How can I prevent the repetition of the words I'm looking for regardless of their order?

Question

I've spent like one hour trying to make a better regex but that's not my cup of tea... I need a regex which will do the following (can provide more if needed):

Spd_Engine          #Ok
speedengine         #Ok
enginespd           #Ok
Engine_speed        #Ok
aps_speed_engine    #Ok
engine_speed        #Ok
engine_trq          #Not Ok
speed_rpm           #Not Ok

The regex shoud match every line which contains at least (engine && (speed || spd))

So I came up with this:

[e,E]ngine[_]?[s,S]p[e]*d|[a-zA-Z]*[_]*[s,S]p[e]*d[_]?[e,E]ngine

But I feel it can be improved. How can I simplify it?

For convenience, could you please edit your question with a description in words of what your validation check is? It's not obvious why aps_speed_engine is acceptable but engine_trq isn't. — shree.pat18, Dec 05 '14 at 09:46
Inside character class you need not use commas, `[eE]` the correct format. But you can use `i` flag to ignore the case checking — nu11p01n73R, Dec 05 '14 at 09:48
This isn't code review. "Improve" is opinion-based. RegEx is overkill, `input = input.ToLower(); if ((input.Contains("speed") || input.Contains("spd") && input.Contains("engine")) { ... }` will do the same and is more readable. — CodeCaster, Dec 05 '14 at 09:51
@nu11p01n73R I didn't manage to find the correct syntax for the `i` flag using [regex hero](http://regexhero.net/) — Thomas Ayoub, Dec 05 '14 at 09:51
@CodeCaster Sorry, as this is not the first question about reviewing regex in SO I took my chance — Thomas Ayoub, Dec 05 '14 at 09:52
I'm not a fan of regex an I bet your maintenance programmers wont be either. I'm with codecaster on this — Kell, Dec 05 '14 at 09:56
You can start by asking a more specific question. "I feel" and "simplify" aren't very factual. _"How can I prevent the repetition of the words I'm looking for regardless of their order?"_ is a more precise question. — CodeCaster, Dec 05 '14 at 09:56

nu11p01n73R · Accepted Answer · 2014-12-05T10:08:37.037

9

You can use look aheads to simplify the regex a lot as

^(?=.*spe*d)(?=.*engine).*

^ Anchors the regex at the start of the string
(?=.*spe*d) positive look ahead. Checks if the string contains spe*d
(?=.*engine) another postive look ahead. Checks if the string contains engine
.* matches the entire string

Regex Example

OR

^(?=.*spe*d).*engine.*

Droping the second look ahead

Regex Example

Notes on [e,E]ngine[_]?[s,S]p[e]*d|[a-zA-Z]*[_]*[s,S]p[e]*d[_]?[e,E]ngine

[e,E] the commas inside character class does not mean e comma E. You can modify it as

[eE]
[_]? There is no advantage in adding a single character in character class. It is simlar as wirting _?
i flag i can be used to ignore cases while matching the regex

edited Dec 05 '14 at 10:08

answered Dec 05 '14 at 09:54

nu11p01n73R

26,397
3
39
52

2

I like http://www.regexr.com/ more – Emre Acar Dec 05 '14 at 09:57
And also, i was just playing with your expression and after i delete second ?= nothing changed in the result :/ i cant write it here because two asterisks are counting as bold here. – Emre Acar Dec 05 '14 at 10:03
1

@EmreAcar There are offcourse different ways you can acomplish the same. :) – nu11p01n73R Dec 05 '14 at 10:05

score 0 · Answer 2 · answered Dec 05 '14 at 09:56

0

So if shorter version of this regex is count as simplification then it is what came to my mind

[e,E]ngine_?[s,S]ped|[a-zA-Z]_[s,S]ped_?[e,E]ngine

VS

[e,E]ngine[]?[s,S]p[e]*d|[a-zA-Z]*[]*[s,S]p[e]*d[_]?[e,E]ngine

answered Dec 05 '14 at 09:56

Sebastian 506563

6,980
3
31
56

How can I prevent the repetition of the words I'm looking for regardless of their order?

2 Answers2