News for hackers. Stuff that matters.
Re:The New ‘Absent Operator’ in Ruby’s Regular Expressions by perlgeek on Thursday March 23, 2017 @10:01AM (#13939764)
I'm pretty sure you can emulate the absent operator with look-ahead expressions. Untested: (?=((?!exp).)*$)

That is: a zero-width assertion where at every character until the end exp doesn't match.

It's a nice idea to include this in the standard, and also illustrates how convoluted regular expression syntax has become over the years, mostly because it wasn't designed to be extensible.

Re:The New ‘Absent Operator’ in Ruby’s Regular Expressions by rntz on Thursday March 23, 2017 @10:55AM (#13940236)
You propose:

    (?=((?!exp).)*$)
I don't believe you're right, but can't be sure, because you haven't given a method of translating an arbitrary regular expression containing a negation (?~E) into one without.

For example, how would you translate:

    (?~exp)a.*
which asks for "anything not exp, then an `a`, then anything"?

    (?=((?!exp).)*$).*a.*
is wrong, since it doesn't match "aexp".

    (?=((?!exp).)*$)a.*
is also wrong, for the same reason.

    (?=((?!exp).)*a)a.*
is wrong, because it doesn't match "expba".

EDIT: argh, HN formatting rules about asterisks.

Re:The New ‘Absent Operator’ in Ruby’s Regular Expressions by mnarayan01 on Thursday March 23, 2017 @11:38AM (#13940697)
I believe he's saying you'd want:

  (?=((?!exp).)*a.*)
I'm a little dubious that full equivalence is maintained (particularly w.r.t. to e.g. #match), but my brain isn't up to handling convoluted REs this early.

Edit: Another hangup I have is how would you do something like:

  ((?~E1)|E2)E3
Obviously you can distribute by duplicating E3, but then it's pretty clear what the absent operator buys you.
Re:The New ‘Absent Operator’ in Ruby’s Regular Expressions by rntz on Thursday March 23, 2017 @02:23PM (#13942564)

    (?=((?!exp).)*a.*)
is also wrong; it doesn't match "expba", but

    (?~exp)a.*
does.
Re:The New ‘Absent Operator’ in Ruby’s Regular Expressions by mnarayan01 on Thursday March 23, 2017 @11:07PM (#13946285)

  /(?=((?!exp).)*a.*)/ =~ "expba"
  # => 1
for me at least (using #match is also truthy, but the matched string is not what I'd guess the absent operator would give).
Re:The New ‘Absent Operator’ in Ruby’s Regular Expressions by rntz on Sunday March 26, 2017 @07:34PM (#13964016)
Huh! Now I am super confused.
Re:The New ‘Absent Operator’ in Ruby’s Regular Expressions by al2o3cr on Thursday March 23, 2017 @10:48AM (#13940164)
See also the second comment on the original issue:

https://bugs.ruby-lang.org/issues/5588

Re:The New ‘Absent Operator’ in Ruby’s Regular Expressions by runako on Thursday March 23, 2017 @10:08AM (#13939837)
I haven't completely ingested this post yet, but this seems to be incorrect:

> I'm pretty sure you can emulate the absent operator with look-ahead expressions. Untested: (?=((?!exp).)*$)

From TFA: "Note that this is not the same as a negative look-behind or look-ahead — we’ll see how shortly."

Re:The New ‘Absent Operator’ in Ruby’s Regular Expressions by falsedan on Thursday March 23, 2017 @10:30AM (#13940021)
The article explains that (?~foo) is not the same as (?!foo), since the latter fails the match if it matches, but the former succeeds if it can match anything that's not 'foo' (including 'oo'). The first examples anchor the patterns, so that this behaviour:

  "foo" ~= /(?~foo)/ # => 1
isn't revealed.
Re:The New ‘Absent Operator’ in Ruby’s Regular Expressions by perlgeek on Thursday March 23, 2017 @10:26AM (#13939990)
Well, it's not the same as a simple positive or negative look-ahead, but that's not what I proposed either.