Enhance TokenStore methods with includeComments option

[kaicataldo]

Having used the new TokenStore token traversal methods this week (while working on this PR), I think we should add an includeComments option to the following: getTokenByRangeStart(), getTokens(), and getTokensBetween().

The API I'm envisioning is:

getTokenByRangeStart(index, {
    includeComments: true
});

getTokens(node, {
    beforeCount: 0,
    afterCount: 0,
    includeComments: true,
    filter: filterFn
});

getTokensBetween(leftToken, rightToken, {
    padding: 0,
    includeComments: true,
    filter: filterFn
});

The proposed change to getTokenByRangeStart() is not a breaking change, but the other two are (though I suppose we could do this in a backwards compatible manner still). I'd ideally like to use getTokens in the aforementioned PR here.

Also wanted to note that the beforeCount, afterCount, and padding options are all undocumented and look like they were included to avoid a breaking change.

cc @mysticatea

[mysticatea]

Actually, getTokens and getTokensBetween are supporting includeComments and filter options. Though those object-style options are not supporting padding option.

interface CountOptions {
    includeComments?: boolean
    filter?: (token: Token|Comment) => boolean
    count?: number
}

interface TokenStore {
    //...

    getTokens(node: Node): Token[]
    getTokens(node: Node, beforeCount: number, afterCount: number): Token[]
    getTokens(node: Node, filter: (token: Token) => boolean): Token[]
    getTokens(node: Node, options: CountOptions): (Token|Comment)[]
    
    getTokensBetween(left: Node|Token|Comment, right: Node|Token|Comment): Token[]
    getTokensBetween(left: Node|Token|Comment, right: Node|Token|Comment, padding: number): Token[]
    getTokensBetween(left: Node|Token|Comment, right: Node|Token|Comment, filter: (token: Token) => boolean): Token[]
    getTokensBetween(left: Node|Token|Comment, right: Node|Token|Comment, options: CountOptions): (Token|Comment)[]
}

[kaicataldo]

@mysticatea Ah, sorry then - thanks for the explanation. I'd still like to propose the change to getTokenByRangeStart(), but since that's no longer a breaking change, I'll just open a PR (have one ready based on our last discussion @mysticatea).