regex lookarounds: how to extract a full match with optional lookahead

Question

to describe the problem:

I have some text with mail-header-lines like

From: me
To: you
Subject: welcome, this is a long line of subject with two 
         lines of text
Attachements: welcome.jpg, foo.pdf

the last line ('Attachements: welcome.jpg, foo.pdf') is OPTIONAL. so maybe this text only looks like

From: me
To: you
Subject: welcome, this is a short line of subject

I need to extract the Subject-Line(s), without the text 'Subject:'. leading and trailing whitespace are no problem.

the only operation I can use is a SINGLE QT-Regex-call that returns a FULL MATCH ONLY.

great, isn't it ?

I tried with success

(?<=Subject:)(?:\s*)(.*)(?=Attachements:)

but how do I make the 'Attachements:'-line optional ?

When there is no Attachement:-line, I expect the text/string to end with the Subject:-line(s).

any idea ?

to be honest: I don't know. I'm using a product that is created with Qt (5.7.1). And they allow for just a single line of REGEX in their interface. :-( — Marc, Feb 14 '18 at 13:44
Let's assume this is `QRegularExpression`. Try [`(?s)Subject:\s*\K.*?(?=\RAttachements:|$)`](https://regex101.com/r/NbRBF9/3). — Wiktor Stribiżew, Feb 14 '18 at 13:52
Else, try [`(?<=Subject:)\s*((?:(?![\r\n]Attachements:).)*)`](https://regex101.com/r/HXMdQ4/1) — Wiktor Stribiżew, Feb 14 '18 at 13:54

mrzasa · Answer 1 · 2018-02-14T14:27:38.193

0

You can use negative lookeahead (?!...) forAttachments

(?<=^Subject: )(?:(?!^Attachements:)[\s\S])+

Btw. I've changed .* to [\S\s]* to allow newlines in the subject.

edited Feb 14 '18 at 14:27

answered Feb 14 '18 at 11:35

mrzasa

thank you. I also tried the variants with \z. but: your sample does not contain the 'Attachement:'-line. and it does not cut out only the Subject:. – Marc Feb 14 '18 at 13:48
In fact, `([\S\s](?!Attachements:))+` is not a correct tempered greedy token, you should temper the dot before it matches a char, use `(?:(?!Attachements:)[\s\S])+` – Wiktor Stribiżew Feb 14 '18 at 14:16
thanks @WiktorStribiżew, could you tell why is that? - my regexp seem to work correctly (but I'm eaget to learn) – mrzasa Feb 14 '18 at 14:18

score 0 · Accepted Answer · answered Feb 14 '18 at 14:57

You may use

(?<=Subject:)\s*((?:(?![\r\n]Attachements:).)*)

The pattern matches:

(?<=Subject:) - a positive lookbehind that matches a position inside a string where there is a Subject: substring immediately to the left of the current location
\s* - 0+ whitespace chars
((?:(?![\r\n]Attachements:).)*) - Capturing group 1 that matches any char (QRegExp regex . pattern even matches line breaks), zero or more repetitions as many as possible (*), that does not start a CR/LF + Attachments: char sequence. This construct is called a tempered greedy token.

2 Answers2