`[`, `(`, `,`, `;` should be allowed before the `@user` tag

[albert-ying]

Please select if your request is either something new or an enhancement

• [x] Enhancement of an existing Feature.
• [ ] Request of a new feature.

Please select the area your request applies to. (Multiple selections are Possible. You can leave blank if you're not sure.)

• [ ] Workspace - VSCode workspace, vaults, Intellisense/autocomplete, Dendron settings
• [ ] Views - Dendron Preview, Tree View, Side Panels in the UI
• [ ] Lookup - Dendron's Lookup Command
• [ ] Schema - Dendron Schemas
• [ ] Pod - Data import from / export to Dendron
• [ ] Publish - External Site Publish
• [x] Markdown - Markdown features, such as syntax support and features

Is your feature request related to a problem? Please describe

Currently @user tag doesn't work whenever there is non-space character before it. However, sometimes people may want to tag people in following ways:

- xxxx (@user is handling this)
- @user1,@user2
- @user1;@user2
- [@citekey1;@citekey2] #for using user tag for citation notes

All of these doesn't recognize the tag as link.

Describe the solution you'd like

Special characters, including [, (, ,, ; should be allowed before the @user tag

[kevinslin]

thanks for the issue. have added it as a crop candidate

[aleksey-rowan]

This issue is featured in this week's round of CROP.

[rlh1994]

I haven't been able to build proper tests to be able to submit an actual PR for this, but if someone does come to implement it - from manual testing you should only need to do is swap out the regexs in the following :

https://github.com/dendronhq/dendron/blob/9efdf2c453e822972e5cb749e558b00de572005b/packages/engine-server/src/markdown/remark/userTags.ts#L30-L44

becomes:

export const USERTAG_REGEX = new RegExp(
  // Require one of ,;([  or whitespace before to avoid something like [email protected]
  `^(?<=^|[\\s,;\\(\\[])(?<tagSymbol>@)(?<tagContents>` +
    `${GOOD_MIDDLE_CHARACTER}*` +
    `${GOOD_END_CHARACTER}` +
    `)`
);
/** Same as `USERTAG_REGEX`, except that that it doesn't have to be at the start of the string. */
export const USERTAG_REGEX_LOOSE = new RegExp(
  // Require one of ,;([  or whitespace before to avoid something like [email protected]
  `(?<=^|[\\s,;\\(\\[])(?<userTag>@)(?<userTagContents>` +
    `${GOOD_MIDDLE_CHARACTER}*` +
    `${GOOD_END_CHARACTER}` +
    `)`
);

and https://github.com/dendronhq/dendron/blob/9efdf2c453e822972e5cb749e558b00de572005b/packages/engine-server/src/markdown/remark/userTags.ts#L82-L96

becomes

  function locator(value: string, fromIndex: number) {
    // Do not locate a symbol if the previous character is non-whitespace (except one of ,;([ ).
    // Unified cals tokenizer starting at the index we return here,
    // so tokenizer won't be able to reject it for not starting with a non-space character.
    const atSymbol = value.indexOf("@", fromIndex);
    if (atSymbol === 0) {
      return atSymbol;
    } else if (atSymbol > 0) {
      const previousSymbol = value[atSymbol - 1];
      if (!previousSymbol || /[\s([,;]/.exec(previousSymbol)) {
        return atSymbol;
      }
    }
    return -1;
  }