CODEWITHSUNDEEP
javascriptregex
Pal Singh
Pal Singh

Reputation: 1974

Regex match character before and after underscore

I have to write a regex with matches following:

I came up with the following regex, but it doesn't seems to work

/^[a-zA-Z0-9]+(\b_|_\b)[a-zA-Z0-9]+$/

Test case:

HelloWorld // Match
Hello_World //Match
Hello _World // doesn't match
Hello_ World // doesn't match
Hello _ World // doesn't match
Hello_World_1 // Match
He110_W0rld // Match
Hello - World // Match
Hello-World // Match
_HelloWorld // doesn't match
Hello_-_World // match

Upvotes: 3

Views: 7849

Answers (4)

Whothehellisthat
Whothehellisthat

Reputation: 2152

Oh! I love me some regex!

Would this work? /^[a-z]$|^[a-z](?:_(?=[^ ]))?(?:[a-z\d -][^ ]_[^ ])*[a-z\d -]*[^_-]$/i

I was a tad unsure of rule 4--do you mean underscores can have a space before or after or neither, but not before and after?

Upvotes: 0

trincot
trincot

Reputation: 350137

You could use this one:

^(?!^[ _-]|.*[ _-]$|.* _|.*_ )[\w -]*$

regex tester

For the test cases I used modifier gm to match each line individually.

If emtpy string should not be considered as acceptable, then change the final * to a +:

^(?!^[ _-]|.*[ _-]$|.* _|.*_ )[\w -]+$

Meaning of each part

  • ^ and $ match the beginning/ending of the input
  • (?! ): list of things that should not match:
    • |: logical OR
    • ^[ _-]: starts with any of these three characters
    • .*[ _-]$: ends with any of these three characters
    • .* _: has space followed by underscore anywhere
    • .*_: has underscore followed by space anywhere
  • [\w -]: any alphanumeric character or underscore (also matched by \w) or space or hyphen
  • *: zero or more times
  • +: one or more times

Upvotes: 2

Wiktor Stribiżew
Wiktor Stribiżew

Reputation: 626738

You may use

^(?!.*(?:[_-]$|_ | _))[a-zA-Z][\w -]*$

See the regex demo

Explanation:

  • ^ - start of string
  • (?!.*(?:[_-]$|_ | _)) - after some chars (.*) there must not appear ((?!...)) a _ or - at the end of string ([_-]$), nor space+_ or _+space
  • [a-zA-Z] - the first char matched and consumed must be an ASCII letter
  • [\w -]* - 0+ word (\w = [a-zA-Z0-9_]) chars or space or -
  • $ - end of string

Upvotes: 8

Markus Laire
Markus Laire

Reputation: 2947

What about this?

^[a-zA-Z](\B_\B|[a-zA-Z0-9 -])*[a-zA-Z0-9 ]$

Broken down:

^               
[a-zA-Z]        allowed characters at beginning
(
 \B_\B          underscore with no word-boundary
|                 or
 [a-zA-Z0-9 -]  other allowed characters
)*
[a-zA-Z0-9 ]    allowed characters at end
$

Upvotes: 1

Related Questions