Finding sentences with open parenthesis that and not closed

[Robin Cruise]

good day. How do I find unclosed parenthesis?

ex. What is taxe d'habitation and (do I have to pay it?...

after ? should be )

[Scott Sumner]

@Robin-Cruise

Perhaps something like this, if your ( and your sentence-ender (e.g. ? but also could include . or ! or…) is on the same line:

Find-what zone: (?-s)\(.+?[?!.]
Replace-with zone: $0\)
Search mode: Regular expression

[Scott Sumner]

Ach…that doesn’t seem to quite do it…maybe this one is better:

Find-what zone: (?-s)\(.+?(?:(\))|[.?!])(?!\))
Replace-with zone: $0?1:\)

…or maybe I should just give up… :-)

[Robin Cruise]

hello Scott Sumner. Thanks for answer. Your regex seems to be fine for my exemple, but it’s not about search and replace, only search I need.

because in another case, such as this:

What is taxe d'habitation and (do I have to pay its products anytime?

your regex will add a close parenthesis of the end of sentence (but not where I want it). Not so good. So I need only to find the sentences with open parenthesis that and not closed. Just to find them, not to close them.

[guy038]

Hello @Robin-cruise, @scott-sumner and All,

From your last post and assuming that :

• A line may contain several (....) correct blocks

• Any (....) block is contained in a single line, only

After some tests and building complicated regexes, I thought, as you only need to perform a search, that the best regex would be, simply :

SEARCH \([^(\r\n]*?\)|\(|\)

This regex matches any correct (....) block OR any non-balanced ( OR ) parenthesis !

Remark : As correct blocks are searched first, this, automatically, avoids to find their boundaries themselves. Thus, the regex engine correctly looks for any non-balanced remaining parenthesis ;-))

Give it a try, with the sample test, below :

This is a sentence with TWO consecutive blocks between parentheses            00 00
This is a sentence with TWO consecutive blocks between ) parentheses          00 01
This is a sentence with TWO consecutive ( blocks between parentheses          00 10
This is a sentence with TWO consecutive ( blocks between ) parentheses        00 11

This is a ( sentence with TWO consecutive blocks between parentheses          10 00
This is a ( sentence with TWO consecutive blocks between ) parentheses        10 01
This is a ( sentence with TWO consecutive ( blocks between parentheses        10 10
This is a ( sentence with TWO consecutive ( blocks between ) parentheses      10 11

This is a sentence with TWO ) consecutive blocks between parentheses          01 00
This is a sentence with TWO ) consecutive blocks between ) parentheses        01 01
This is a sentence with TWO ) consecutive ( blocks between parentheses        01 10
This is a sentence with TWO ) consecutive ( blocks between ) parentheses      01 11

This is a ( sentence with TWO ) consecutive blocks between parentheses        11 00
This is a ( sentence with TWO ) consecutive blocks between ) parentheses      11 01
This is a ( sentence with TWO ) consecutive ( blocks between parentheses      11 10
This is a ( sentence with TWO ) consecutive ( blocks between ) parentheses    11 11

Cheers,

guy038

[Scott Sumner]

@Robin-Cruise said:

but it’s not about search and replace, only search I need.

You are right; I misread your original post. But…if someone gives you a search+replace expression, can’t you just use the search part if that’s all you want to do? :-)

@guy038, likely overkill… :-D

[Robin Cruise]

thank you guy038 .

But in case I want to find only the LINES that contains only one single parenthesis, not both of them? Like in this case:

This is a sentence with open parentheses  ( blocks between  ... 
This is a sentence with close parentheses  ) blocks between ...

not this

This is a ( sentence with TWO ) consecutive blocks between parentheses        11 00

[guy038]

Hi, @Robin-cruise and All,

In case of an UNIQUE expected block of text, between parentheses, strangely, the regex seems a bit more complicated !

SEARCH ^[^(\r\n]*\K\)|\((?!(?-s).*\))

Notes :

• This regex has two alternatives, separated with the | regex symbol :

  • ^[^(\r\n]*\K\), which searches for a ) character, if NO ( character, nor \r or \n has been found, before, from beginning of current line

  • \((?!(?-s).*\)), which searches for a ( character, if NO ) character can be found, further on, on current line

Just test it with the sample text, below :

This is a sentence with ONLY ONE block between parentheses          00
This is a sentence with ONLY ONE block ) between parentheses        01
This is a sentence ( with ONLY ONE block between parentheses        10
This is a sentence ( with ONLY ONE block ) between parentheses      11

)                                                                   01
(                                                                   10
()                                                                  11

Cheers,

guy038

[Robin Cruise]

thank you

[Robin Cruise]

and on small thing, if I want to exclude all the lines that contains ); how cand I do this?

 This is a sentence with ( ONLY ONE block  between parentheses       
 This is a sentence with ONLY ONE block ) between parentheses

NOT THIS:

 This is a sentence with ONLY ONE block between parentheses  );

I try to add this to your regex, but doesn’t work (?![\);])

^[^(\r\n]*\K\)|\((?!(?-s).*\))(?![\);])

[guy038]

Hi, @Robin-cruise and All,

In that case, the search of the ending parenthesis ( ) ) must have the additional condition that is NOT be followed with a semicolon ( ; ). Thus, the negative look-ahead (?!;) must be added after the literal ending parenthesis \)

So, the regex becomes :

SEARCH ^[^(\r\n]*\K\)(?!;)|\((?!(?-s).*\))

BR

guy038

P.S. :

And, with the text, below :

This is a sentence ( with ONLY ONE block between parentheses );

Do you expect to match the opening parenthesis ( ( ) or to ignore it ? Presently, it does not match the (, because of the regex (?!;) !

[Robin Cruise]

and, in the future, if I want to use negative look-ahead for other sings like },{ or [,] what negative look-ahead should I use?

because, just for testing, I try to change ); with )} so as to find those lines that contains only )}

 This is a sentence with ONLY ONE block between parentheses  )}

so, in this case I should use something like this (?!\}), correct?

^[^(\r\n]*\K\)(?!\})|\((?!(?-s).*\)) but is not working

[Robin Cruise]

got it (?![\}]) or (?![\{])

So this will find all lines that contains only curly bracket ); or ){ or )}

^[^(\r\n]*\K\)(?!;)(?![\}])(?![\{])|\((?!(?-s).*\))

Thank you very much Guy038

[guy038]

Hi, @Robin-cruise and All,

Of course, you may add these 3 look-aheads, consecutively, after the litteral \) ending parenthesis, as you did :

(?!;)(?![\}])(?![\{])

Indeed, while evaluating each condition, in each look-ahead, the regex engine location does NOT change ( It is just between the ) and its next character ! )

However, you can, also, use the unique look-head (?![;{}]) ! In addition, when inside a character class [....], the { and } braces are just literal characters :-))

Recall :

Inside a character class [....], 4 characters, only, have a special meaning :

• The character ^, which must be at any position but the first, to be considered as literal or at any position if preceded with the \ escape symbol

• The character ], which must be the very first character, after ], to be taken as literal or at any position if preceded with the \ escape symbol

• The character -, which must be at the very beginning or at the the very end of the character class to be considered as literal or at any position if preceded with the \ escape symbol

• The character \, which can be at any position of the character class, if preceded, itself, with an other \ escape symbol, to be taken as a literal character

• All the other chracters, inside a character class [....], are just literal chars !

---

To sum up, assuming an unique block (.....) per line, the regex ^[^(\r\n]*\K\)(?![;{}])|\((?!(?-s).*\)) would find :

• The ending ) parenthesis, if not followed with a ;, a } or a { character AND if a ( parenthesis has not been found, before, in current line

• The starting ( parenthesis, if a ) parenthesis cannot be found, further on, in current line

Cheers,

guy038