`tokenizeLine` is not a function

Hello, I’m Denis, an Italian developer ! I’m trying to craft an advanced postfix completion helper for vscode : https://github.com/Denis-net/vscode-definitive-postfix

I wanted to activate my suggestions only when not inside comment blocks (tokens), I tried regexs but I didn’t manage to distinguish strings from comment or line comment to block comment etc.

( for example:

var x=5; // /*
x++;
// */
var aString = "//"; x++;

)

I discovered that I need a syntax analyzer/tokenizer and after a long search i came across this intresting project. But I missed something important, because I can’t make it working. I read about tmLanguage and plist, I found them on vscode.env.appRoot/extensions/*/syntaxes and on textmate repo.

In extension.ts.loadParser(:110) method I call the parseRawGrammar, then I store the result in my global singleton(SnippetCollector.Current) field.

In completionProvider.ts.isInsideComment(:46) method I tried to use the parser, but It raises an exception :

“parser.tokenizeLine is not a function”

so it didn’t create the expected object.

Furthermore, in the loadGrammar.then()(extension.ts:137) the grammar arg is null, while inside the promise I can load a rawGrammar object.

Could you help me ? Maybe you know what I’m doing wrong, or if there is another way of doing it.

Thanks

[ghost]

I tried to compile it but I had issues with node/electron environment (not understood well)… anyway.

I found a compiled version in C:\Users\User\AppData\Local\Programs\Microsoft VS Code\resources\app\node_modules.asar.unpacked\oniguruma\build\Release So I copied it into my folder and… the good news is : It works ✨ the bad news is : don’t work as I expected 😔

Using plist:

Using tmLanguage I can get more information:

But what I would really like to have are the built-in scope inspector’s data : With It i could understand exactly where I am (more specifically if user cursor is in a comment)

I think I had misunderstood the vscode-textmate plugin. Or I didn’t make it working …

Anyway Thank you @msftrncs for the great help !

For patching the doc (Readme.md) what can we do ? Because isn’t specified that developers need the Registry and that the addGrammar().then() will return the real Grammar object. Not speaking about this oniguruma build dependency.

Or can i just close this issue for posterity ?

[ghost]

Ok, The best I can do is this: I opened this pull request. Hope this could help in fixing the Readme so other developers could save time. Thanks again !