jedireza / reserved-subdomains

:package: A list of names that should be reserved in multitenant apps.

reserved-subdomains

A list of names that should be reserved in multitenant apps.

Install

$ npm install reserved-subdomains

API reference

• Properties
  • names
  • nameMap
  • patterns
  • patternsRegex
• Methods
  • isNotValid(name)
  • isValid(name)

Properties

names

An array of strings. Ex: ['www', 'ftp', 'mail', ...].

nameMap

An object where keys are names. Ex: { www: true, ftp: true, mail: true, ...}.

patterns

An array of regular expression strings. Ex: ['ww[a-z0-9-]+', 'ftp[0-9]+', 'mail[0-9]+', ...].

patternsRegex

An array of regular expression objects. Ex: [/ww[a-z0-9-]+/, /ftp[0-9]+/, /mail[0-9]+/, ...].

Methods

isNotValid(name)

Returns true if the name is not valid where:

• name - the name to test.

isValid(name)

Returns true if the name is valid where:

• name - the name to test.

Inspired by

https://github.com/nkkollaw/reserved-subdomains

License

MIT

reserved-subdomains's Issues

Single-letter words?

I noticed most single-letters (a,b,c, etc) are in the list but it's missing some. Is that purposeful or accidental?

Could we adjust this pattern?

In https://github.com/jedireza/reserved-subdomains/blob/master/patterns.json#L20
Is there a reason why a-z is included in the pattern whereas all earlier patterns only do [0-9]?

I have a situation where someone legitimately wants to have a subdomain with ww in the beginning and while I could just fork it would be nice if the pattern:

"ww[a-z0-9-]+" could be changed to "ww[0-9]+"

In the actual main package in npm.
https://www.npmjs.com/package/reserved-subdomains

Thanks for considering!

Separate by category?

This is really handy but I was wondering what you thought about separating out the "external companies" into a different list. I'm not sure I care much if someone creates riteaid.myapp.com but I definitely don't want them creating admin.myapp.com or login.myapp.com.

Right now I'm deleting everything I can find that is external-company related from the list. Would you be interested in a pull request that separates these?

Match whole strings instead of partial strings when using patterns

I feel it would be better if patterns matched only whole strings instead of partial strings. For example^m[0-9]+$ instead of m[0-9]+. Because the following harmless names end up failing validation:

momentum1
kalashnikov1
bmx1
ghost1
humans1

This could be remedied by changing the pattern list to:

[
    "^db[0-9]+$",
    "^dc[0-9]+$",
    "^dev[0-9]+$",
    "^dns[0-9]+$",
    "^ftp[0-9]+$",
    "^host[0-9]+$",
    "^m[0-9]+$",
    "^mail[0-9]+$",
    "^mx[0-9]+$",
    "^ns[0-9]+$",
    "^server-[0-9]+$",
    "^server[0-9]+$",
    "^smtp[0-9]+$",
    "^static[0-9]+$",
    "^test[0-9]+$",
    "^v[0-9]+$",
    "^vpn[0-9]+$",
    "^web[0-9]+$",
    "^ww[0-9]+$"
]

Should I send a PR?