fix(message-parser): Link parser issues

[hugocostadev]

Proposed changes (including videos or screenshots)

I've done several changes to grammar.pegjs to improve the link parser, such as:

• Added a library called tldts that validates if it's a valid TLD (Top-Layer-Domain) this fixed several texts dat had a dot between and that was not a link to become a link
• Fixed some special chars like : -, _ and others to be accepted in the link
• Removed the digit restriction to start the domain or TLD
• Removed last comma or dot that follows an URL to be parsed as a part of the LINK.

Previous result:

Current Results (NEW) :

Benchmark results

develop branch (current parser)

this PR (proposed parser)

Issue(s)

RocketChat/Rocket.Chat#27376
RocketChat/Rocket.Chat#27484
RocketChat/Rocket.Chat#26476
RocketChat/Rocket.Chat#27078

Further comments

TC-294

[Matt-something]

You're lacking any testing for single word domains using DNS to resolve back to an IP

https://singleword/directory/word should be recognised as a URL

[hugocostadev]

You're lacking any testing for single word domains using DNS to resolve back to an IP

https://singleword/directory/word should be recognised as a URL

Meanwhile you can use as this [https://singleword/directory/word](https://singleword/directory/word) , we accepts file path (the pattern of this link) inside the reference link