Get a portion of text delimited by cariage return with regex

Question

I have a text like:

This patch requires:

Patch 1.10-1
Patch 1.11-2

Notes:

I want to extract

Patch 1.10-1
Patch 1.11-2

with the following regex:

This patch requires\:[\r\n](.*)[\r\n]Notes\:

But nothing is matched.

Why ?

`[\r\n]` > `[\r\n]+` and use a **DOTALL modifier** or `.` will not match across line breaks. See [demo](https://regex101.com/r/I97ebV/2), use `'~This patch requires:\R+(.*?)\R+Notes:~s'` and grab `$match[1]`. — Wiktor Stribiżew, Feb 07 '18 at 09:36
I used `This patch requires\:[\r\n]+(.*)[\r\n]+Notes\:` but same thing, nothing returned — Snake Eyes, Feb 07 '18 at 09:37
PHP ... I tested in regex.com and seems if I switch to Single Line option then works otherwise not... — Snake Eyes, Feb 07 '18 at 09:39
Not working without `/This patch requires\:[\s\S]*?Notes\:/gi` , I have to add `s` for single line like: `/This patch requires\:[\s\S]*?Notes\:/gis` — Snake Eyes, Feb 07 '18 at 09:41
Single Line option or `DOTALL` is required because you are `.*` that is supposed to match multiline string. Without `s` you can use `[\s\S]*?` also — anubhava, Feb 07 '18 at 09:42
You don't need `s` flag if you don't use `.` in your regex. Hence `[\s\S]`. — revo, Feb 07 '18 at 09:42
`This patch requires\:[\s\S]*?Notes\:` does not capture the text I wish. It captures everything. — Snake Eyes, Feb 07 '18 at 09:43

Jan · Answer 1 · 2018-02-07T09:49:18.160

For the sake of alternative solutions:

^\QThis patch requires:\E\R+
\K
(?:^Patch.+\R)+

This says:

^\QThis patch requires:\E     # match "This patch requires:" in one line and nothing else
\R+                           # match empty lines + newlines
\K                            # "Forget" what's left
(?:^Patch.+\R)+               # match lines that start with "Patch"

In PHP:

<?php

$regex = '~
    ^\QThis patch requires:\E\R+
    \K
    (?:^Patch.+\R)+
         ~xm';

if (preg_match_all($regex, $your_string_here, $matches)) {
    // do sth. with matches
}

See a demo on regex101.com (and mind the verbose and multiline modifier!).

score 0 · Answer 2 · edited Feb 07 '18 at 09:53

0

What about (?<=[\r\n])(.+)(?=[\r\n])

Demo

edited Feb 07 '18 at 09:53

Wiktor Stribiżew

484,719
26
302
397

answered Feb 07 '18 at 09:39

Mohammad Javad Noori

1,059
9
23

Note that `(?=[\r\n]+)` will work the same as `(?=[\r\n])` since `.*` is a greedy subpattern. Trimming will be required as an additional step. – Wiktor Stribiżew Feb 07 '18 at 09:46
thanks, can you edit ? – Mohammad Javad Noori Feb 07 '18 at 09:50
I have. Still, what do you mean by "what about"? Answers are helpful when you explain the approach, and here, you are using quite a different approach than OP has. You have two matches now, while OP expects to extract a single 1+-line string. – Wiktor Stribiżew Feb 07 '18 at 09:54
You right , I can speak English a little ,so what should I use? – Mohammad Javad Noori Feb 07 '18 at 09:58
[This regex](https://regex101.com/r/I97ebV/3). – Wiktor Stribiżew Feb 07 '18 at 10:00
thank you @WiktorStribiżew – Mohammad Javad Noori Feb 07 '18 at 10:10

revo · Accepted Answer · 2018-02-07T10:12:15.857

But nothing is matched.

DOT . doesn't match newline characters so .* doesn't go beyond one line. For that you need to set DOTALL s modifier on or try an alternative [\s\S] (which literally means any character without any exception).

preg_match('~This patch requires:\s*([\s\S]*?)\s*^Notes:~m', $text, $matches);
echo $matches[1];

Note: don't use greedy-dot (.*) if multiple matches are going to occur instead go lazy .*?

By the way this is supposed to be the most efficient regex you may come across:

This patch requires:\s+((?:(?!\s*^Notes:).*\R)*)

Live demo

Get a portion of text delimited by cariage return with regex

3 Answers3