tommoa / spa Goto Github PK

For some reason, AppInspect does not seem to be installing properly for regression testing.

Tasks:

• Investigate possible reasons why this would happen
• Fix whatever issues cause this

At the moment, we don't restart the Splunk instance after CI is run.

This means that any changes made to app do not get reflected in the runtime until Splunk is manually restarted. We should automatically restart Splunk after CI completes.

Updating unit tests to work after the code refactor.
Adding GUI tests with Selenium WebDriver.

Added a page to allow the user to add websites and view the list of websites on the blacklist.

It also monitors the SPA/blacklists folder for CSV files to add to the blacklist index. A button on the blacklist page will push the new websites into the KV store to allow for lookups.

@khuang247 and I just need to sit down and clean this whole codebase up, the javascript library Splunk uses is Require.js - and the syntax needs to be standardised across the whole app. This should only take a few hours and should be done in the final weeks before documentation is written up. I'm not a fan of Require.js and there seems to be alot of inconsistences (syntactically) across the application.

We actually already have the key set. There are other GitHub repositories (see this one) have finangled figuring out how to push to a remote hsot (even a different git branch).

As of AppInspect 2.0.0 (released 11 Oct, 2019), Python 3 is now supported as a language for AppInspect. As such, we can now upgraade the version of Python used for integration tests.

We need to have an icon to make AppInspect pass. The required sizes are:

• 72px*72px (/static/appIcon_2x.png)
• 36px*36px (/static/appIcon.png)

In order to be more efficient we can simply cache the JWT on the frontend - it's not the end of the world if this doesn't happen, I don't think it's unreasonable to assume only a few people will be accessing TIM in the UWA context. However, it is still inefficient - just a bit of a pain to implement.

It'd be good to have a consistent and nice style to both our commits and our code.

The majority of our code will be in JavaScript (for the frontend app) and Python (possibly for the backend). We should make sure that for both of these languages, we are consistent with both Splunk's coding standards, but also our own.

• Commit style guide
• JavaScript style guide
• Python style guide

Experiment and develop a POC to demonstrate the ability to integrate a google maps screen into the dashboard. Will just need to get a Google Maps key and figure out how to integrate maps into Splunk.

Need to add in view which essentially just lists all the alerts

AppInspect currently fails when trying to verify the application - likely because of noise in the directory.

• Clean up the directory to not include non-needed files

A CSV can be exported from Callista, which contains the name and contact information of students. This can be integrated into our profiles.

AppInspect error:

Check that when decompressed the Splunk App directory name matches the 
    app.conf [package] stanza's id property. For cloud, id property must exists; 
    For others, it's ok if no id property if 'check_for_updates = True' or no 
    'check_for_updates'. 
        FAILURE: The `app.conf` [package] stanza does not exist. Please 
            disable `check_for_updates` or set the `id` property in the 
            [package] stanza. File: default/app.conf 

Proof of concept to demonstrate the ability for SPA to communicate with TIM via REST API.

Need to add in View which visualises the data returned by the /statistics endpoint from TIM

Currently, I've set TIM up with a single endpoint and some static data. The next step will be integration between "real time" alerts that TIM creates and the SPA displaying these alerts. This will require a number of steps

• creating the UI page to display alerts
• creating the JS "polling" code to retrieve a set amount of alerts
• creating the UI component which will expand to display the relevant information for each alert

Need to find an appropriate library that can do testing. I don't think actually testing the code will help us since it's so tightly coupled with the UI and the Splunk database, but maybe a library which can test UI interactions? Need to research this.

Added in a /login request to retrieve a valid JWT before accessing API. This is inefficient - but it's a solution that will do for now. Additionally, the username and password for accessing this API will be exposed on the front-end - not sure what the best way to deal with this is. @Tommoa I think it may be possible to use environment variables since we're not building the app and then deploying it.

The Modal takes a while to "kick in" e.g. the contact details take a while to appear - it would be nice to have a "loading bar" to show that content will be delievered soon

The current SPA interface is bland and uninteresting. There is a need for a simple, clean and attractive design to be implemented with CSS. CSS files can be found under appserver/static/css. Some standard for font size, weights, colors e.t.c will also need to be created.