How to use regex to match paragraphs containing a specific word?

Question

I tried the answer of this post Regex. Find paragraph containing some word, in my case that would be

((?!\n\n).)*(cat)

, but this don't work.

How can I use PCRE2 regular expressions (PHP >= 7.3) to match all paragraphs in my text that contain the word "cat", where each paragraph is separated by two consecutive line breaks (It is allowed to have one line break in a paragraph but not two)?

For example, if the input text is as follow

Paragraph 1 wepfowfpo
fww efwf

Paragraph 2 wefwf32321
!@d r33tcat54, 333!..

Paragraph 3 4t4t022
-`121231ere3r3cat342232
$ 4t0g cat rdwd203  
$$333

Paragraph 4 222cocdo3

Then the desired ouput is

Paragraph 3 4t4t022
-`121231ere3r3cat342232
$ 4t0g cat rdwd203  
$$333

I tried to use something like \n\n.*(?=cat)cat.*\n\n, but this match only those lines contain "cat".

bobble bubble · Accepted Answer · 2023-04-20T13:25:37.870

How about splitting the string into paragraphs and matching those containing cat.

preg_grep('/\bcat\b/i', explode("\n\n", $str));

See this PHP demo at tio.run - The word bundary \b prevents from matching tcat5.

If you can't use PHP functions, following a regex-only idea for (?m) multiline mode.

^(?:.+\n)*.*?\bcat\b.*(?:\n.+)*

See this demo at regex101 - Further add i flag to ignore case (also match e.g. Cat).

regex	explained
`(?m)`	flag for multiline mode to make `^` match line start too
`^(?:.+\n)*`	at `^` start repeat the `(?:` non capturing group `)` `` any amount* of times, containing: `.+` greedily match one or more chars up to `\n` newline - part that matches lines before (if available, use of atomic group instead non capture can be more efficient here: demo)
`.?\bcat\b.`	`.?` matches lazily* any characters up to `\bcat\b` (using word bundaries) `.*` rest of line
`(?:\n.+)*`	matches any remaining lines in the paragraph where `.+` prevents to skip over `\n\n`

How to use regex to match paragraphs containing a specific word?

1 Answers1