Re: AI Overlords

[morgon]

My perl tells me:

No such class new at poem line 20, near "To our new"
syntax error at poem line 20, near "To our new"
Execution of poem aborted due to compilation errors.

So amazingly only the last line actually produces an error, but nevertheless I'd say that is enough to disqualify it as Perl poetry (even though I also welcome Alphago's first win today).

[hippo]

So amazingly only the last line actually produces an error

Without semicolons this is all just one statement. The single statement has a lot of commas which IIRC are evaluated from right to left and therefore the last line of the prose is the first phrase that perl will actually try to parse in any detail.

[morgon]

I was surprised that it would not bail out at the full-stops.

Is there any way to see how perl attempts to parse a script that eventually turns out not to be valid (B::Deparse will not help then I suppose...)

[hippo]

I was surprised that it would not bail out at the full-stops.

The full stop is the concatenation operator. This is valid perl:

$ perl -ce 'a . b;'
-e syntax OK

If your only punctuation is the full stop and the comma you can get away with most things. Just avoid reserved words near punctuation.

$ perl -ce 'Twas brillig, and the slithy toves Did gyre and gimble in 
+the wabe; All mimsy were the borogoves, And the mome raths outgrabe'
-e syntax OK

Of course, you couldn't end that with a full stop because it's a binary operator. HTH

[Anonymous Monk]

enough to disqualify it as Perl poetry

Just imagine an __END__ at the top of the poem, and its valid perl --- now imagine having to actually write the __END__ as part of the poem is silly

[morgon]

Ah right.

Implicitly assuming an __END__ at the beginning of a poem creates a whole new world of poetry.

So eg the whole quran is in reality just one big perl poem.

[hdb]

Oh yes! But then imagine all the poems with an implicit __DATA__ at the beginning.