REGEX: Match strings that contain a specific word but not strings that merely contain that word-CodePudding

For example, I want to match all strings that contain the word 'cat' or 'dog' such as concatenation, doghouse, underdog, catastrophe, or endogamy. But I want to exclude the words dogs or cats from being matched. I tried this task using the following regex.

\\w*(cat|dog)(s(?=\w ))*\

But this regex doesn't help me select whatever is after the s. Is there some other way to achieve this? Any help is appreciated.

CodePudding user response：

This regex should work

const regex = /\b(?!cats|dogs)[a-z]*(?:cat|dog)[a-z]*\b/gi;
const m = 'concatenation, doghouse, underdog, catastrophe, endogamy should match, but not cats and dogs.'.match(regex);
console.log(m);

Output:

[
  "concatenation",
  "doghouse",
  "underdog",
  "catastrophe",
  "endogamy"
]

Explanation of regex:

\b -- word boundary
(?!cats\b|dogs\b) -- negative lookahead for just cats or dogs
[a-z]* -- optional alpha chars
(?:cat|dog) -- non-capture group for literal cat or dog
[a-z]* -- optional alpha chars
\b -- word boundary

CodePudding user response：

If you also don't want to match dogsdogs you might write the pattern as:

\b(?!\w*(?:cats\b|dogs\b))\w*(?:cat|dog)\w*

The pattern matches:

\b a word boundary
(?! Negative lookahead, assert that to the right is not
- \w*(?:cats\b|dogs\b) Match optional word characters followed by the word cat or dog followed by a word boundary
) Close the lookahead
\w*(?:cat|dog)\w* Match cat or dot between word characters

Regex demo

If a lookbehind assertion is supported, and you also want to allow other non whitespace characters, you can use \S to match a non whitespace character instead of \w that matches a word character.

(?<!\S)(?!\S*(?:cats\b|dogs\b))\S*(?:cat|dog)\S*

See another Regex demo

CodePudding user response：

I understand your requirements as: match everything that has cat/dog anywhere in word apart from the specific words 'cats' and 'dogs'

  \b(?!cats\b|dogs\b)(?=\S*cat\S*|\S*dog\S*)\S*\b

(very) Rough human translation: Find a point where a word isn't cats or dogs (ending with word boundary) and then find a point where a word has cat or dog (either at start, middle, or end) then match everything till the end of the word from that point

Note: flavour - PCRE2