Great idea, I released an update which adds the keyboard shortcut Ctrl+Shift+U to parse problems/contests, it should be available in a few minutes. Whenever the green plus icon is visible, that shortcut will work too.

If anybody runs into similar problem, it was somehow because my Intellij ended up with "Build project automatically" unchecked. I followed steps here and now everything works great, thanks for the tremendous plugin!

https://zeroturnaround.com/software/jrebel/quickstart/intellij/enable-automatic-compilation-in-intellij-idea/

Unfortunately, LeetCode cannot be added. This is because LeetCode forces solutions to be written in a pre-defined method with a pre-defined signature (method parameters and return type), which is something that Competitive Companion does not support. The chances of it being implemented in the future is pretty low, considering how none of the publically available tools that Competitive Companion sends data to are capable of handling such constraints at the moment.

Besides that, LeetCode doesn't seem to have any structure in their problem descriptions, which makes it pretty much impossible to parse something like the sample testcases.

I have just released version 2.5.0 which adds a problem parser for Virtual Judge. Your browser should auto-update within 24 hours.

Unfortunately, I could not find a way to parse all problems in a contests at once due to how Virtual Judge renders contest problems dynamically. You can however parse all problems in a contest separately, just like how you would parse out-of-contest problems.

It's also worth noting that while the problem parser can parse the sample testcases for a lot of problems, it will not work correctly for all of them. This is because Virtual Judge does not have a standarized format for problem descriptions, so the extension has to look for signs of sample testcases like <pre> and <span style="font-family: some-monospace-font;"> tags. This isn't a rock-solid solution, but it works for the majority of the problems. Feel free to create an issue on GitHub when you find a problem that isn't parsed correctly.

Thanks for the suggestion, but I'm not planning on changing this. I think "AThreatreSquare" is better than "1A" when you're participating in a contest with multiple problems. Especially when having to differentiate between different problems having the full name instead of just the letter can be very helpful.

I have just fixed the USACO Training issue you mentioned. I'll look into the UVa issue later this week because UVa seems to be down right now, and I'll publish a new release once that's done.

That's odd, I released 2.17.2 about a week ago to make the extension work with the Facebook Coding Competitions platform with the upcoming Qualification Round you linked. When I released it everything seemed to work fine on archived problems, so either the live problems are displayed somewhat differently or the platform as a whole got some last-minute updates in preparation of the Qualification Round.

Either way, I am on vacation right now so I'll have a look at it at the beginning of next week.

You're right, I forgot to change that when I switched it from fileName to pattern (I removed the regex type 1.5 years ago). This is fixed in 4904656 which will be released in the next few days.

Competitive Companion only parses problems and sends the extracted data to external tools like CHelper, JHelper, Hightail, acmX and others. If you don't have a tool that listens for Competitive Companion data running, then clicking on the + button does nothing.

Because it's not clear from your comment whether you are already using a tool which listens for Competitive Companion data:
- If you are using such a tool and nothing happens when you press on the + button, send a bug report to the developer of the tool you're using.
- If you are not using such a tool you'll have to start using one if you want to make use of Competitive Companion. For Sublime Text the combination of FastOlympicCoding and FastOlympicCodingHook should work (note that neither of those tools are made by me so I can't provide support for them).

At the moment I don't think there are vim specific tools, although you could try IDE-agnostic tools like Caide or Hightail. You could also create your own tool which is documented here (for inspiration you can watch the first ~12 minutes of ecnerwala's Google Kickstart 2020 Round E video in which he explains his setup, which includes vim and Competitive Companion).

I created a parser for TLX/TOKI and sent a pull request: https://github.com/jmerle/competitive-companion/pull/122

Yes please fix the extension for Codeforces. It was so helpful. Now it's quite painful to manually copy and match the outputs.

The answer is two-fold:

1. Competitive Companion only parses problems and sends the parsed data to external tools, it does not create the C++ files you're talking about (that's handled by said external tools). Competitive Companion does not control how these tools name the files they create.
2. Competitive Companion aims to support a large number of judges, and does so by parsing the common parts of all these problems. This means it parses things like time/memory constraints and sample cases (which almost all judges have), but not things like difficulty or success rate (which only a few judges have). I'm not planning to change this. However, Competitive Companion does send external tools the URL to the parsed problem, so technically it's possible for these tools to retrieve the difficulty of Codeforces problems themselves.