Regular expression not capturing optional group - regex

Question

I'm using the following regular expression pattern:
.*(?<line>^\s*Extends\s+#(?<extends>[_A-Za-z0-9]+)\s*$)?.*
And the following text:
Name #asdf
Extends #extendedClass
Origin #id
What I don't understand is that both of the caught group results (line and extends) are empty, but when I remove the last question mark from the expression the groups are caught.
The line group must be optional since the Extends line is not always present.
I created a fiddle using this expression, which can be accessed at https://regexr.com/4rekk
EDIT
I forgot to mention that I'm using the multiline and dotall flags along with the expression.

Answer 1

It's already been mentioned that the leading .* is capturing everything when you make your (?<line>) group optional. The following is not directly related to your question but it may be useful information (if not, just ignore):
You need to be careful elsewhere. You are using ^ and $ to match the start and end of lines as well as the start and end of the string. But the $ character will not consume the newline character that marks the end of a line. So:
'Line 1\nLine 2'.match(/^Line 1$^Line 2/m) returns null
while
'Line 1\nLine 2'.match(/^Line 1\n^Line 2/m) returns a match
So in your case if you were trying to capture all three lines, any of which were optional, you would write the regex for one of the lines as follows to make sure you consume the newline:
/(?<line>^\s*Extends\s+#(?<extends>[_A-Za-z0-9]+)[^\S\n]*\n)?/ms
Where you had specified \s*$, I have [^\S\n]*\n. [^\S\n]* is a double negative that says one or more non non-white space character excluding the newline character. So it will consume all white space characters except the newline character. If you wanted to look for any of the three lines in your example (any or all are optional), then the following code snippet should do it. I have used the RegExp function to create the regex so that it can be split across multiple lines. Unfortunately, it takes a string as its argument and so some backslash characters have to be doubled up:
let s = ` Name #asdf
Extends #extendedClass
Origin #id
`;
let regex = new RegExp(
"(?<line0>^\\s*Name\\s+#(?<name>[_A-Za-z0-9]+)[^\\S\\n]*\\n)?" +
"(?<line>^\\s*Extends\\s+#(?<extends>[_A-Za-z0-9]+)[^\\S\\n]*\\n)?" +
"(?<line2>^\\s*Origin\\s+#(?<id>[_A-Za-z0-9]+)[^\\S\\n]*\\n)?",
'm'
);
let m = s.match(regex);
console.log(m.groups);
The above code snippet seems to have a problem under Firefox (an invalid regex flag, 's', is flagged on a line that doesn't exist in the above snippet). See the following regex demo.
And without named capture groups:
let s = ` Name #asdf
Extends #extendedClass
Origin #id
`;
let regex = new RegExp(
"(^\\s*Name\\s+#([_A-Za-z0-9]+)[^\\S\\n]*\\n)?" +
"(^\\s*Extends\\s+#([_A-Za-z0-9]+)[^\\S\\n]*\\n)?" +
"(^\\s*Origin\\s+#([_A-Za-z0-9]+)[^\\S\\n]*\\n)?",
'm'
);
let m = s.match(regex);
console.log(m);