musings | 2012-06-02 - Interposing dynamic library

I have a written a large data processing and plotting application using PyQt5 with Matplotlib and I am plagued by old plots reappearing when creating new ones. I have seen many questions answered regarding this but none of the suggested methods work for me.

I have reduced the problem down to the following simplified code:

In below code:

Suppose that library includes arbitrarily large number of functions; one function per file:

• doSomethingAlpha.ts
• doSomethingBravo.ts
• doSomethingCharlie.ts
• ...

Some of functions are refers to global __LANGUAGE_CODE__:

... Or how to change $.Str value from token sigil { ... } idependently from the matched text. Yes I'm asking how to cheat grammars above (i.e. calling) me.

I am trying to write a Slang for Raku without sigil.

So I want the nogil token, matching anything to return NqpMatch that stringifies: $.Str to '$'.

Currently, my token sigil look like that

Google's security guidelines for Android app developers has the following:

WebViews do not use addJavaScriptInterface() with untrusted content.

On Android M and above, HTML message channels can be used instead.

Near as I can tell, "HTML message channels" refers to things like createWebMessageChannel(), WebMessagePort, WebMessage, and kin.

However, they do not provide any examples. All they do is link to a WhatWG specification, which is rather unclear. And, based on a Google search for createWebMessageChannel, it appears that this has not been used much yet — my blog post describing changes in the Android 6.0 SDK makes the top 10 search results, and I just mention it in passing.

addJavascriptInterface() is used to allow JavaScript in a WebView to call into Java code supplied by the app using the WebView. How would we use "HTML message channels" as a replacement for that?

I have a set of microservices implemented via Spring Boot that communicate via HTTP REST. I am attempting to set up integration tests using the Citrus Framework. I have successfully written and executed integration tests for some of the services, but am struggling with a more complicated use case.

The scenario is as follows:

I have a notional "mission planning" application that involves 3 services: a mission planner, route generator and a route assessor. The 2 route services are called by the mission planner to do some of the work necessary to plan a mission. The route services make no REST calls.

I have set up my maven POM file to run the integration test when the Maven verify step is performed. The failsafe and Spring Boot maven plugins are configured to do this.

The integration tests for the route services work as expected, but the test for the mission planner does not. I suspect one possible cause is that the route services have not been started. So, after a little research, I inserted calls in the integration test logic to start the 2 services before doing the test work, as well as stopping them afterwards.

I'm not 100% sure whether I have taken the correct steps to start the services properly. I'm also curious whether there is a way to know whether they have actually started successfully. I can't find any documentation that addresses this.

In any case, either they are not being started, or there is something wrong with their configuration, or there is something else causing issues. When I run the test, it fails with an indication that the mission planner is unable to connect to the route generator (which is called first).

Here are the networking details:

• Everything is currently running on localhost.
• There is NO SSL/TLS involved currently.
• The route generator will be listening on port 9110.
• The route assessor will be listening on port 9120.
• The mission planner is configured to access the route services using the above ports.

My workplace does use a HTTP proxy - would this affect localhost communications?

Mission planner

EndpointConfig.java:

Create the checkbox as Vue component, herewith:

1. No logic inside checkbox component allowed: all event handlers and also checked property are fully depends on external logic, which could be the vuex store.
2. We should not watch the checkbox "checked" state: checked it or not, it depends on, again, external logic, e. g. vuex state or getter.

The checkbox component has checked and onClick properties, which value are off course, could be dynamic.

Template in Pug language:

Following on a great system for using an enum-like replacement for Django choices (http://musings.tinbrain.net/blog/2017/may/15/alternative-enum-choices/) I have a project that uses a class with a custom metaclass that allows me to do list(MyChoices) (on the Class itself) to get a list of all the enum choices. The relevant part of the code looks something like this:

I followed this tutorial and made a typo where I was supposed to create a user for my django apps to connect as;

I was supposed to run su - postgres -c "createuser www-data -P" but I ran su - postgres -c "createuser www-dtata -P".

I dont want to proceed until I remove that user, which I don't know the command for. I found and tried DROP USER after searching around, but the terminal returned -su: DROP: command not found.