And operator in REGEX

Question

I am trying to find two separate groups of text within a document using REGEX in an application. Example:

(facility services|MAFS|MFA|facility|facilities)
(agreement|lease)

I only want to identify documents that have a match to one word in both sets of text, such as facility and agreement. How would I write that in REGEX?

`((facility services|MAFS|MFA|facility|facilities)(agreement|lease))` — zer00ne, Jun 26 '18 at 20:06
`facility|agreement` seems the logical choice. What is `both sets of text` ? — , Jun 26 '18 at 20:11
Currently it's hard to get your intentions. Please provide an example. — revo, Jun 26 '18 at 20:13
I need to find document that have at least one of the words ((facility services|MAFS|MFA|facility|facilities) AND at least one of the words in (agreement|lease) — Rochelle Sides, Jun 26 '18 at 20:23
There's no need for both `facility` and `facility services`. If it matches `facility services`, it also matches `facility` by itself. — Barmar, Jun 26 '18 at 20:26
Never heard of that, is it http://activenavigation.com/? What language do you program in when you use it? — Barmar, Jun 26 '18 at 20:27
They can be in any order, Active Navigation is a file analysis software, that has a UI to write regex to find groups of words within a document in order to identify document type — Rochelle Sides, Jun 26 '18 at 20:30
I used this expression: (facility services|MAFS|MFA|facility|facilities) (agreement|lease) and it found zero results. However this is a line of text within the document: FACILITIES EXPANSION AND RISK SHARING AGREEMENT — Rochelle Sides, Jun 26 '18 at 20:31
The need for a matching both sets of text is to eliminate documents that aren't agreements or leases — Rochelle Sides, Jun 26 '18 at 20:33

score 1 · Answer 1 · answered Jun 26 '18 at 20:31

If you're just looking for two matches, just search for both of them in either order using alternation.

((MAFS|MFA|facility|facilities)[\s\S]*(agreement|lease))|((agreement|lease)[\s\S]*(MAFS|MFA|facility|facilities))

If there are more patterns this doesn't scale well because of combinatorial explosion, so lookaheads are the solution. See Regular Expressions: Is there an AND operator?

score 0 · Answer 2 · 2018-06-26T21:00:19.020

This is commonly known as Out-Of-Order matching.

If you get into a situation where you have more than 2 sets, then the only
efficient way to do it is with an engine that does conditional constructs.

But, this is it for your 2 sets (?:.*?\b(?:(?(1)(?!))(facility|MAFS|MFA|facilities)|(?(2)(?!))(agreement|lease))\b){2}

Readable version

 (?:
      .*? 
      \b          
      (?:
           (?(1)
                (?!)
           )
           (                        # (1 start)
                facility
             |  MAFS
             |  MFA
             |  facilities
           )                        # (1 end)
        |  
           (?(2)
                (?!)
           )
           ( agreement | lease )    # (2)
      )
      \b 
 ){2}

And operator in REGEX

2 Answers2