How to extract last occurence of string, between two delimiters from multiple delimiters, using regex?

Question

I am trying to extract

`Something is "670px ?
am"` 

from my input string:

`100000000670
? "Result was 670670px"
? Something is "670px ?
am"
? `

But I can't seem to get it quite right.

If I use the following regex:

/\n\?(.*?)\n\?\s*$/s

Then the extracted string is:

`"Result was 670670px"
? Something is "670px ?
am"`

And If i use:

/\n\?((?!\n\?)*?)\n\?\s*$/s

Then there is no match. I've tried other regexes, but none seem to do the trick.

I want to extract all the characters that occur between the last two question marks that start on a newline. There may be other question marks present in the characters to be extracted. The input string always ends in a question mark followed by a space. What regex will extract the relevant string?

Answer 1

I want to extract the last string that occurs between two question marks which start on a newline. There may be other question marks in the middle but I want to ignore those. The input string always ends in a question mark followed by a space.

 const input = `100000000670 ? "Result was 670670px" ? "670px ? am" ? `; const regex = /\\n\\?((?:[^\\n]|\\n(?!\\?))*\\n)\\? $/; const match = input.match(regex); console.log(match[1]);

If you want to exclude the leading space and trailing newline just use trim() or trivially adjust the regex. I am just following the letter of the quote I included above, and interpreting "last string" as all characters between the question marks.

Further explanation on request.

Answer 2

To match the last string that occurs between two question marks which start on a newline:

\n\?[^\S\n]*(.*(?:\n(?!\?).*)*)\n[^\S\n]*\?[^\S\n]*$

• \\n\\? Match a newline and ?
• [^\\S\\n]* Match optional whitespace chars without a newline
• ( Capture group 1
  • .* Match the rest of the line
  • (?:\\n(?!\\?).*)* Match all lines that do not start with ?
• ) Close group
• \\n[^\\S\\n]*\\?[^\\S\\n]* Match a newline and a ? between optional spaces
• $ End of string

Regex demo

 const regex = /\\n\\?[^\\S\\n]*(.*(?:\\n(?!\\?).*)*)\\n[^\\S\\n]*\\?[^\\S\\n]*$/; const s = `100000000670 ? "Result was 670670px" ? Something is "670px ? am" ? `; const m = s.match(regex); if (m) { console.log(m[1]); }

Answer 3

This would work:

"[^"]+"(?=\s+\? $(?![\r\n]))

• "[^"]+" - get double-quoted content
• (?=\\s+\\? $(?![\\r\\n])) - ahead of me must be the question mark + space which terminates the string
  • JS does not support the "end of string" meta-character \\Z so we emulate it with $(?![\\r\\n]) which translates to "no more line breaks after the end of the line"

https://regex101.com/r/lzd2gY/1/

Answer 4

One strategy might be to split the string based on the delimiters, creating an array. Then, pop the last element off of the array. (Or maybe the next-to-last.)

An advantage of such logic, if it applies here, would be that it is fairly obvious what is happening. "Regex spaghetti-code" is a thing to be avoided whenever possible.