Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/jbms/finance-dl/llms.txt

Use this file to discover all available pages before exploring further.

finance-dl’s interactive mode drops you into a live IPython shell with a fully initialised scraper instance, a real browser window, and direct access to Selenium helpers — all wired together in one command. This page explains how to launch interactive mode, what variables are available in the shell, and common patterns for debugging login flows and multi-factor authentication.

​
Launching Interactive Mode

Add the --interactive (or -i) flag to any finance_dl.cli invocation: 
python -m finance_dl.cli \
  --config-module finance_dl_config \
  --config mybank \
  --interactive
Interactive mode always opens a visible browser window. The --interactive flag implies --visible — you do not need to pass both.

​
What’s in the Shell

When the shell opens, the namespace is populated with the module’s own globals plus a set of context-specific variables injected by the scraper module’s interactive() context manager.

​
For Selenium-based scrapers

Most scrapers use scrape_lib.interact_with_scraper, which yields:
VariableTypeDescription
scraperscrape_lib.ScraperThe active scraper instance (same object as self)
selfscrape_lib.ScraperAlias for scraper, matching the scraper’s own method signatures
Byselenium.webdriver.common.by.ByLocator strategy constants (By.XPATH, By.ID, etc.)
Selectselenium.webdriver.support.ui.SelectHelper for interacting with <select> elements
Keysselenium.webdriver.common.keys.KeysKeyboard key constants (Keys.RETURN, Keys.TAB, etc.)
Access the raw WebDriver through self.driver to call any Selenium API directly.

​
For OFX-based scrapers

The finance_dl.ofx module’s interactive() context yields:
VariableTypeDescription
ofx_paramsdictThe OFX connection parameters from your config
output_directorystrThe output path from your config
instofxclient.InstitutionAn already-connected Institution object ready for queries

​
Module-level globals

All functions and variables defined at the top level of the scraper module are also available in the shell namespace, so you can call internal helpers directly.
The shell is launched with %load_ext autoreload and %autoreload 2 already active. Any changes you make to the scraper source file on disk are reloaded automatically before the next statement executes — no need to restart the shell to pick up edits.

​
Common Interactive Patterns

​
Run the full scraper

For most scrapers, calling self.run() executes the complete scraping sequence from inside the shell: 
self.run()

# Go to a specific URL
self.driver.get('https://www.mybank.com/login')

# Get the current page title
self.driver.title

​
Find page elements

# Find all password inputs on the page
self.driver.find_elements(By.XPATH, '//input[@type="password"]')

# Find a button by visible text
self.driver.find_elements(By.XPATH, '//button[contains(text(), "Sign In")]')

# Find an element by ID
self.driver.find_element(By.ID, 'username')

​
Interact with elements

# Type into an input field
field = self.driver.find_element(By.ID, 'username')
field.send_keys('myusername')
field.send_keys(Keys.TAB)

# Click a button
self.driver.find_element(By.ID, 'submit-btn').click()

# Select a dropdown option
from selenium.webdriver.support.ui import Select
dropdown = Select(self.driver.find_element(By.ID, 'account-type'))
dropdown.select_by_visible_text('Checking')

​
Handle MFA manually

If a scraper stops at an MFA prompt, complete the challenge directly in the visible browser window and then continue the scraping session from the shell: 
# Complete the MFA prompt manually in the browser, then:
self.run()
For scrapers that support a profile_dir parameter, set it in your config to a persistent directory. Chrome will save cookies and session data there, so completed MFA sessions survive restarts. On subsequent runs the scraper can resume without re-authenticating:
# In your CONFIG_ function
def CONFIG_mybank():
    return {
        'module': 'finance_dl.mybank',
        'profile_dir': '/home/user/.config/finance-dl/profiles/mybank',
        # ...
    }

​
Reload Edited Code Without Restarting

Because %autoreload 2 is active, you can edit a scraper module in your editor and call it again in the same shell session without losing your browser state: 
# Edit finance_dl/mybank.py in your editor, then back in the shell:
self.run()   # picks up your changes automatically
This makes interactive mode the fastest way to iterate on scraper code — no need to restart Chrome, log in again, or navigate back to the page you were debugging.

Build docs developers (and LLMs) love

Get started for freeTalk to us