Help with Regex

Pangolin

My knowledge with Regex is limited to add a text at the beginning of a line or the end of it.

So I tried with ChatGpt to get a code to do what I want but without any success. I hope you will understand my explanations on what I’m looking to do.

First EOL conversion is on LF.

I have over text with over 30 000 lines.

With two kinds of listing

First :

Movie Title
Director
Actor 1
Actor 2
Actor 3
and so on

Second

Movie Title
Actor 1
Actor 2
Actor 3
and so on

What I want to do is:
When “Director” is present remove the first line “Movie Title” and all the lines that begins with “actor”.

If “Director” is not present. Keep the paragraph.

Thanks for your help

Mark Olson

@Pangolin
Looks like this should work:
Replace (Movie Title:.+(?:\R|\Z))(Director:.+(?:\R|\Z))?((?:Actor.*?:.+(?:\R|\Z))*) with ?2\2:\1\3
Make sure . matches newline is turned off.

Explanation:

(?:\R|\Z) matches any newline (including, CR, CRLF, LF) or the end of the file.
There are three capture groups, one for the movie title, one for director, and one for zero or more actors.
In the replacement, the syntax ?2\2|\1\3 means we should replace with the second capture group (director) only if director was matched. Otherwise, we should replace with the other two capture groups (movie title and any number of actors).

I was able to convert

Movie Title: Baz to the Bone
Director: Bazeven Bazlberg
Actor 1: baz bazsson
Actor 2: foo foosson
Actor 3: bar barsson

Movie Title: Foo Manchu
Actor 1: foo foostein
Actor 2: bar barstein
Actor 3: baz bazstein

Movie Title: Conan the BarBarian
Director: Baron Dracula
Actor 1: baz bazmeister
Actor 2: foo foomeister
Actor 3: bar barmeister

to

Director: Bazeven Bazlberg

Movie Title: Foo Manchu
Actor 1: foo foostein
Actor 2: bar barstein
Actor 3: baz bazstein

Director: Baron Dracula

Note that a ? after the second capture group is optional for this use case. The find/replace works fine either way, and performance is slightly faster if you don’t include the ?.

However, not including the ? comes at the cost of failing to match anything where the director is not listed, which could be problematic for related use cases.

guy038

Hello, @pangolin, @mark-olson and All,

Here is an other formulation ! Just one constraint : add one / two line-break(s) at the very end of your file !

SEARCH (?-si)^Movie Title.*\RDirector.*\R(?:Actor.*\R)*\R+

REPLACE Leave EMPTY

Notes :

The strings Movie Title, Director and Actor must have this exact case
The strings Movie Title, Director and Actor can be followed or not with a name
The string Actor may not be present in a paragraph block
The end of each block to delete may contain several consecutive line-breaks

Test it, for instance, against this text :

Movie Title
Director
Actor
Actor
Actor




Movie Title
Actor 1
Actor 2
Actor 3

Movie Title
Director

Movie Title
Director
Test 1
Test 2
Test 3


Movie Title
Director ...
Actor ...
Actor ...
Actor ...

Best regards,

guy038

Pangolin

Thank you very much Mark and Guy for taking the time to answer my request. It is really appreciated!

Have a nice evening