Error handling example, eg auto-screenshot of webpage if element is missing [done]

[marcelocecin]

hi Ken
as suggested here: aisingapore/TagUI#939
I also suggest creating this function in this repository.
suggestion to change tagui.py

import requests
from requests_toolbelt.multipart.encoder import MultipartEncoder

def rpa_error_cannot_find(error):
    snap('page', error+'_error.png')
    url = 'https://webhook.site/53388dfe-1308-45ae-8c88-c31b89c10cbb'
    m = MultipartEncoder( fields={'error': '[RPA][ERROR] - cannot find ' + error, 'image': (error+'_error.png', open(error+'_error.png', 'rb'), 'text/plain')} )
    r = requests.post(url, data=m, headers={'Content-Type': m.content_type})
    print('[RPA][ERROR] - cannot find ' + error)

substitute in all functions that used the print('[RPA][ERROR] - cannot find ' + element_identifier) function for rpa_error_cannot_find(element_identifier), example:

def click(element_identifier = None, test_coordinate = None):
    if not _started():
        print('[RPA][ERROR] - use init() before using click()')
        return False

    if element_identifier is None or element_identifier == '':
        print('[RPA][ERROR] - target missing for click()')
        return False

    if test_coordinate is not None and isinstance(test_coordinate, int):
        element_identifier = coord(element_identifier, test_coordinate)

    if not exist(element_identifier):
        rpa_error_cannot_find(element_identifier)
        return False

    elif not send('click ' + _sdq(element_identifier)):
        return False

    else:
        return True

example sample.py

import tagui as r
r.init(visual_automation = True)
r.url('https://ca.yahoo.com')
r.click('webhook_test')
r.type('ybar-sbq', 'github')
search_text = r.read('ybar-sbq')
print(search_text)
r.click('ybar-search')
r.wait(6.6)
r.snap('page', 'results.png')
r.snap('logo', 'logo.png')
r.url('https://duckduckgo.com')
r.type('search_form_input_homepage', 'The search engine that doesn\'t track you.')
r.snap('page', 'duckduckgo.png')
r.wait(4.4)
r.close()

results console:

[RPA][ERROR] - cannot find webhook_test
github

result received at https://webhook.site

thanks !

[kensoh]

Hi @marcelocecin thanks for sharing your solution! I've added a link from readme to your implementation - https://github.com/tebelorg/RPA-Python#api-reference

I can't merge to master because different users will have different ways to handle the error, to a different API endpoint for eg. However, if having this style of error handling is preferred by many users, then I could explore adding with a dummy function or let the function read from a predefined error-handler Python script.

Below are some possible iterations to the version that you are using. For eg, taking snapshot of webpage only or the entire screen when visual automation is used. Also, replacing spaces in the error input parameter may be a good practice to deal with spaces in the given identifier because filenames with spaces are not ideal on Linux / macOS.

if not _visual():
     snap('page', error+'_error.png')
else:
     snap('page.png', error+'_error.png')