View Code? Open in Web Editor
NEW
Backend for the HelperLine.
License: MIT License
backend's People
Contributors
Watchers
backend's Issues
Simple backend fix inside get_all_helper_data
.
This is required from our frontend but not using the plain API.
Add method for dequeuing calls from these queues according to the filters.
Possibility to store a comment with each call (in call record itself).
The helper can make notes about a call (no formatting just plan text).
When call is rejected that comment is removed.
Add all database fields - without scheduling.
Test out call forwarding with Twilio.
Implement the actual forwarding.
Add REST-endpoint to toggle forwarding.
Make subdomain backend.helperline.io
point at backend
Try it first on other domain.
Modify flack_backend/__init__.py
for easy switch between Testing and not Testing (On GAE as well as locally).
A dedicated database for testing.
Dedicated environment variables.
Conceptualize list of features to test.
Change routes according to the new standard v1.
Do not deploy yet - or if so just with --no-promote
Commits associated with this issue are accidentily mixed up with issue #41 .
You can simply export the full API documentation as one static html file. This can be a template for the flask backend at the route /v1/docs
Add version numbers to api:
.com/backend/... -> .com/v1/backend/...
This makes it easier to further develop the backend separately from the backend.
The Release CW 15 will be the API version v1.
Whenever trying to save a call comment (onBlur of the comment text field) it says "invalid session token".
Add condition to edit-call-functions or block hotline requests right away.
Currently we are host the whole webapp and the hotline on Heroku.
However until now Heroku does not offer any Covid-19 related funding, whereas GCP does.
We could easily switch to the GCP AppEngine. Deploying and scaling seems to be super easy and fast.
Better status responses about what went wrong for all API endpoints.
From {"local": True/False} to {"call_type": "local"/"global"}.
Currently with most requests, the result gets send back to the client in the form of the full account/calls data. The frontend state changes only a little bit (e.g. fulfilling a call).
A full fetch after every operation is totally unnecessary!
Goal: Further reduced server load by reducing the amount fetching data that the client already has.
Create Test environment: Separate Environment variables and separate testing database.
Add test scripts to backend, testing all API endpoints of the backend.
Unit tests for enqueuing and dequeuing already exists but they are not fully up to date!
Add every - for now - incoming call to the local/global call queue.
Actually all three queues (see details tab) are in one single collection with clever querying.
Add all database fields.
Test out the Twilio SMS API.
Implement the actual SMS verification trigger.
Add REST-endpoint to enter phone number/resend verification.
Replace the current survey with the actual hotline
Receive calls and store/update the caller profile as well as the call record in our database.
For now pure html must do it.
Future: Twilio offers templates for emails.
Currently most directories have their own file support_functions.py
.
This is very confusing because these support functions do not get used exclusively by this directory but also across directories.
Goal of this refactoring: Support Functions into separate directory and well sorted with clean imports.
Online = Calls get forwarded. Going online is only possible if the number is confirmed!
Add one field (e.g. "service") to api key documents.
Really simple fix inside rest_call.py
.