Extension causes high cpu load

[nkprince007]

• Issue Type: Performance
• Extension Name: markdown-table-formatter
• Extension Version: 2.1.3
• OS Version: Linux x64 5.9.11-3-MANJARO
• VSCode version: 1.52.1

⚠️ Make sure to attach this file from your home-directory:
⚠️/tmp/fcrespo82.markdown-table-formatter-unresponsive.cpuprofile.txt

Find more details here: https://github.com/microsoft/vscode/wiki/Explain-extension-causes-high-cpu-load

[fcrespo82]

Sorry for the looooong delay on this.

Are you still seeing this issue?

If yes, can you please upload the cpuprofile file.

[brunnerh]

I might have a way to reproduce this:

• Create new JavaScript file
• Write HTMLElement
• Call command Go to Definition

This should open up the type declaration file lib.dom.d.ts, which is huge (my version has around 19k lines).
Doing this causes some very long or infinite processing, essentially taking out the extensions host.

Why is the extension running on non-markdown files, though?

[fcrespo82]

Thanks for the feedback.

Why is the extension running on non-markdown files, though?

It shouldn't, I'll look into it as soon as possible.

[fcrespo82]

Hey @brunnerh, I couldn't reproduce the way you told. Maybe you enabled the extension to the typescript language by mistake?

Please look at your settings, look for the key markdown-table-formatter.markdownGrammarScopes and tell me if it was the case, please.

I will continue to search for other ways to mitigate this kind of problem, but for now I will add a confirmation dialog to avoid the extension being enabled on other languages by mistake...

[brunnerh]

Oh, you are correct, for some reason this setting has been modified, but i do not know why. Possibly because at some point i created tables in documentation texts and wanted to format them.

"markdown-table-formatter.markdownGrammarScopes": [
	"markdown",
	"csharp",
	"svelte",
	"typescript"
],

Still, the extension does not seem to handle large files too well and/or different languages cause it trouble. Maybe that could be improved?

[fcrespo82]

Of course it should handle large files better, this is now my priority.

Thanks again.

[fcrespo82]

Hey @brunnerh and @nkprince007 I think I was able to solve the high CPU load.

Can you folks test with this pre-release version, and give me feedback?

@brunnerh now it should parse the 19k file in less than 2s, it was the results I got in my tests.

Thanks in advance.

[brunnerh]

Tested it and it appears to not cause any issues anymore. 👍

[fcrespo82]

Glad to hear. I'll make some more tests today, and then release it later today, or tomorrow.

Thanks for the feedback and helping identifying the problem.

[fcrespo82]

Should be fixed in v2.2.2