什么正则表达式可以在一个短语中捕获2个确切的“单词”？

Question

I'm trying to capture words constant in a string. That constant is:

1. a word
2. followed by one separator (whitespace, dot, dash or underscore)
3. another word
4. and then a separator (see #2) or the end of the line or string.

For the sake example let's say I'm looking for "Bob 1", in the following strings:

Hello, I'm Bob 1 --> Should capture Bob 1
Hello, I'm Bob 11 --> Should capture nothing (Bob 1 is not at the end or followed by a separator)
Hey, it's Bob-1 over there --> Should capture Bob-1
Hey, it's Bob - 1 over there --> Should capture nothing (Bob should be followed only by one separator not 3 like here)
Bob.1 --> Should capture Bob.1
Bob_1 rules! --> Should capture Bob_1

I have a regex that mostly works:

/Bob[\s._-]1[\s._-]/ig

In the second list I don't know how to add the end of the string in the possible characters... Which ends in only the last line in the live demo below that should be a match and that isn't captured.

See the live demo .

I use pcre (in PHP).

Answer 1

I'm not using PHP, but the following matches for me:

\bBob[\s.\-_]1\b

It is making use of \\b which matches against a word boundary. I found that I had to escape the dash inside the square brackets, which isn't something you are doing but that may be a difference between the regex engines we are using.

Answer 2

This works

https://regex101.com/r/ezikuP/2

(?<!\\S)Bob[\\s._-]1(?![^\\s._-])

Formatted

 (?<! \S )               # Whitespace boundary
 Bob                     # Word 1
 [\s._-]                 # Special seperator
 1                       # Word 2
 (?! [^\s._-] )          # Special seperator boundary

Answer 3

Which ends in only the last line in the live demo below that should be a match and that isn't captured.

For that you need a positive lookahead.

Regex: Bob[\\s._-]1(?=[\\s._-])

• (?=[\\s._-]) will only look for give character class and won't match/capture it.

Regex101 Demo

Answer 4

In the second list I don't know how to add the end of the string in the possible characters.

You can use this regex with anchor $ to assert end of string:

/\bBob[\s._-]1(?:[\s._-]|$)/m

OR if you don't want to match next character after 2nd word then use a lookahead:

/\bBob[\s._-]1(?=[\s._-]|$)/m

([\\s._-]|$) will assert presence of given (one of whitespace, DOT, Underscore, Hyphen) characters or end of line $ .

It is safer to add \\b before Bob to match exact word Bob and avoid matching HelloBob

RegEx Demo