Comment by pavel_lishin

1 day ago

Isn't exactly their complaint? It's called an array, referred to consistently everywhere as an array, but it just ... isn't.

Apparently array is short for associated array:-)

  • In PHP the array index supports different key types and there's various optimizations that need to happen when the indexes are all numeric, mixed, or all strings or anything else. Technically it's called associative as soon as even one of the indexes isn't numeric. Internally though they are always numeric and anything non integer is hashed internally with DJB2 (Daniel J. Bernstein) hash algorithm and then stored. Using a non numeric index is slightly slower for that reason.

    • But even if you only have numeric keys, those keys don't need to be consecutive, or start at 0.

Better than calling it a hash.

  • I don't think it is, tbh.

    Perl's hashes are a complete mystery to me still, but at least it lets me know that it's not just a linear, uh, well, array.

    • > Perl's hashes are a complete mystery to me still

      They're unordered mappings from strings to arbitrary values ("scalars" in Perl jargon). In this sense they're just like an object in JavaScript.

      Where this gets a little weird is that Perl arrays and hashes are fundamental types distinct from scalars - you can't put a hash into a $variable without taking a reference to it first, for instance. But that's more a matter of Perl being picker about the value/reference distinction than a hash-specific thing.