Regex - A expert should solve this in seconds

Question

Hi There I were wondering if anyone can help me with a regex

I want to insert the following into a database:

(#text1#,#text2#,#text3#,#text4#,#text5#,#text6#, #text7#, #text8#, #text9#), (#text1#,#text2#,#text3#,#text4#,#text5#,#text6#, #text7#, #text8#, #text9#), (#text1#,#text2#,#text3#,#text4#,#text5#,#text6#, #text7#, #text8#, #text9#);

but sometimes I will not have 9 textfields that I can place into my database

(#text1#,#text2#,#text3#,#text4#,#text5#,#text6#, #text7#, #text8#, #text9#),

(#text1#,#text2#,#text3#,#text4#,#), <<<--- String breaks and mess up my insert

(#text1#,#text2#,#text3#,#text4#,#text5#,#text6#, #text7#, #text8#, #text9#);

Is there a way that I can write a regex to delete if the line is not having a start tag ( and and end * edit ** The lines will always have the start tag (# and always have a closing tag #)

I have tried /^(#.?#,#.?#,#.?#,#.?#,#.?#,#.?#,#.?#,#.?#,#.*?#)$/ig

but do not really have an idea how to do it?

Can anyone please help me...

I have created a page where you can insert a REGEX to see if your solution workds

Answer 1

You could try it with this:

/^\(([\s]*#[^#]+#,?){9}\)[,;]$/

edit:

In perl, if you want to remove occurrences of any pattern of your above set that has less than 9 #\d#'s, you can use the following:

$string =~ s/\(([\s]*#[^#]+#[\s]*,?){0,8}\)[,;]*//g;

It allows for spaces at either end of the #\d#, an optional comma separating them within the parens, and either a comma or a semi after the group. Your resulting $string will be the list of 9-token groups from your input string, as they appear in the original.

Answer 2

/^\((#\w+#,?\s?){9}\)$/ matches exactly 9.

Answer 3

Deleting from the string has some complications, such as what to do with the error line if it is last - contains the ;

line = ^\s*\(.*\)[,;]\s*$
a string token = #[\w\s]*#
a list of tokens = token(?:\s*,\s*token)
7 or less items {0,7}
a list of 8 or less tokens = token(?:\s*,\s*token){0,7}

Making

^\s*\(#[\w\s]*#(?:\s*,\s*#[\w\s]*#){0,7}\)\s*[,;]\s*$

with which you want to replace these lines with nothing globally treating the string as multiple lines /match/replace/gm

/^\s*\(#[\w\s]*#(?:\s*,\s*#[\w\s]*#){0,7}\)\s*[,;]\s*$//gm

If you have set your string character to # for the purpose of the insert then the token could be simplified to #[^#]+#

In your example short line the last token only has the one # which I have not allowed for here so far, nor an entirely empty element which might be acceptable to your sql parser.