from pytanis import GSheetsClient
   client = GSheetsClient()
   
   # New way (recommended)
   from pytanis import get_storage_client
   storage = get_storage_client()
   

For New Projects¶

1. Install only what you need
2. Use the new abstraction APIs
3. Configure providers in config.toml

Benefits¶

• Lighter installations: Only install dependencies you actually use
• Flexibility: Easy to switch between providers
• Testing: Use local storage for tests, Google Sheets for production
• Future-proof: Easy to add new providers without changing your code