[BUG] Rule matching is too sensitive

[shaonianruntu]

Rule matching is too sensitive, and the matching content will not only focus on the annotation, but also match the code.

For example, following is my configuration of todo-tree:

"todo-tree.general.tags": [
    // "BUG",
    // "HACK",
    // "FIXME",
    // "TODO",
    // "XXX",
    // "[ ]",
    // "[x]",
    "?",
    "^",
    "~",
    "-",
    "+"
  ],
  "todo-tree.highlights.customHighlight": {
    // flag of question
    "?": {
      "type": "text-and-comment",
      "foreground": "Gold",
      "icon": "question",
      "gutterIcon": true,
      "iconColour": "Gold"
    },

I hope it can match such annotation format like “//~ [balabala]”.

However, there is destructor function in C++, and the function name begins with ~，like "~functionName()". These statements are not annotation, but they will also be overmatched by todo-tree. I think it is a big error.

And also, todo-tree will do not only match "//~ [balabala]", but also match the statement like "//~~~", or "~~~~~" in C++.

*I hope it can only match the statement like "//~ comment" or "// ~ comment" in C++. (regex: ^[FLAG]\x20.$) rather than match "~", "~function", "~~", "//~~".* (regex: ^[FLAG].$)

Is there something wrong with my rule configuration? Or is it a problem that todo-tree matching is too sensitive.

Hoping your help.

Thanks & Regards Nan

[Elias-Graf]

Could potentially be related to #667. E.g. that the extension has issue with special regex characters?

(Yes, I've deleted a previous comment, because I've misread what you wrote)

[shaonianruntu]

Could potentially be related to #667. E.g. that the extension has issue with special regex characters?

(Yes, I've deleted a previous comment, because I've misread what you wrote)

No related.

This is really difficult to express.

But what I mean,

the commonly used C++ annotation format is //.

If I set + as a todo-tree tag. The result I want is to mark is sentences begin with //+ or // +. e.g, return 0; // + to add some function before here

But the result now is that it will mark all the "+" in the file, not only labeling + under the annotation statement, but also the plus sign in the operation expression. e.g. a = 1 + 2;

It seems that the regular object is +, not //~ or // ~.

[maxaxehax]

@shaonianruntu This works perfectly fine for me. You have set the wrong regex expression, where you are not prefixing your + tag with the // line-comment opener for your language of choice.

Use regexr.com and the "todo-tree.regex.regex" setting and double check your settings.

[shaonianruntu]

@shaonianruntu This works perfectly fine for me. You have set the wrong regex expression, where you are not prefixing your + tag with the // line-comment opener for your language of choice.

Use regexr.com and the "todo-tree.regex.regex" setting and double check your settings.

Thank you very much. You are right! I should add // before myself define tags, like // todo or // +, ... rather than todo or +

Thanks